many bluray/hd-dvd titles are released as 16bit soundtracks instead of 24bit. although master is usually 24bit, would it make a difference to consumers?


1) is there any data that supports 24bit superiority? journals? blind tests? abx tests of any sort?


2) is it realistic to expect a difference with consumer level equipment?

16bit gives about 97 dB
24bit gives about 145 dB

denon AVR-2307CI $799 has 100db
denon AVR-5805CI $7,200.00 has 117db analog


***3) most importantly, is it realistic to expect a difference under real world movie viewing conditions (i.e. with the volume not turned way up)

24bits adds resolution of low level signals. for music i can see how the low resolution would matter, but for movies people typically normalize the volume for dialogue.




does anyone even have anecdotal evidence? if anyone does believe 24bit makes a difference

please describe how you came to that conclusion:

1) what equipment used
2) what materials listening to (hd-dvd/blu-ray/xyz, what time point)
3) were you listening at normal volume (normal=whatever volume *you* would watch the whole movie at)
4) nature of difference you heard


or do people simply want 24bit audio for peace of mind, like monster cables?

Personally, I couldn't care less about 24-bit audio on HD DVD/Blu-ray.

20-bit might be a bonus, but truthfuly 16-bit is fine for me.

Same goes for 96 KHz audio. I don't care about it. 48 KHz is fine as far as I'm concerned.

I guess what I'm saying is that 48 KHz/20-bit seems like the best bang for the bits.

I guess what I'm saying is that 48 KHz/20-bit seems like the best bang for the bits.

I agree with you on 20/48 audio. While the honest truth is for 99% of consumers 16/48 is fine bumping it to 20/48 doesn't cost much in terms of bits and makes the media capable of higher resolution if and when receivers are.

24bit is nothing but marketing.
96KHz would be nice but the problem is that unlike the jump from 16 -> 20 or 20->24 the change between 48Khz and 96Khz requires double the bandwidth and double the space for very little benefit. IMHO that space would be better spent on higher video bitrates, extras, or additional languages.

There's a definite jump in audio quality going from 16/44.1 (CD) to 24/96 (DVD-A) (and just a wee bit more for 24/192). Too much information is lost at the lower bit rates particularly for well recorded classical/jazz. How applicable this is for movie soundtracks or whether a "tweener" sampler rate (e.g. 20/48) would work better for movies is an open question which is probably highly dependent on the resolving power of the playback system.

But to remove any doubt, I'd want the audio to be at the same level as the original master (24/48 for most studio recordings ?). I'm hoping the recording labels start releasing HD video music concerts with 5.1 24/96 soundtracks. Worried about available space on disc ? Use Blu-ray !

Worried about available space on disc ? Use Blu-ray !

No I am not worried! Use HD DVD! ;)

Yeah, for concerts and stuff I can see them using 24/96, even if just for marketing purposes.

For movies though I think it's a waste of space.

No I am not worried! Use HD DVD! ;)

I keep looking at King Kong as an example why I don't like HD-DVD. They created a reference video demo disc but couldn't put in a TrueHD or PCM soundtrack because of space constraints. Maybe in the future we'll have reissue using a dual layer Blu-Ray disc with 24/48 uncompressed PCM soundtrack.

I keep looking at King Kong as an example why I don't like HD-DVD. They created a reference video demo disc but couldn't put in a TrueHD or PCM soundtrack because of space constraints. Maybe in the future we'll have reissue using a dual layer Blu-Ray disc with 24/48 uncompressed PCM soundtrack.

This kind of smear is getting old. Did any other Universal title around that period have a lossless soundtrack? NO, so why would you assume it just could not fit?

But lets take this one step further, the primary track that is on there is a 1.5Mbit/s DD+ track at 24bit/48kHz. So actually they could have easily fit a TrueHD lossless track at 16bit/48kHz as they average about 1.4Mbit/s.

But back to the topic of the thread, I actually prefer the high bitrate lossy (DD+ @ 1.5) encode with 24bit resolution, over lossless at only 16bit.

I keep looking at King Kong as an example why I don't like HD-DVD. They created a reference video demo disc but couldn't put in a TrueHD or PCM soundtrack because of space constraints. Maybe in the future we'll have reissue using a dual layer Blu-Ray disc with 24/48 uncompressed PCM soundtrack.


source?

16 bit is 88db or so.
24 bit is 105 db or so.

There really is no such thing as 20 bit audio, so 24 is better use of the same space.

16/48 PCM is usually "lossy" by a huge amount when compared to the master.

16 bit is 88db or so.
24 bit is 105 db or so.

There really is no such thing as 20 bit audio, so 24 is better use of the same space.

16/48 PCM is usually "lossy" by a huge amount when compared to the master.

Can you explain what you mean there is no such thing as 20 bit audio. All audio is analog and can be represented by digital values. A wider value range (2^24 vs 2^16) results in more precision. Kinda like trying to express pi as a decimal. More decimal places equals more precision. Saying there is no such thing as 20bit audio is like saying pi can ONLY be expressed to 16 or 24 places. No other values are pi.

There is nothing magical about 16, 24, or 32 bit values. You could have 27bit audio is you really wanted to.

Can you explain what you mean there is no such thing as 20 bit audio. All audio is analog and can be represented by digital values. A wider value range (2^24 vs 2^16) results in more precision. Kinda like trying to express pi as a decimal. More decimal places equals more precision. Saying there is no such thing as 20bit audio is like saying pi can ONLY be expressed to 16 or 24 places. No other values are pi.

There is nothing magical about 16, 24, or 32 bit values. You could have 27bit audio is you really wanted to.

20 bit audio gets stored as 24 bit audio with a 4 bit pad. It takes up the same amount of space, and it's all that useful in this situation, and 24 bit convertors only have 20 bits of resolution anyway. How's that for confusing.

There really is no such thing as 20 bit audio, so 24 is better use of the same space.Except that HDCD made an attempt at getting 20 bits -- kinda in a virtual sense.

http://en.wikipedia.org/wiki/HDCD

20 bit audio gets stored as 24 bit audio with a 4 bit pad. It takes up the same amount of space, and it's all that useful in this situation, and 24 bit convertors only have 20 bits of resolution anyway. How's that for confusing.
I remember Amir commented on this before, BR supprts 16, 24, but not 20 bits; but HD DVD supprts 20 bits. So 20 bits is useful on HD DVD but not BD.

I want lossless, plain and simple.

I want lossless, plain and simple.
Well, that leaves you with very few options for movie watching ATM. Hell, 99% of CDs aren't "lossless".

does anyone even have anecdotal evidence? if anyone does believe 24bit makes a difference

please describe how you came to that conclusion:

1) what equipment used
2) what materials listening to (hd-dvd/blu-ray/xyz, what time point)
3) were you listening at normal volume (normal=whatever volume *you* would watch the whole movie at)
4) nature of difference you heard


or do people simply want 24bit audio for peace of mind, like monster cables?
In the absence of answers, can we assume the supporters of 24 bit use Monster or better cables too? :rolleyes:

The consistent lack of proof proves to me it isn't better.

^^^ Don't forget the power cord when you are redoing interconnects. :rolleyes:

In the absence of answers, can we assume the supporters of 24 bit use Monster or better cables too? :rolleyes:

The consistent lack of proof proves to me it isn't better.
In blind tests people notice the difference between 16 and 24 bit. The differences between 44.1 and 48 are less noticable and very few people notice the difference between 48 and 96 and above.

In blind tests people notice the difference between 16 and 24 bit. The differences between 44.1 and 48 are less noticable and very few people notice the difference between 48 and 96 and above.
Do you have references?

Do you have references?
I use Google.

IMHO the difference between 48 kHz and the higher ones (96 & 192 kHz) is the ease of doing proper low pass filtering without adding phase shift impact. Much easier to do a perfect brick wall filter to just pass 24 kHz analog and below with almost zero phase shift.

But you can throw more $ at a low pass filter and still get the same minimal phase shift impact when only 48 kHz is used for the encode. Plus you can also oversample to 192 kHz just before the A/D conversion and use 192 kHz conversion rate DACS.

Based on my own listening with audiod (e.g. SACDs vs CDs, and other various sources) I doubt most consumers will notice the difference. I can hear a slight difference (a somewhat "fuller" sound) but haven't gotten a friend to set up a blind test yet.

The difference between dd 5.1 (or DD+) and uncompressed PCM/TruHD is night and day, and for that reason I personally would like to see 20 bit audio, but in regards to the OPs question, I doubt most people would benefit.

OTOH, many people don't or can't take advantage of DTS on regular DVDs, and they still seem to be drawn to the "value-added" marketing proposition, so there may be a benefit from the studios point of view (I suppose focus groups could help with this one).

The difference between dd 5.1 (or DD+) and uncompressed PCM/TruHD is night and day, and for that reason I personally would like to see 20 bit audio, but in regards to the OPs question, I doubt most people would benefit.


Would you mind describing the differences?

I use Google.
So do I. I haven't seen a controlled DBT. If you have give me a link.

There's a definite jump in audio quality going from 16/44.1 (CD) to 24/96 (DVD-A) (and just a wee bit more for 24/192).

Bats of the world - unite !

20 bit audio gets stored as 24 bit audio with a 4 bit pad. It takes up the same amount of space, and it's all that useful in this situation, and 24 bit convertors only have 20 bits of resolution anyway. How's that for confusing.
Not always. Its true that Blu-ray has this unfortunate limitation with LPCM, but the other lossless codecs do not have this limitation.

TrueHD and DTS-HD MA can do any number of bits between 16 and 24. If you want 19 bits of resolution, its no problem (no wasted space here).

There's a definite jump in audio quality going from 16/44.1 (CD) to 24/96 (DVD-A) (and just a wee bit more for 24/192).

Can you explain for us the setup an testing methodology that came to allow you to discern the difference between 96Khz sampling (ie. up to 48Khz frequency) and 192Khz sampling rate (96Khz frequency)?

Thanks.

Did any other Universal title around that period have a lossless soundtrack?

The following Universal were Dolby TrueHD:
End of Days and Fear and Loathing in Las Vegas

As with thier recent combos there were a bunch of problems... But this was most likey due to something that Universal was doing since WB & The Weinstein Company have not had any problems. (Of course The Weinstein discs still take about a minute or two to load... But that a completely different issue...)

fitprod

The following Universal were Dolby TrueHD:
End of Days and Fear and Loathing in Las Vegas


My point exactly, those were released over a month after King Kong. At the point in time that KK was released Universal had not released ANY TrueHD titles.

I just think its wrong that people claim that KK is proof that TrueHD tracks don't fit on long titles on HD DVD, as its entirely false on multiple counts as explained in my last post.

Just so we can put the idea to bed here are a couple concepts:

16 bit audio allows a waveform to be described as 65,536 (2^16) discrete values.
20 bit audio allows for 1,048,576 discrete values.
24 bit audio allows for 16,777,216 discrete values.

Dynamic Range is 20* log(2^bit resolution)
So we get 96db for 16bit, 120db for 20bit, and 144db for 24bit.

The receiver's components would need a noise floor of -144db to avoid losing the additional resolution gained. Try finding a consumer grade device with s/n ratio of 144db. Don't both You won't find one. The DACs like any silicon device generate heat and as that causes noise which limits the SNR (signal to noise) and masks the lower order bits gained by the 4 extra bits.

To put it in one sentance any component with less than 120db SNR simply can't see the higher resolution offered by 24bit sampling. Any receiver with less than 96db SNR can't see the higher resolution offered by 20bit sampling.

A 20bit audio file (120db range) and 24bit audio file (144db range) output over a receiver with -105db snr can never have dynamic range >105db.

Even the 120db dynamic range obtained by 20bit is beyond consumer grade components. So why go 20bit? Because the 96db offered by 16bit is rather low for high end systems. A setup with SNR of >96 and <120 would benefit some from the higher resolution.

Since no consumer grade device has SNR >120db there is no technical reason to go beyond 20bit. The studios will release movies in 24 bit for marketing only because consumers buy "stuff" with bigger numbers without ever understanding how insanely complex it is go get a snr that high. They might even put a sticker on the cover that says 16X the audio resolution compared to DVD. Just because it fools J6P don't let it fool you. You will gain nothing from the extra 4 bits except wasted space.

Just so we can put the idea to bed here are a couple concepts:

16 bit audio allows a waveform to be described as 65,536 (2^16) discrete values.
20 bit audio allows for 1,048,576 discrete values.
24 bit audio allows for 16,777,216 discrete values.

Dynamic Range is 20* log(2^bit resolution)
So we get 96db for 16bit, 120db for 20bit, and 144db for 24bit.

The receiver's components would need a noise floor of -144db to avoid losing the additional resolution gained. Try finding a consumer grade device with s/n ratio of 144db. Don't both You won't find one. The DACs like any silicon device generate heat and as that causes noise which limits the SNR (signal to noise) and masks the lower order bits gained by the 4 extra bits.

To put it in one sentance any component with less than 120db SNR simply can't see the higher resolution offered by 24bit sampling. Any receiver with less than 96db SNR can't see the higher resolution offered by 20bit sampling.

A 20bit audio file (120db range) and 24bit audio file (144db range) output over a receiver with -105db snr can never have dynamic range >105db.

Even the 120db dynamic range obtained by 20bit is beyond consumer grade components. So why go 20bit? Because the 96db offered by 16bit is rather low for high end systems. A setup with SNR of >96 and <120 would benefit some from the higher resolution.

Since no consumer grade device has SNR >120db there is no technical reason to go beyond 20bit. The studios will release movies in 24 bit for marketing only because consumers buy "stuff" with bigger numbers without ever understanding how insanely complex it is go get a snr that high. They might even put a sticker on the cover that says 16X the audio resolution compared to DVD. Just because it fools J6P don't let it fool you. You will gain nothing from the extra 4 bits except wasted space.
WHich takes us back to 16 bit converters having an 88 db DR and a 24 bit converter having an 105 db DR. Most preamps, amps and speakers can handel that.

You do bring up another interesting fact that most ignore here.
Most audio before 1994 and after 1999 is 24 bit. There is alot of talk here about 16 bit PCM being "lossless". Let's check that.
Original audio 24 bit 16,777,216 discrete values
New audio 16 bit 65,536 discrete values

New audio loses 16711680 pieces of information to become the PCM "lossless" audio on 90% of BD biscs.

My point exactly, those were released over a month after King Kong. At the point in time that KK was released Universal had not released ANY TrueHD titles.

I just think its wrong that people claim that KK is proof that TrueHD tracks don't fit on long titles on HD DVD, as its entirely false on multiple counts as explained in my last post.

Don't confuse them with fact. :D

"Everyone" knows HD DVD didn't have enough space. Just like the fact that most BD use 16/48 PCM which means they didn't have enough space for 24/48. Apparently 50GB is not enough for high quality audio either. Guess we should throw both BD and HD DVD away and instead distribute movies on 1TB HVD discs. HVD for the win in 2014!

Seriously the idea that it was a decision by the company to pursue high bit DD+ instead of TrueHD is just too hard for the bluboys to believe.

Just so we can put the idea to bed here are a couple concepts:

16 bit audio allows a waveform to be described as 65,536 (2^16) discrete values.
20 bit audio allows for 1,048,576 discrete values.
24 bit audio allows for 16,777,216 discrete values.

Dynamic Range is 20* log(2^bit resolution)
So we get 96db for 16bit, 120db for 20bit, and 144db for 24bit.

The receiver's components would need a noise floor of -144db to avoid losing the additional resolution gained. Try finding a consumer grade device with s/n ratio of 144db. Don't both You won't find one. The DACs like any silicon device generate heat and as that causes noise which limits the SNR (signal to noise) and masks the lower order bits gained by the 4 extra bits.

To put it in one sentance any component with less than 120db SNR simply can't see the higher resolution offered by 24bit sampling. Any receiver with less than 96db SNR can't see the higher resolution offered by 20bit sampling.

A 20bit audio file (120db range) and 24bit audio file (144db range) output over a receiver with -105db snr can never have dynamic range >105db.

Even the 120db dynamic range obtained by 20bit is beyond consumer grade components. So why go 20bit? Because the 96db offered by 16bit is rather low for high end systems. A setup with SNR of >96 and <120 would benefit some from the higher resolution.

Since no consumer grade device has SNR >120db there is no technical reason to go beyond 20bit. The studios will release movies in 24 bit for marketing only because consumers buy "stuff" with bigger numbers without ever understanding how insanely complex it is go get a snr that high. They might even put a sticker on the cover that says 16X the audio resolution compared to DVD. Just because it fools J6P don't let it fool you. You will gain nothing from the extra 4 bits except wasted space.

To add to the above, we have the relative loudness of common sounds:
120dB - Jet takeof @ 200ft
115db - Loud rock concert @ 10ft
95db - Loud classical music
65db - Av. room conversation @ 3ft
50db - quiet suburban house at night
40db - quiet auditorium
25db - quiet whisper @ 5 ft
20db - rustling leaves

There is a difference of
70db between a jet takeoff and a quiet suburban home
80db between a jet takeoff and a quiet auditorium
95db between a jet takeoff and a whisper
100db between a jet takeoff and rustling leaves

My point exactly, those were released over a month after King Kong. At the point in time that KK was released Universal had not released ANY TrueHD titles.Sorry, but that isn't true. They were released before "King Kong". And "Miami Vice" was announced as having TrueHD, but then didn't get it. That came out 3 weeks after "King Kong".

--Darin

Would you mind describing the differences?

Sure. The best example I can think of offhand is Glory Road. If you switch from DD 5.1 to PCM durin the opening game at the highschool gym, the sound opens up tremendously. It sounds like the echos of the bouncing balls and cheering crowd are coming from around you, rather than out of the speakers. I have done a pseudo blind test with this track, where I switched the tracks without telling friends what I was doing, and I've gotten several (colorfully worded) exclamations about how "real" the sound became.

The difference is of a somewhat smaller magnitude with my HD DVD TrueHD tracks, but I don't believe this is any lacking on the part of the format or the codec, rather its that the digital/analog converters (DACs) on the A1 are not quite as good as the ones on the Panny (I can sort of test this soon, as I have some trueHD tracks on BD disks coming).

But the main difference in both cases has to do localization of the sound. With less compressed audio the sound seems to emenate from a larger area. Of course I have $10,000+ invested into my speakers and audio electronics. I would imagine on most setups, for example, that the differences in DACs between the A1 and the Panny would be all but invisible.

I'm also not convinced that there is much of a difference audible even to me with the higher bit-rate PCM (compared to lower rate PCM), but I'd rather the bottleneck be in my equipment (which I can upgrade) than the source, hence my preference for higher sampling rates, even if the audible dfference is currently negligable.

To add to the above, we have the relative loudness of common sounds:
120dB - Jet takeof @ 200ft
115db - Loud rock concert @ 10ft
95db - Loud classical music
65db - Av. room conversation @ 3ft
50db - quiet suburban house at night
40db - quiet auditorium
25db - quiet whisper @ 5 ft
20db - rustling leaves


This piece of data is the complementary key to namechamps data. When you take the s/n ratio of the typical high-end (oxymoron?) component at 110 dB and subtract out the room noise contribution (40-50 dB), you have the best case dynamic range/resolution you could possibly observe with that equipment/room combination: a mere 60-70 dB! The only exception is the extreme amplification scenario, which would only then expose you to hearing loss risks, and possibly the police giving you a housecall.

They could give us 24-bit or 24,000 bit audio, and it will make zilcho difference to the end user who is only able to render the audio in a window of about 12 bits (60-70 dB).

Given that piece of data, it is fairly certain that even 16/44 pcm is a healthy bit of overkill for that listening room scenario. "16/44" may not be glamorous a spec these days, but it remains mathematically relevant for the typical equipment and useage scenarios the typical consumer will encounter.

So we get 96db for 16bit, 120db for 20bit, and 144db for 24bit.Excellent analysis! Completely agree. Thanks! :)

Edit: To put it in perspective IIRC the THX reference standard for movie soundtrack playback had 105 dB SPL peak for normal channels and 115 dB peak for the LFE channel.

Excellent analysis! Completely agree. Thanks! :)

Edit: To put it in perspective IIRC the THX reference standard for movie soundtrack playback had 105 dB SPL peak for normal channels and 115 dB peak for the LFE channel.

Agree if you want, but it's wrong.

Agree if you want, but it's wrong.

What is with the oneliners GoodBooty does your ISP charge you on a per charecter basis?

If you think something is wrong explain why. Give a link, or a compelling arguement. Show an example or a piece of equipment. Do something. The most annoying post are the ones like yours that says something (not exactly sure which part) is wrong with no foundation.

Do you think the dynamic range formula is wrong? If so then you better just not explain since that is a fundamental basis for designs of audio recording, digital conversion (both ADC and DAC), and audio playback systems.

TU nc, ee^3, and even Mr H!!! Next we have to debunk Area 53

WHich takes us back to 16 bit converters having an 88 db DR and a 24 bit converter having an 105 db DR. Most preamps, amps and speakers can handel that.

You do bring up another interesting fact that most ignore here.
Most audio before 1994 and after 1999 is 24 bit. There is alot of talk here about 16 bit PCM being "lossless". Let's check that.
Original audio 24 bit 16,777,216 discrete values
New audio 16 bit 65,536 discrete values

New audio loses 16711680 pieces of information to become the PCM "lossless" audio on 90% of BD biscs.

If your 16 bit DAC is only 88 dB, get another, better one. Just because one DAC model/design has a bad SNR doesn't mean they all do.

Sampling rate: The only reason for a higher sampling rate is to make the analog reconstruction filter cheaper. Humans still can't hear the high frequencies though, so it doesn't make sense to store them (it'd be like storing an UV / infrared color channel for movies). You can just double the sampling rate digitally right before the analog filter.

Btw, is there a 1:1 correspondence between audio SNR and sound pressure level? If so, wouldn't it take a sound system blasting at 96dB in an otherwise completely quiet (0dB room) to even enjoy 16-bit audio to the fullest? Doesn't sound very healthy... according to Wikipedia:
Hearing damage (due to short-term exposure) 120 dB
Hearing damage (due to long-term exposure) 90 dB

If mixed and mastered right and with care, I do believe 24 bit/96 kHz and above resolution does help.

Now, if only we could convince most of the studios to release these masters via lossless compression or LPCM!!!

When you take the s/n ratio of the typical high-end (oxymoron?) component at 110 dB and subtract out the room noise contribution (40-50 dB), you have the best case dynamic range/resolution you could possibly observe with that equipment/room combination: a mere 60-70 dB! The only exception is the extreme amplification scenario, which would only then expose you to hearing loss risks, and possibly the police giving you a housecall.


There are two pieces I want to add to this.

Humans can hear below the noise floor. This is how we hear conversation in a noisy room. This means, we could possibly have a better than 60-70 dB range.

Our ear's frequency response depends on loudness. So, if we just make it louder to get higher dynamic range - it is distorted to our ear - since the sounds would be louder than they would be in their natural surroundings. This means to get real lossless sound we need to turn down the volume - ofcourse that would help in not getting that police visit too.

Some people assume that 96/24 (or whatever) is the magic bullet to audio nirvana. No it isn't. We still need good recording / mastering and good equipment (esp. speakers and that most expensive of all components - the room).

Ofcourse nothing new. I must have posted this thing some 10 times by now.

Oh, one more thing. 24 bit resolution with real microphones (and electronics) is impossible. With 24 bits we are going below the thermal noise of molecules at room temperature.

Yes, we can "hear into the noise", but therein is the realization that you are no longer listening to pure, unadulterated sound, at that point (10 dB into the noise, sure, but 30/40/50+ dB into the noise?...that's getting a "bit" optimistic, right?). All this extra resolution/detail of "24-bit audio" has been mangled and obfuscated by the very noise level of the room. Once you have this noise floor in place, there is no accuracy or determinance of the signal below it (because the room is now the adhoc LSB of the digital waveform). The accuracy/resolution that remains for the reference waveform of this audio piece is essentially what you would have got with "mere" 16-bit sound (12-bits if you want to be precise).

Sure. The best example I can think of offhand is Glory Road. If you switch from DD 5.1 to PCM durin the opening game at the highschool gym, the sound opens up tremendously. It sounds like the echos of the bouncing balls and cheering crowd are coming from around you, rather than out of the speakers. I have done a pseudo blind test with this track, where I switched the tracks without telling friends what I was doing, and I've gotten several (colorfully worded) exclamations about how "real" the sound became.

DD vs. PCM..........Did you use an SPL meter to match levels?

An article with some nice measurements of movie loudness.
http://www.cbc.ca/consumers/market/files/health/loud_movies/index.html

DD vs. PCM..........Did you use an SPL meter to match levels?

Absolutely not, although perhaps I'll give it a shot now that you bring it up. Otoh, I have played with volume, and it's definitely not what provides better localization with sound. Inbdeed, cranking poorly mixed stereo just makes bad localization worse on poor recordings.

Absolutely not, although perhaps I'll give it a shot now that you bring it up. Otoh, I have played with volume, and it's definitely not what provides better localization with sound. Inbdeed, cranking poorly mixed stereo just makes bad localization worse on poor recordings.

No, volume adjustments within a range of typical listening levels won't improve localization or imaging, but it's well understood that when presented with two single-sourced samples differing only in volume, listeners consistently express a preference for the louder example. The effect is quite pronounced, and has been observed even when the relative SPL differs by only .1 dB. Unfortunately, that level of precision isn't practically achievable with inexpensive, consumer-grade instruments.

Sure. The best example I can think of offhand is Glory Road. If you switch from DD 5.1 to PCM durin the opening game at the highschool gym, the sound opens up tremendously. It sounds like the echos of the bouncing balls and cheering crowd are coming from around you, rather than out of the speakers. I have done a pseudo blind test with this track, where I switched the tracks without telling friends what I was doing, and I've gotten several (colorfully worded) exclamations about how "real" the sound became.

To me this sounds more like they mastered the soundtrack a little differently. Did they create the lossless soundtrack and then do the encode with no other changes, or is it possible that they did some different processing to each. If the two were prepared separately, then I bet you'd get people who prefer one and people who prefer the other, aside from fidelity. I have a friend who always chooses the DTS soundtrack on regular DVDs because he thinks it's the best. I've found that when I switch between DD 5.1 and DTS that sometimes I prefer DTS and sometimes I prefer DD 5.1.

Btw, is there a 1:1 correspondence between audio SNR and sound pressure level? If so, wouldn't it take a sound system blasting at 96dB in an otherwise completely quiet (0dB room) to even enjoy 16-bit audio to the fullest? Doesn't sound very healthy... according to Wikipedia:
Hearing damage (due to short-term exposure) 120 dB
Hearing damage (due to long-term exposure) 90 dB

Not exactly. SPL is based on RMS or Root Mean Squared. It is the square root of the average of the instantaneous values squared. So it would be possible for a movie to have an SPL of say 90 and have some peaks of of 110+. Still the track would need to be played at a level that would likely have your neighbors calling the cops. eecubed had a good article on what levels movie theaters play content at and even at those levels 70-80db SPL with peaks up to 95db it is unlikely they are getting more than low 100db of dynamic range at best. Well < than 20bit resolution.

The other issue is that in mixing most studios chose to intentionally compress the dynamic range anyways. The reason is that otherwise the loudness in quiet scenes (two people whispering) would seem too quiet for listeners and too loud in loud scenes (large explosions).

Yet another issue is the actual sound recording. The best mikes have a dynamic range of around 130db which is less than what would be needed to true recording @ 24bit.

So put it all together
1) recording: mikes don't have enough dynamic range
2) content: most movies just don't have content that has a dynamic range >100-120db. Music maybe but not movies.
3) ADC: The absolute max resolution is based on the snr of the studios Analog to Digital Converter. My understanding is that this is in the 120db range so despite it being 24bit you are not getting anymore digital information than a 20bit recording.
4) mixing: studios compress dynamic range to make movie enjoyable for many systems
5) playback: DAC in most consumer grade equipment <105db SNR
6) listening environment: most consumers don't listen to movies in a sound stage and thus have environmental issues that increase noise floor.
7) maximum volume: to get the full dynamic range loudness would need to be increased to a level that would not be possible, enjoyable, or legal for most consumers.

If any part of the chain is broken then the track may have 24bits in it but you aren't getting a 24bit resolution. Honestly 18bit (108db) would be enough because of the DAC limitations on consumer grade equipment. Since 18 isn't a very "sellable #" 20 bit would be a good compromise between marketing and value.

Despite all this I still honestly think within a few years most tracks will be labeled 24bit since it only costs about 20% more bits and sounds very "audiophile" and it will be a marketable way to make HD seems better than DVD. It will impress J6P and even many people on AVS. If it makes you feel good fine but remember if any part of the chain above is broken you simply are not getting 24bit resolution.

I keep looking at King Kong as an example why I don't like HD-DVD. They created a reference video demo disc but couldn't put in a TrueHD or PCM soundtrack because of space constraints. Maybe in the future we'll have reissue using a dual layer Blu-Ray disc with 24/48 uncompressed PCM soundtrack.


Or Universal studios doesn't put Dolby TrueHD tracks on any of their movies. Only one of the 97 Universal movies released or soon to be released has a TrueHD track. Universal simply doesn't do the TrueHD thing. Kinda like how Warner Bros. didn't put lossless on Happy Feet or Superman Returns on Blu-Ray, does that mean they didn't have the space?

No one's hearing or equipment is going to resolve 24-bit audio.

A better focus for this thread would be whether you or your equipment can ever resolve more than 16-bit. If at any point during a film, your equipment or ears can resolve the differences between 16-bit and 18-bit or 20-bit sound, then there is a benefit offered by 24/48 tracks.

No one's hearing or equipment is going to resolve 24-bit audio.

A better focus for this thread would be whether you or your equipment can ever resolve more than 16-bit. If at any point during a film, your equipment or ears can resolve the differences between 16-bit and 18-bit or 20-bit sound, then there is a benefit offered by 24/48 tracks.
There is way more to sound than dynamic range. In every test of music I have seen people can tell the difference in 16 and 24 bit music. I have never seen a test on movies.

Kinda like how Warner Bros. didn't put lossless on Happy Feet or Superman Returns on Blu-Ray, does that mean they didn't have the space?


no, but then again what every HD DVD supporter misses is that it is not magic. disk capacity- disk content tels us how much is left both of those numbers are easy to get for any title. I don't know about Happy feet (did not see its numbers) but SM has tons of unused space, enough for even PCM let alone DTHD or DTS MA. As for Happy feet even if you assume it did not on the BD-25, then they could have gone with BD-50 and there would have been more then enough space. On the other hand KK is on an HD DVD 30 and there is not enough place for DTHD, so even if they wanted too, they copuld not have added it unless they got rid of something (lower video, get rid of some IME, get rid of other languages.....)

In other words WB could have released those two on BD with lossless added to them but Universal could not release KK on HD DVD with lossless added to it.

There is no magic if one can or cannot it is simple math E+L<=30 (E=Existing on diskm and L=lossless) or E+L<=25 or E+L<=50
we have E for every movie and evenb though L can varry the range is relatively easy to find

There is way more to sound than dynamic range. In every test of music I have seen people can tell the difference in 16 and 24 bit music. I have never seen a test on movies.
Here is a great opportunity to convince a lot of people of the benefit of 16+ bit encoding. Share with us the conditions of the test.

no, but then again what every HD DVD supporter misses is that it is not magic. disk capacity- disk content tels us how much is left both of those numbers are easy to get for any title. I don't know about Happy feet (did not see its numbers) but SM has tons of unused space, enough for even PCM let alone DTHD or DTS MA. As for Happy feet even if you assume it did not on the BD-25, then they could have gone with BD-50 and there would have been more then enough space. On the other hand KK is on an HD DVD 30 and there is not enough place for DTHD, so even if they wanted too, they copuld not have added it unless they got rid of something (lower video, get rid of some IME, get rid of other languages.....)

In other words WB could have released those two on BD with lossless added to them but Universal could not release KK on HD DVD with lossless added to it.

There is no magic if one can or cannot it is simple math E+L<=30 (E=Existing on diskm and L=lossless) or E+L<=25 or E+L<=50
we have E for every movie and evenb though L can varry the range is relatively easy to find

Yeah they could, just because they didn't doesn't mean they couldn't. Like I said before, Universal doesn't include lossless tracks in their movies. Only one of Universals 97 HD DVD movies has Dolby TrueHD. They aren't going to change the way they release movies because a few Blu-Ray fanboys complain about a 30 GB disc.

...but Universal could not release KK on HD DVD with lossless added to it.

What is the bit rate of the DD+ soundtrack on King Kong?

Not to my Ear
LOL
;)

Yeah they could,

how show me the math
Like I said before, Universal doesn't include lossless tracks in their movies. Only one of Universals 97 HD DVD movies has Dolby TrueHD.
and none of that shows how they could
They aren't going to change the way they release movies because a few Blu-Ray fanboys complain about a 30 GB disc
what does one have to do with the other.

What is the bit rate of the DD+ soundtrack on King Kong?

why?

how show me the math

YOU show ME the math.

I asked a question. What is the bit rate of the DD+ soundtrack on the HD DVD of King Kong? I think I know the answer, but I'm not certain. When I am certain, I may be able to show you the math.

from what I hear 1.5mbps (but then again that would be the max)

What is the bit rate of the DD+ soundtrack on King Kong?
24/48

Okay, I found what I was looking for...

King King carries a Dolby Digital Plus soundtrack with 24-bit/48 kHz resolution at 1.5 mb/s.

According to the Dolby info bfdtv has been posting...

Further information, according to Dolby Labs:

2 hours @ 5.1 16/48 TrueHD = 1.26 Gbytes @ 1.4 Mbps ABR
2 hours @ 5.1 24/48 TrueHD = 3.06 Gbytes @ 3.4 Mbps ABR

...Universal could have easily used a lossless Dolby TrueHD 5.1 track with 16-bit/48 kHz resolution, the same as most Sony Blu-ray 5.1 PCM tracks, using the same, or less space as the higher resolution Dolby Digital Plus track.

Gee, you think maybe someone at Universal actually has a working brain? Naw. :rolleyes:

how show me the math

and none of that shows how they could

what does one have to do with the other.


Guy above me did

DolbyTrueHD 16/48 = 1.4 mbps ABR (average bitrate)

Track on King Kong is DD+ is 1.5 mbps

And Universal releasing a single HD DVD movie with TrueHD indicates that they aren't interested in the addition of TrueHD tracks of their films. Just get over it.

no wont work, try again

1) you got rid of the lossy that is mandatory if you don't want PO customers Bitching that there is no sound because they set up the player as DD straight to their receivers.

2) you are wrong in your math the DD+ is max as I pointed out in my previous post so the abr is not 1.5 for it.

DD+ is not VBR. TrueHD is, but an ABR of 1.4 is not a stretch to conclude would fit in the same (or less) space as 1.5 mb CBR.

KK also includes a Spanish and a French 5.1 track, neither of which are mandatory. Including a second English 5.1 track would have been a simple matter of choice.

Unless you have some more concrete data, how about you just shut the hell up.

...Universal could have easily used a lossless Dolby TrueHD 5.1 track with 16-bit/48 kHz resolution, the same as most Sony Blu-ray 5.1 PCM tracks, using the same, or less space as the higher resolution Dolby Digital Plus track.

Gee, you think maybe someone at Universal actually has a working brain? Naw. :rolleyes:The ABR of 16/48 is 1.4Mbps, but that includes peaks of almost 3.0Mbps.

During most of the track, there may be no audible difference between 16/48 TrueHD (or 16/48 LPCM) and 1.5Mbps DD+, but something is obviously lost on complex scenes which push close to 3.0Mbps. Since the average 16/48 TrueHD track requires less total space than 1.5Mbps DD+, capacity isn't the issue; the issue is bandwidth.

Remember, until we get dynamix muxing, everything is still encoded separately and muxed together at the end. If a title has a VC-1 feature that peaks at 22Mbps, an IME feature that peaks at 4Mbps, and three Dolby Digital tracks @ 640Kbps each, then there isn't going to be sufficient bandwidth for a 16/48 TrueHD track, which may peak at 3.0Mbps during 30 seconds of the entire film.

If authoring could be done more efficiently, perhaps 16/48 and 24/48 TrueHD tracks would be more common. The average VC-1 feature encode is probably <15Mbps ABR, the average VC-1 IME encode is just ~1.5Mbps ABR, and 24/48 TrueHD peaks at just ~5Mbps. Just looking at those numbers, HD-DVD's 30Mbps mux rate seems more than adequate. Most titles lack TrueHD simply because studios haven't made it a priority, imo.

Remember, until we get dynamix muxing, everything is still encoded separately and muxed together at the end. If a title has a VC-1 feature that peaks at 22Mbps, an IME feature that peaks at 4Mbps, and three Dolby Digital tracks @ 640Kbps each, then there isn't going to be sufficient bandwidth for a 16/48 TrueHD track, which may peak at 3.0Mbps during 30 seconds of the entire film.

No argument there, but we (all of we) can only speculate on whether TrueHD could have fit on the KK disc or not. Only the people that encoded that disc know that for certain. Knowing Universal's choice to go with higher resolution lossy DD+ instead of lower resolution lossless THD, I consider it a very fair speculation that TrueHD was not used as a choice, not a necessity.

What pisses me off is people like this "AnthonyP" spouting off their own speculation, without support of any kind, as though it is fact.

...Universal could have easily used a lossless Dolby TrueHD 5.1 track with 16-bit/48 kHz resolution, the same as most Sony Blu-ray 5.1 PCM tracks, using the same, or less space as the higher resolution Dolby Digital Plus track.

...with 16-bit audio instead of 24-bit.

It seems like you're assuming that lossy 24-bit is worse than lossless 16-bit, but I certainly have never seen any data to back that up.

Given all the complaints about the KK audio, I've yet to hear anyone complaining about any issues in the actual audio experience. Just ungrounded specsmanship.

It seems like you're assuming that lossy 24-bit is worse than lossless 16-bit, but I certainly have never seen any data to back that up.

No, not at all, Ben. You misunderstood me. Just the opposite, in fact. Given a choice between lossless THD at 16-bit or lossy DD+ at 20 or 24-bit and the sort of bit rate Uni uses (1.5 mb), I'll take the DD+ every time.

In my opinion, I believe Universal has come to this very conclusion.

no wont work, try again

1) you got rid of the lossy that is mandatory if you don't want PO customers Bitching that there is no sound because they set up the player as DD straight to their receivers.

2) you are wrong in your math the DD+ is max as I pointed out in my previous post so the abr is not 1.5 for it.

Robert George already presented the facts and you still ignore them. Present your data and stop the talking point bs. This is the AVS forum not some political forum.

DD+ is not VBR. TrueHD is, but an ABR of 1.4 is not a stretch to conclude would fit in the same (or less) space as 1.5 mb CBR.

DD+ is not CBR, DD is CBR on DVD and HD DVD because of the way DVD was built.

KK also includes a Spanish and a French 5.1 track, neither of which are mandatory. Including a second English 5.1 track would have been a simple matter of choice.

1) I know it is 1.5 for English but Universal has used lower bitrate on other languages, are you sure they did not do the same here? No one here ever talks about other languages

2) so now you just removed something important (i.e. keep 2 english and get rid of other languages) considering the discussion was about adding it and not removing/changing content that was there.

It seems like you're assuming that lossy 24-bit is worse than lossless 16-bit, but I certainly have never seen any data to back that up.

Ben, he was just trying to pretend it could fit, nothing more. He had to remopve something so remove the lossy, pretend there are no BDs with lossless 24/48 and then say "see it could fit" BS

Robert George already presented the facts and you still ignore them. Present your data and stop the talking point bs. This is the AVS forum not some political forum.

there are no facts in his post just a lot of BS marketing

...considering the discussion was about adding it and not removing/changing content that was there.

No, the discussion was about whether Universal COULD or COULD NOT put a TrueHD track on the HD DVD of King Kong. You have yet to present a single shred of evidence, other than your opinion, that they COULD NOT.

And I don't work in marketing.

No, the discussion was about whether Universal COULD or COULD NOT put a TrueHD track on the HD DVD of King Kong. You have yet to present a single shred of evidence, other than your opinion, that they COULD NOT.

And I don't work in marketing.

None will be forthcoming, either to support that contention, or this little gem, which didn't even include a *( not to scale ) disclaimer:

my graph
worst quality…………………………………best quality
DD..DD+..................................................... ..........DTHD

I'm guessing this was retrieved from a very dark, clammy, foul-smelling source.

there are no facts in his post just a lot of BS marketing

Actually the BS marketing is your job ;) You'd do better on a political forum where you can get away with what you do. It is clear that Universal could of put a TrueHD track in King Kong if they wanted to. However Universal would clearly rather release films with DD+ @ 1.5 mbps. Universals fault for the lack of TrueHD on King Kong, not HD DVDs just get over it.

No, the discussion was about whether Universal COULD or COULD NOT put a TrueHD track on the HD DVD of King Kong.

but you did not "put it on" obviously they can remove stuff and replace it for example they can remove the video and most likely add DTHD 24/192, or they can split the movie on two disks and then capacity would not be an issue and they could add DTHD for all three languages.....

you never even tried (obviously because you know we are right) that they could add DTHD to what is there. No one will buy the movie for DTHD or DD+ most fanboys are convinced DD+ is just as good as DTHD in the first place, and if someone wants the movie then they will grudgingly buy it even if there is no lossless (I did that with Happy Feet) on the other hand a French speaker that might think of buying because he watched it in the theatre in French and wants it in French won’t buy it if they lose the French track, that is what studios mean when they say diminishing returns, and say sorry, no lossless for you.

No one will buy the movie for DTHD or DD+ most fanboys are convinced DD+ is just as good as DTHD in the first place, and if someone wants the movie then they will grudgingly buy it even if there is no lossless (I did that with Happy Feet)...
Why grudigingly? Was there any deficiency with the audio on Happy Feet?

Why grudigingly? Was there any deficiency with the audio on Happy Feet?The BD25 only had Dolby Digital 5.1 EX in English, Spanish, French.

The HD30/9C combo had those tracks in DD+ 5.1 EX and also a English Dolby TrueHD 5.1 track.

No one will buy the movie for DTHD or DD+ most fanboys are convinced DD+ is just as good as DTHD in the first place, and if someone wants the movie then they will grudgingly buy it even if there is no lossless (I did that with Happy Feet) on the other hand a French speaker that might think of buying because he watched it in the theatre in French and wants it in French won’t buy it if they lose the French track, that is what studios mean when they say diminishing returns, and say sorry, no lossless for you.
What is this, a contest for the longest, meaningless sentence?

The BD25 only had Dolby Digital 5.1 EX in English, Spanish, French.

The HD30/9C combo had those tracks in DD+ 5.1 EX and also a English Dolby TrueHD 5.1 track.

Call me crazy, but I'd take the original language (you know, the one the creators of the film had in mind, when they....I don't know...CREATED IT....) in a lossless format before any others. And not just English...ANY title in any language. Dubs are irritating, and not the intended creation of those who made the movie.

FilmMixer already talked about technical reasons why you don't want to try lossless in other languages other than the original track. (I'll try to find it...not that it will ever stop a certain someone from claiming that there has to be at least 5 lossless languages on each title :rolleyes: )

Call me crazy, but I'd take the original language...No argument from me! The animated movie in question had mostly native English speaking actors doing the voices and songs.

No argument from me! The animated movie in question had mostly native English speaking actors doing the voices and songs.

Of course, but we all KNOW there should be Romanian and Swahili lossless tracks for it! :rolleyes:

I recently read about CD/DVD-A/SACD comparisons using a double blind listening test. The subjects could not tell the difference between any of the formats or resolutions (16-bit vs. 24-bit). The differences heard in Hi-Rez audio formats could be the result or different mastering that is better/louder. For instance, I wonder why the Sea Change DVD-A by Beck sounds so amazing. Is it the 24 bits, or good mastering with a strong LFE track?

So if bitrate doesn't improve sound quality, is it the lossless compression that makes the difference?

So if bitrate doesn't improve sound quality, is it the lossless compression that makes the difference?
Not per se. Obviously overly lossy encoding can hurt, but once it's lightly encoded enough that you can't hear a difference, there isn't a difference you can hear.

That threshold is lower per channel for movie soundtracks compared to 2-channel audio, since there's so much more going on to mask any noise, plus the distraction of having an image playing :).

I recently read about CD/DVD-A/SACD comparisons using a double blind listening test. The subjects could not tell the difference between any of the formats or resolutions (16-bit vs. 24-bit). The differences heard in Hi-Rez audio formats could be the result or different mastering that is better/louder. For instance, I wonder why the Sea Change DVD-A by Beck sounds so amazing. Is it the 24 bits, or good mastering with a strong LFE track?

Where did you read about this article ? Is there a link ? Thx.

Just thought I'd chime in here. As a recording professional, I can tell you that the industry standard is 24 bit. It allows engineers to achieve much more headroom in recordings. You can perceive more detail and things sound more natural with 24 bits. In my opinion, and many an engineer as well, more bits are better than a higher sampling rate. 24/48 is widely used on commercial recordings. 24/96 and higher are great, but audibly the difference between 48/96k rates are minimal. I typically record @ 24/96 on location and 32/48 in my studio.

I'm not going to comment on the sound quality of the codecs being discussed here, but just thought I'd pass on some factual theory from the pro-audio world.

Tweakheadz has a great article on 16 vs 24 bit audio. Sorry, I can't post the link due to my post count. There are also many examples of this on the web.

Chris

Tweakheadz has a great article on 16 vs 24 bit audio. Sorry, I can't post the link due to my post count. There are also many examples of this on the web.http://www.tweakheadz.com/16_vs_24_bit_audio.htm

There are a lot of forum threads over on tweakheadz where professional sound engineers and mixers discuss the improvements with 24-bit as well.

Just thought I'd chime in here. As a recording professional, I can tell you that the industry standard is 24 bit. It allows engineers to achieve much more headroom in recordings. You can perceive more detail and things sound more natural with 24 bits. In my opinion, and many an engineer as well, more bits are better than a higher sampling rate. 24/48 is widely used on commercial recordings. 24/96 and higher are great, but audibly the difference between 48/96k rates are minimal. I typically record @ 24/96 on location and 32/48 in my studio. Several other studio mixers have said the same thing on this forum.

There is no question that many people can reliably distinguish between 16/48 and 24/48 in a carefully controlled environment (studio quality speakers, treated room, etc). Plenty of double-blind tests have shown that in the past, some of which are linked on wiki. As you can see on this forum, a lot of members believe that the average person with the average setup in the average room isn't going to tell a difference, because of their equipment, room acoustics, or SNR of the listening environment. That may well be true, but I don't think Blu-ray and HD-DVD should cater to the lowest common denominator.

The electronics for audio and video reproduction improve with every passing year. With the full resolution of the studio master on every disk, we can be certain we are getting the best sound possible with a particular title. Further, that way, the experience continues to improve as we upgrade with future generations of A/V equipment.

The appended chart shows how DACs used in CE equipment have steadily improved over the past several years.

I've been lurking for years and do indeed have an above average set up. My studio is in my home and it's far from ideal as far as treatments, isolation and such. I do own nice gear but the point I was attempting to make is that there are real-world benefits to higher bit rates that I believe will come across typical, consumer equipment. We are talking about timbre, smoothness, warmth, etc. Certainly not something that the casual enthusiast will hear, but it's there if you listen for it.

There are too many factors as well as opinions to fully rationalize this debate, but I just wanted to get in my two cents, for what it's worth. I've always loved audio and sometimes feel that people expect a little too much from the world of home entertainment. There are many great sounding titles available and lots of bad ones as well. I'd be happy enough just to have consistantly good quality audio across the board no matter what the format.

Chris

IMHO the difference between 48 kHz and the higher ones (96 & 192 kHz) is the ease of doing proper low pass filtering without adding phase shift impact. Much easier to do a perfect brick wall filter to just pass 24 kHz analog and below with almost zero phase shift.

But you can throw more $ at a low pass filter and still get the same minimal phase shift impact when only 48 kHz is used for the encode. Plus you can also oversample to 192 kHz just before the A/D conversion and use 192 kHz conversion rate DACS.

My own belief is that the increase in the sampling rate is in fact more important than the increase in bits. 16 bits is enough to give a high signal-to-noise, provided the data is properly dithered on encode.

But, the 44.1 kHz sampling rate of CD is simply death to high quality music reproduction. Transient accuracy is destroyed. A frequency analysis will only show some phase shifting in the below-20kHz region, but in the time domain transients look bad. There is ultrasonic ripple before the transient even happens. Nothing in nature works like that, and it isn't suprising that our brains notice something is wrong in the highs. While we can't hear ultrasonic sine-waves, there is research that indicates our brain can notice the difference in sound when ultrasonics associated with harmonics of audible sound, or with transients are added or removed. How this works, I'm not sure -- it could be some sort of nonlinearity in the playback devices, or in our ears themselves, that makes accurate waveform reconstruction important, even beyond 20 kHz.

The truth is that 24/96 simply sounds better than CD, even on the same playback. I think studies have shown that we can resolve up to about 18 or 19 (accurate) bits, but likely not beyond that. But, it's the high frequencies that are most improved when going to high sampling rates. Instead of ripping your ears off, the sound becomes anolog-like. This isn't a result of reconstruction filters, because the same CD signal can be oversampled, and use the same filters. CD simply won't sound as good.

Whether differences can be heard on movies, which have dialog normalization, and are usually processed by THX during playback (bringing down the highs a bit), I'm not sure. I think movie soundtrack quality is more likely to be a function of the audio engineering than small changes like this.

But on a music track, or music video, I would want at least 48 kHz, and preferably 96.

A frequency analysis will only show some phase shifting in the below-20kHz region, but in the time domain transients look bad. There is ultrasonic ripple before the transient even happens.Very interesting! But phase or time domain transients -- somewhat the same to me -- Yes I would agree that you want the "reconstruction filter" to be completely devoid of doing anything to the audio passband (<24 kHz).

As to having to pass harmonics of the highest audio freqs -- that is beyond my ordinary knowledge, but I do remember some discussion about that either on other AVS forums or in audio magazines. Maybe it was stuff I read ten (20?) years ago on the usenet? Never did hear that any formal conclusions or findings were reached concerning the need for high frequency harmonic content (>24 kHz).

There are a lot of forum threads over on tweakheadz where professional sound engineers and mixers discuss the improvements with 24-bit as well.

Several other studio mixers have said the same thing on this forum.

There's no argument about recording and mastering at >16-bit as being an extremely good thing. I'm not an audio pro, but I've done enough audio-for-video in the 16-bit and now the 32-bit worlds to know the difference when you have to multiple generations of effects that all need to get dithered to 16-bit.

But if you just have to do one last dither, in the final normalized mix, with the time to get it right, I think 16-bit can produce great sound.

That said, when I'm encoding to our WMA Pro codec from 24-bit source, I'll always encode as 24-bit, just because that's easier than worrying about that optimal 16-bit dither. That's another plus for DD+ - it isn't a huge birate hit to move to >16-bit, unlike PCM (especially in BD).

Just thought I'd chime in here. As a recording professional, I can tell you that the industry standard is 24 bit. It allows engineers to achieve much more headroom in recordings. You can perceive more detail and things sound more natural with 24 bits.

No doubt more is always better but at what point is it just wanted.
Is comparing 16bit to 24bit the right comparison to make?
What about comparing 16bit to 18 or 20bit?

Increasing the bit depth to 20bit from 16bit results in a 16X increase in dynamic range. Since no consumer grade audio DAC have SNR > 120db they simply can't resolve anything greater than 20bit. So when you are comparing a 16bit sample to a 24bit sample in reality you are comparing a 16bit sample to a 20bit sample. The least 4 significant bits take up space but are simply lost in the noise floor.

If you get the chance try comparing some 20bit samples to 24bit sample in a double blind test.

I guess another way to look at 16bit vs 24 bit would be like asking an employee what is the minimum salary they would accept:
A) $10,000 annually (16bit)
B) $2.5 million annually. (24bit)

Given only 2 choices most people would pick B. Does that mean none of the employees would accept the position for $160K (20bit)?

I wonder if in 10yrs we will have the same arguments about the so called merits of 32bit audio with it's 192db dynamic range which approaches the limits of max sound pressure possible on planet earth (196db SPL I believe)?

We are talking about timbre, smoothness, warmth, etc.

Definitely what the room and speaker affect the most (for a given recording).

---------

Here is the essential problem with subjectivists. They want to say that somehow when it comes to hi-end audio laws of physics do not hold good. Esp. laws of physics and the engineering that created the equipment in the first place.

I've no idea why someone would fall for this kind of snake oil :(

Why grudigingly? Was there any deficiency with the audio on Happy Feet?

Yes. I watched it last night and the audio felt very constrained compared to PCM tracks. I'd have happily ppaid an extra $5 to get a pcm or TrueHD track on it. Anyone who thinks DD+ sounds as good is either convincing themselves, or doesn't have good enough audio gear to tell the difference.

I can see why a distortion free frequency response to above hearing limits is good to keep distortion down below it, but I have doubts of any supersonic pschoacoustic benefit.

I like this link for ultrasonic ring tones http://www.ultrasonic-ringtones.com/ "from around 17kHz upwards we get into the range that only the under 20's can hear! "

I can't hear the 12 kHz tone any more FWIW, and probably the reason for my doubt.

If you get the chance try comparing some 20bit samples to 24bit sample in a double blind test.The same tests that found many could distinguish between 16/48 and 24/48 also found that individuals could not distinguish between 20/48 and 24/48.

No doubt more is always better. Is comparing 16bit to 24bit the right comparison to make. What about comparing 16bit to 18bit or 20bit.

Increasing the bit depth to 20bit from 16bit results in a 16X increase in dynamic range. Since no consumer grade audio DAC have SNR > 120db they simply can't resolve anything greater than 20bit. So when you are comparing a 16bit sample to a 24bit sample in reality you are comparing a 16bit sample to a 20bit sample. The least 4 significant bits take up space but are simply lost in the noise floor.The upcoming 2007 mid-range receivers from Denon and Onkyo use the new Burr-Brown PCM1796 (http://focus.ti.com/lit/ds/symlink/pcm1796.pdf) with a 123dB SNR. In contrast, the high-end models from 2006 used the Burr-Brown PCM1791 (http://focus.ti.com/lit/ds/symlink/pcm1791a.pdf) with a 113dB SNR. Now, of course this is just one component in the chain, but it's an example of how the parts in consumer A/V products continue to improve to deliver greater performance at lower price points.

No doubt more is always better. Is comparing 16bit to 24bit the right comparison to make. What about comparing 16bit to 18bit or 20bit.

Increasing the bit depth to 20bit from 16bit results in a 16X increase in dynamic range. Since no consumer grade audio DAC have SNR > 120db they simply can't resolve anything greater than 20bit. So when you are comparing a 16bit sample to a 24bit sample in reality you are comparing a 16bit sample to a 20bit sample. The least 4 significant bits take up space but are simply lost in the noise floor.

If you get the chance try comparing some 20bit samples to 24bit sample in a double blind test.

I guess another way to look at 16bit vs 24 bit would be like asking an employee what is the minimum salary they would accept:
A) $10,000 annually (16bit)
B) $2.5 million annually. (24bit)

Given only 2 choices most people would pick B. Does that mean none of the employees would accept the position for $160K (20bit)?

What I've gathered from the professionals that have posted here is that the difference between 16 and 20 bit is much more audible to most than the difference between 20 and 24 bit. On my lame old receiver and speakers, and noisey room, I bet they'd all sound the same.

I wish there were a recent test where unknowing listners were allowed to listen to all 3 (16,20,24) with the same material, and if they could even identify them, in the best of conditions.

EDIT: sorry, just saw the posts about the tests above

No, volume adjustments within a range of typical listening levels won't improve localization or imaging, but it's well understood that when presented with two single-sourced samples differing only in volume, listeners consistently express a preference for the louder example. The effect is quite pronounced, and has been observed even when the relative SPL differs by only .1 dB. Unfortunately, that level of precision isn't practically achievable with inexpensive, consumer-grade instruments.

Absolutely true, but not really relevant in this case, as 1) the PCM tracks are not louder, and 2) there are several very different listening levels used in my home theater (depending on audience composition...e.g. fiance = quieter), and the quieter PCM still invariably demonstrates better localization and imaging than louder compressed tracks. Still, it's worth a test when I get some free time.

Anyone who thinks DD+ sounds as good is either convincing themselves, or doesn't have good enough audio gear to tell the difference. I've heard DD+ soundtracks that were better than TrueHD and I do have a room and system that easily reveals the differences.
Do you?
I see no info in your profile or a link to pics of your setup.

Here is the essential problem with subjectivists. They want to say that somehow when it comes to hi-end audio laws of physics do not hold good. Esp. laws of physics and the engineering that created the equipment in the first place.

I'm sorry, how exactly are you qualified?? I don't believe that I forced my personal experience on anyone during my post. Calling a complete stranger whom you have never met or even conversed with a subjectivist is a pretty big smack in the face. Thanks for the thread crap.

I've heard DD+ soundtracks that were better than TrueHD and I do have a room and system that easily reveals the differences.
Do you? I see no info in your profile or a link to pics of your setup.

I personally tend to believe that anything > DD+ is into very rapidly diminishing returns, but I also think the "+" is a bit misleading. DD+ is more significantly improved over plain vanilla DD than that symbol might lead you to expect, IMHO.

I'm sorry, how exactly are you qualified??

Qualified for what ?

Anyway, not really picking a fight with you - just talking about subjectivists - not esp. about you. I'll edit the post to clarify that. BTW, what you wrote sounds - exactly - like what I've seen from subjectivists for about a decade now ... here, in audioasylum, and the newsgroups before that.

I'm sorry, how exactly are you qualified?? I don't believe that I forced my personal experience on anyone during my post. Calling a complete stranger whom you have never met or even conversed with a subjectivist is a pretty big smack in the face. Thanks for the thread crap.

What makes you so sure that he was referring specifically to you, and why are you so defensive? Have you ever read "Stereophile?" The BS piles up pretty fast and deep in their columns and reviews.

Back on track. If you can hear the difference between an mp3 and a 16bit/44.1 cd, than you should be able to hear the difference between 16 and 24 bits. There's an audio theory that says even though we might not be able to hear certain higher frequencies, they do exist and may have an effect on the lower frequncies that we do hear. It's all perception.

By the same token, there are devices that capture audio with super high sampling rates such as 5.6 MHz and the bit rate is simply 1. So I guess it's ultimately a trade-off.

All of this is completely subjective. One person may hear what another can't. The absolute bottom line is how it sounds to me (or you.)

Chris

All of this is completely subjective. One person may hear what another can't. The absolute bottom line is how it sounds to me (or you.)
Chris
Not if you want to put in place standards that you can't prove. Sorry

There's an audio theory that says even though we might not be able to hear certain higher frequencies, they do exist and may have an effect on the lower frequncies that we do hear. It's all perception.

If these inaudible harmonics have any perceptible effect, then a double blind procedure shouldn't impact its' illustration either way. Double blind procedures utilizing well encoded MP3s at a sufficient bit-rate have suggested that they are audibly indistinguishable from the 16/44 PCM program material from which they were sourced.

Nataraj,

Sorry about my defensiveness. I too for many years have cringed at the "audio mysticism" and it's claims. I just strive to get the facts out and it seems as though many people buy into this stuff. Believe me, I am not some sort of audio snob and also apologize if I came across as one. I am a lover of audio and obviously pretty passionate about it. My goal is to increase awareness about how good it can and should be.

Chris

Chris

If you can hear the difference between an mp3 and a 16bit/44.1 cd, than you should be able to hear the difference between 16 and 24 bits.

That makes no sense at all. If you can make out the difference between 1 meter and one centimeter doesn't mean you can make out the difference between 1 micrometer and 1 nanometer.

There's an audio theory that says even though we might not be able to hear certain higher frequencies, they do exist and may have an effect on the lower frequncies that we do hear. It's all perception.

And has that been proven ?

All of this is completely subjective. One person may hear what another can't. The absolute bottom line is how it sounds to me (or you.)

Sure. But can he really hear the difference or is he just imagining he hears a difference ? And how do we tell ?

Afterall would you buy medicine that has proven to be no different in clinical trials than the placebo ?

I personally tend to believe that anything > DD+ is into very rapidly diminishing returns,DD+ @ 1.5mbps. Agreed.
but I also think the "+" is a bit misleading.At times, yes.
DD+ is more significantly improved over plain vanilla DD than that symbol might lead you to expect, IMHO. Agreed.

Also how the soundtrack was recorded makes a huge difference and with DD+ and TrueHD it is much easier to distinguish these differences, especially being able to tell the difference when dialog is recorded on a "real" location rather than post-processing where it sounds like a voice in box.
It's still clear but there's no "air" , depth or ambiance to it.

Qualified for what ?

Anyway, not really picking a fight with you - just talking about subjectivists - not esp. about you. I'll edit the post to clarify that. BTW, what you wrote sounds - exactly - like what I've seen from subjectivists for about a decade now ... here, in audioasylum, and the newsgroups before that.

You talking about all those audio recording engineers and audio listening enthusiasts who'd swear 24/96 PCM audio is a huge improvement over 16/44.1k CD ? If so, there are a heck of lot of people who are delusional; including myself. In my experience, 24/96 playback thoroughly trashes 16/44.1 with any well-recorded music in my system.

Believe what you want. A lot of us are convinced high resolution audio is desirable, even with movie soundtracks. If you ever get a chance, try to audition the Eagle's "Hotel California" DVD-V w/DTS soundtrack vs. the DVD-A 24/96 soundtrack on a decent system.

And has that been proven ?

As I stated, it's a theory not a fact.


That makes no sense at all. If you can make out the difference between 1 meter and one centimeter doesn't mean you can make out the difference between 1 micrometer and 1 nanometer.

It makes perfect sense. It's just that the difference is not that great. But there is a difference.

And why has this all of a sudden turned into a thread about "If a tree falls in the forest...." routine? Do you guys actually listen and enjoy or do you just sit around and pick apart every possible angle until there is nothing left to listen to?

You talking about all those audio recording engineers and audio listening enthusiasts who'd swear 24/96 PCM audio is a huge improvement over 16/44.1k CD ? If so, there are a heck of lot of people who are delusional; including myself. In my experience, 24/96 playback thoroughly trashes 16/44.1 with any well-recorded music in my system.

Believe what you want. A lot of us are convinced high resolution audio is desirable, even with movie soundtracks. If you ever get a chance, try to audition the Eagle's "Hotel California" DVD-V w/DTS soundtrack vs. the DVD-A 24/96 soundtrack on a decent system.

Finally....

You talking about all those audio recording engineers and audio listening enthusiasts who'd swear 24/96 PCM audio is a huge improvement over 16/44.1k CD ? If so, there are a heck of lot of people who are delusional; including myself. In my experience, 24/96 playback thoroughly trashes 16/44.1 with any well-recorded music in my system.

Believe what you want. A lot of us are convinced high resolution audio is desirable, even with movie soundtracks. If you ever get a chance, try to audition the Eagle's "Hotel California" DVD-V w/DTS soundtrack vs. the DVD-A 24/96 soundtrack on a decent system.

Agreed! Even if a small percentage of people can take advantage of the increased audio resolution, why exclude it or claim there is no advantage?

Now, a poorly recorded/mixed track will always sound bad, no matter how high the resolution is :)

Hey! I do not think that anyone here is defending 16 bit depth versus 20 bits for a good system. 24 bits may be a standard storage & transport format but at playback time it is very hard to use more dynamic range than what 20 bits provides. I have not seen any strong defenders or proponents that "we all" only need 16 bits forever.

Yes. I watched it last night and the audio felt very constrained compared to PCM tracks. I'd have happily ppaid an extra $5 to get a pcm or TrueHD track on it. Anyone who thinks DD+ sounds as good is either convincing themselves, or doesn't have good enough audio gear to tell the difference.
Have you heard the PCM version for comparison? I'm not sure how one coudl discriminate between the mix and the compression just hearing one version.

You talking about all those audio recording engineers and audio listening enthusiasts who'd swear 24/96 PCM audio is a huge improvement over 16/44.1k CD ? If so, there are a heck of lot of people who are delusional; including myself. In my experience, 24/96 playback thoroughly trashes 16/44.1 with any well-recorded music in my system.

Believe what you want. A lot of us are convinced high resolution audio is desirable, even with movie soundtracks. If you ever get a chance, try to audition the Eagle's "Hotel California" DVD-V w/DTS soundtrack vs. the DVD-A 24/96 soundtrack on a decent system.

Once again a comparison between 16 bit and 24 bit is misleading to begin with. That's like saying choose between a gremlin and a ferrari then saying see everyone needs a ferrari. Since the DAC in your receiver cant resolve more than about 100-115db SNR anything beyond 20bit (120db) is simply hidden by the noise in your receiver. The truth is if the same comparison was done between 16 bit and 18 bit most people couldn't tell the difference but maybe a small percentage could under ideal conditions. If you then had those same people in a double blind compare 18bit, 20bit, 22bit and 24bit there simply would be no perceptible increase in fidelity.

Does higher bit depth and sampling provide an improvement over CD quality? Yes.
The fact is the gain is not linear. Just because 16/44.1 is not enough doesn't mean we need to take a giant leap to 24/192 to get better sound.

Have you heard the PCM version for comparison? I'm not sure how one coudl discriminate between the mix and the compression just hearing one version.

Then give me the 24/48 PCM version to try out. "Lossy compression" means just that: "lossy"; even 16/48 PCM is "lossy" if the master tape was at 24/48. Maybe most people with most sound systems can't distinguish/care about the difference, but there will be people who have sufficiently good systems to appreciate the differences. Even 1.5M DTS is a dramatic improvement over your typical DVD-V soundtrack. I'm not convinced with DD+ soundtracks); I've yet to hear a DD+ soundtrack which wow'ed me after I've heard uncompressed PCM. E.g. King Kong's audio track isn't exactly SOTA.

If I could get music concerts with full motion HD video and 5.1 24/96 soundtracks, I'd be all over them.

Have you heard the PCM version for comparison? I'm not sure how one coudl discriminate between the mix and the compression just hearing one version.

A well designed double blind procedure can rapidly segregate real qualitative wheat from snake-oil chaff in audio. Many participants in such tests reguarly express shock and dismay at their inability to consistently identify differences which they earnestly believed had been clearly audible between various sources, sampling rates, etc. in anecdotal circumstances.

A well designed double blind procedure can rapidly segregate real qualitative wheat from snake-oil chaff in audio. Many participants in such tests reguarly express shock and dismay at their inability to consistently identify differences which they earnestly believed had been clearly audible between various sources, sampling rates, etc. in anecdotal circumstances.

It has been proven that when you teach a person what to listen for, their ability to tell the difference improves. A lot of folks go into dbt without the knowledge of how to listen for differences. Small details like reverberation tails, transient attacks, and ambience are not what people are listen for, they always listening to things that are easy to code. This is why dbt are so difficult for most folks.

Small details like reverberation tails, transient attacks, and ambience.Those are the only things -- mainly during quiet passages -- that I notice improved when comparing 640 kbps DD+ vs 16/48 TrueHD on POTO. :)

Have you heard the PCM version for comparison? I'm not sure how one coudl discriminate between the mix and the compression just hearing one version.

That's a red herring. I would easily subject other tracks a blind test (and have done so under informal "trials" already), so you're pleading to ignorance with this (hold the hounds guys, I'm not calling Ben ignorant, that's an actual logical falacy: the appeal to what we don't know). Give us the PCM track so we can tell.

Or are you afraid that would reveal a shortcoming your chosen format?

Edit: Before the hounds are released anyways, I thought I'd link to the Argument from Ignorance on wikipedia so you don't have to take my word for it: http://en.wikipedia.org/wiki/Argument_from_ignorance

I personally tend to believe that anything > DD+ is into very rapidly diminishing returns, but I also think the "+" is a bit misleading. DD+ is more significantly improved over plain vanilla DD than that symbol might lead you to expect, IMHO.

lol did you read the specs on DD+? do you even know anything about it? start with http://www.dolby.com/assets/pdf/tech_library/DPlus_TrueHD_whitepaper.pdf and go on

lol did you read the specs on DD+? do you even know anything about it? start with http://www.dolby.com/assets/pdf/tech_library/DPlus_TrueHD_whitepaper.pdf and go on


Anthony, exactly where did you find the little graphic you provided showing the relative qualitative positions of DD, DD+, and DTHD? Here, this might help you remember:

worst quality.......................................best quality
DD..DD+..................................................... ......... ...................DTHD

I'm guessing that the "white paper" you sourced this from, is sold on rolls in packs of 8 or 12.

Anthony, exactly where did you find the little graphic
In his mind of course. rto, if I may say so, don't bother, it's not worth it.

That's a red herring. I would easily subject other tracks a blind test (and have done so under informal "trials" already), so you're pleading to ignorance with this (hold the hounds guys, I'm not calling Ben ignorant, that's an actual logical falacy: the appeal to what we don't know). Give us the PCM track so we can tell.

Or are you afraid that would reveal a shortcoming your chosen format?

Edit: Before the hounds are released anyways, I thought I'd link to the Argument from Ignorance on wikipedia so you don't have to take my word for it: http://en.wikipedia.org/wiki/Argument_from_ignorance
And would that be an example of the Wikipedia Fallacy, where a post to an authoratative yet unrelated reference is provided instead of an answer? Ironically, I can't find a link for that one :).

So, what do you hear in the Happy Feet audio that you can tell is due to compression, and how can you tell that it's due to compression and not in the original mix?

Yeah they could, just because they didn't doesn't mean they couldn't. Like I said before, Universal doesn't include lossless tracks in their movies. Only one of Universals 97 HD DVD movies has Dolby TrueHD.I hope you will stop repeating that, since it isn't true. Two Universal titles that came out before "King Kong" and had TrueHD were already discussed before your post. And now Universal has put lossless on "Alpha Dog", which comes out May 1st, according to the highdefdigest review. Kind of goes against the argument that Universal is against TrueHD when the do it for 2 titles, announce it for 2 more and cancel it, then do it for another title.
...Universal could have easily used a lossless Dolby TrueHD 5.1 track with 16-bit/48 kHz resolution, the same as most Sony Blu-ray 5.1 PCM tracks, using the same, or less space as the higher resolution Dolby Digital Plus track.

Gee, you think maybe someone at Universal actually has a working brain? Naw. :rolleyes:That was extremely misleading. You should know enough about bandwidth at this point to know how misleading it is to only look at the space. Universal could have put "King Kong" on an HD15. Not without reencoding, but they could have done it. As to whether they could have put a TrueHD track without encoding the video differently, I doubt they could have. Maybe it would have looked the same, but your post above makes it look like they could have since you did the math with the space, when the actually is that it is likely that the current encode wouldn't have worked with TrueHD because of the bandwidth limitations.
Robert George already presented the facts and you still ignore them.While Robert's facts were likely true, it was an incomplete analysis to see if TrueHD would have worked with the current encode and probably misled some people who don't understand the bandwidth issue. Kind of like only looking at whether something will fit into an elevator to answer whether the elevator can be used to move that item from floor to floor, without considering the weight.

--Darin

Anthony, exactly where did you find the little graphic you provided showing the relative qualitative positions of DD, DD+, and DTHD? Here, this might help you remember:

worst quality.......................................best quality
DD..DD+..................................................... ......... ...................DTHD

I'm guessing that the "white paper" you sourced this from, is sold on rolls in packs of 8 or 12.

Are you saying a document from the DOLBY website with information about DOLBY products is considered equal to TOILET PAPER because you cannot refute what he is saying? Am i reading this correctly? Are you guys even movie fans?! I dont think any of you are to be honest!! MY GOODNESS!!

And would that be an example of the Wikipedia Fallacy, where a post to an authoratative yet unrelated reference is provided instead of an answer? Ironically, I can't find a link for that one :).

So, what do you hear in the Happy Feet audio that you can tell is due to compression, and how can you tell that it's due to compression and not in the original mix?

First off, Wikipedia didn't exactly invent the arguement from ignorance, it's a fairly standard inclusion in philosophy courses on rhetoric, logic, etc., and it's been mentioned by everyone from Carl Sagan (Demon Haunted World) to being gently poked fun at by Terry Pratchett.

Obviously I don't have access to the original mix of Happy feet. What I can tell you is that every TrueHD and PCM mix I listen to has faaaar better imaging (and as a result improved localization of sound effects). To be sure, how great the difference is varys depending on the quality of the mix (I asume that lossless audio is in fact that, so I've never attributed any difference to the codec), but there is always a marked improvement. Happy Feet lacks the imaging heard in even fairly mediocre mixes with PCM/lossless.

I assume this is due to compression because 1) the obvious statistical correlation between PCM/lossless disks and imaging, and the lack of both on Happy feet. 2) presumably the artists who worked on the movie (on which the animators and TDs made some cutting edge breakthroughs in using Softimage XSI and mental ray for feature-length movies) were not appreciably less-competent at the audio level (it's possible that they were, but in general animated features tend to have better, not worse audio mixes). 3) There is sound physics behind why compression effects sound quality (pun intended). While lossy compression techniques generally try to remove data you "can't" see or hear, in point of fact the p-waves that create sound are subject to harmonic reinforment and cancellation, and removing sound that you "can't hear" still affects the frequencies that you can.

I grant you that most people probably niether care nor have the equipment to hear much of a difference if they did. But then I didn't think I was buying into the lowest commmon denominater when I plunked down lots of hard-earned cash for these formats.

In my experience, 24/96 playback thoroughly trashes 16/44.1 with any well-recorded music in my system.

This forum is AV Science forum. Without BDT, "in my experience" doesn't mean anything.

Believe what you want.

I don't care about beleafs. We are talking science - not religion.

As I stated, it's a theory not a fact.

If it was a worthwhile theory - it would have been proven in DBT by now.

Do you guys actually listen and enjoy or do you just sit around and pick apart every possible angle until there is nothing left to listen to?

Now this is an actual personal attack against a stranger :p

And this proves you have run out of your "arguments". You can beleive in snake oil all you want, but you will be challenged if you try to peddle it here.

If it was a worthwhile theory - it would have been proven in DBT by now.

Um, harmonic reinforcement and cancellation is extremely basic physics, generally taught to science majors in their second semester of physics. Not exactly rocket science (although some of that is covered too :D )

Because of this, clipping of "inaudible" notes and other forms of compression (including limitations in the original recording, and the range of sounds your speakers and electronics can reproduce) are largely what are responsible for making recorded music sound different from live sound.

I grant that there is a lot of audiophoolery that is passed off as accepted dogma in audiphile publications, but don't confuse that with the accepted science on how sound waves intereact with each other, or the roll of compression on the cues our brains use to determine where the sound is "coming" from.

Does 24bit pcm sound better than 16bit? Without blind tests I honestly cannot say, but I would rather the deficiency be in my equipment, rather than the audio itself. And there is a clear difference between PCM/lossless and compressed tracks, hence Warner (Happy Feet) or other studios' releases that lack them (especially when there is extra space available on one format or the other) is audibly short-changing the very people who most want to support these new formats.

Are you saying a document from the DOLBY website with information about DOLBY products is considered equal to TOILET PAPER because you cannot refute what he is saying? Am i reading this correctly? Are you guys even movie fans?! I dont think any of you are to be honest!! MY GOODNESS!!

That graphic is nowhere to be found in the Dolby white paper, nor does accurately it reflect any aspect of the technical discussion therein.........so, where did it come from? I have a theory. :rolleyes:

Um, harmonic reinforcement and cancellation is extremely basic physics, generally taught to science majors in their second semester of physics. Not exactly rocket science (although some of that is covered too :D )

And the context is ... ?

Does 24bit pcm sound better than 16bit? Without blind tests I honestly cannot say, but I would rather the deficiency be in my equipment, rather than the audio itself.

Thats just blind faith. You don't do engr based on blind faith or spend a lot of money / time over engr for no reason.

It is no different from saying the video should have colors that have not been demostrated to be visible to human eye (similar to when people talk about 96 or 192 kHz).

As I said before somehow people assume 24/96 is the silver bullet to good audio. That is simply not the case. You need good equipment (esp. room and speakers). We should put in the most effort in areas that matter most - speakers & room. And they (studios) should put in most effort in mastering.

I do see some usefulness in better than 16 bit audio. But that is mostly to get significant 16 bits after all the dsp - time, speaker and room corrections that need to be applied.

All these "audio fundamentalists" will have a heart attack when they plot thier speaker/room's frequency response.

Now this is an actual personal attack against a stranger

And this proves you have run out of your "arguments". You can beleive in snake oil all you want, but you will be challenged if you try to peddle it here.

Enjoy living in your fantasy land......

I suddenly remembered why I don't often post. People like you who have to have a bonified, scientific and clinical rationalization for everything instead of just enjoying something for what it is. Last time I checked, you don't need multiple degrees in physics, engineering and psycho acoustics to enjoy home theater.

Thats just blind faith. You don't do engr based on blind faith or spend a lot of money / time over engr for no reason.

If you had to do any significant engineering on this I'd agree. I'm just asking that the highest available sample rate be put on the disk; if anything it takes more work by the technician to resample a track prior to encoding (or in the case of PCM, just authorin it as an additional track).

It is no different from saying the video should have colors that have not been demostrated to be visible to human eye (similar to when people talk about 96 or 192 kHz).

Absolutely wrong. Electromagnetic radiation, which has a dual wave/particle nature, does not engage in harmonic reinforcement on the scale of home theater observation, while sound waves (which is just moving air and muuuuuch larger) do.

As I said before somehow people assume 24/96 is the silver bullet to good audio. That is simply not the case. You need good equipment (esp. room and speakers). We should put in the most effort in areas that matter most - speakers & room. And they (studios) should put in most effort in mastering.

I absolutely agree. I mainly got involved in this thread because I'd like to voice my preference for the best available track to be used (which is simply not 24/96 with lots of material), and for there to always be a PCM and/or lossless track available. Movies without it, like Happy Feet and, frankly, King Kong, are simply less engaging on high-end home theaters than they would be, and that's dissapointing as a consumer.

I agree with the remainder of your post (especially as regards frequency response).

Um, harmonic reinforcement and cancellation is extremely basic physics, generally taught to science majors in their second semester of physics. Not exactly rocket science (although some of that is covered too :D )

Because of this, clipping of "inaudible" notes and other forms of compression (including limitations in the original recording...... are largely what are responsible for making recorded music sound different from live sound.

Um and how do you hear high frequency air pressure changes that the ear doesn't respond to? Like trying to detect AM high frequency radio sidebands by listening to the carrier frequency. Nice invention if you can do it.

The discussion about high sampling rates has been going on a long time.
From a April 2000 article:
http://www.moultonlabs.com/weblog/more/how_high_do_we_really_need_to_hear

The issue of bandwidth is a fascinating one. On the one hand, the high-frequency limits of our hearing seem pretty well established at somewhere between 14 and 20 kHz., depending on our age, gender, exposure to race cars, etc. On the other hand, we KNOW that there is a lot of significant acoustical information extending well above 20 kHz. and speculations abound about the possible effect that such information may have upon our perceptions. Food for thought . . .

Meanwhile, the rules of digital audio bandwidth are pretty unequivocal. The Nyquist frequency (you know, 1/2 the sampling frequency – the upper limit for bandwidth in a digital system) has proved inviolable, and as we’ve gotten more experienced at digital stuff, we’ve come to notice that we’ve got to be really careful to get everything filtered out at some point BELOW the Nyquist frequency to head off sonic nasties. This means, of course, that if there IS something useful for us above 20 kHz., well, the digital audio system is for sure taking it out. Being the paranoid lot we are, we suspect the worst, to wit: that we are losing some ineffable ultrasonic signal illumination that reveals artistic truth and purity. Moan, groan, whimper, whine! Hence the big push for the 96 kHz. sampling rate. We ain’t sure we ain’t losing something, and we sure wanna be sure we ain’t!
Tales From the Audio Crypt
This is all fine and dandy. However, let me share with you two experiences I have had.

Tale #1. About two years ago, while attending an AES Convention, I was in a meeting of the Psychoacoustics Subcommittee of the Technical Council and we were discussing a proposal for a new wide bandwidth digital format (100 kHz.!) when one of the committee members asked a really simple question. “Hey,” he said, “does anyone here know of any studies that convincingly show that humans hear above 20 kHz.?” And we all just sat there. No one raised their hand. Not one of us knew of anything like that!

So much for what the people who are really interested in what and how we hear can tell us. Above 20 kHz? No evidence! Zero, zippo, zilch.

Tale #2. Last Fall, I attended the Sony/Phillips SACD demo at the AES convention in NYC. Now SACD is a 1-bit format with a sampling frequency of slightly less than 3 megahertz, which means the bandwidth is going to be really pretty wide. Before one of the recordings played, the presenter (an engineer from Phillips) murmured “And this is a copy of a special analog recording that was done on a specially tweaked analog deck flat to 50 kHz. . . .” That murmur got me to thinking. After the presentation (which was really pretty nice, by the way), I asked the presenter if anyone had bothered to measure the acoustic bandwidth throughput of the SACD system, which is this geek’s way of asking, “How high a frequency can really pass through this system, from microphone input to loudspeaker output?”

No, they hadn’t measured such a thing, he said, and he had no idea what the figure might be. However, he went on, the recording had been done with B&K 4006s, which were good to about 30 kHz., he figured. He went on to note that to get the bandwidth much higher, you’d have to seriously reduce diaphragm size, which in turn would increase the already troublesome equivalent input noise of the mic.

Now this IS seriously interesting. I’d never thought of it this way, but there is an inverse relationship between noise and bandwidth that, in microphones, is actually meaningful around the limits of our hearing. If we extend the bandwidth, we reduce the resolution! Yet another engineering tradeoff! Damn!!

Meanwhile, I’ve been hard at work on wide-dispersion lenses for tweeters, and am passionately interested in their power response. The default tweeter we use for prototypes is roughly flat to 30 kHz. However, we’ve found that at that frequency the tweeter’s beamwidth is down to a few degrees and its power (relative to, say, 15 kHz.) is practically nothing.
What About Total System Response?
What does this all mean from a system bandwidth standpoint? Well, we’ve got to define “the system.” It includes the microphone, all the audio devices, and finally, the loudspeakers.

Music recording microphones are, at their best, limited to perhaps 30 kHz., and the really quiet microphones that we’ve come to know and love are going to cave in at about half that. Analog audio gear, except for recorders, can go casually to >>100 kHz., no problemo. Digital gear is band-limited to 20 kHz., unless it’s 96 kHz. stuff, in which case it’s good to about 45 kHz. Good loudspeakers make it to about 15 kHz. with any kind of reasonable power output. That pretty much describes the high-frequency limits of the system – all the stages seem to do OK to 15 kHz., but it gets ratty and inconsistent above that. And humans seem to hear up to about 17 kHz. on average, but few of us hear past 20 kHz. at best.

This is a good point. Before you can convince the subjectivist that the harmonic reinforcement/cancellation effect of the sampling rate has a negligible effect, you have to make them realize that their very own ears act as a low pass filter that will enforce its own harmonic reinforcement/cancellation well before the sampling rate does. ;)

Um and how do you hear high frequency air pressure changes that the ear doesn't respond to? Like trying to detect AM high frequency radio sidebands by listening to the carrier frequency. Nice invention if you can do it.

Why is this hard to explain? You do not hear the higher frequencies; instead, because they amplify or partially cancel certain audible frequencies, it's this modification of the audible harmonic frequencies that you are hearing. Alteration of harmonically reinforced (or cancelled) sounds changes their timbre.

And Mr. Hanky: the fact that your auditory apparatus (from the ear to your brain) acts as a low pass filter is irrelevant to what I am saying, since A) harmonic interaction has already occured to the pressure waves by the time your organic "low pass filter" engages, abd B) it's always the same, and as a constant will not effect apparent differences in imaging.

I am not supporting the subjectivist paradigm; I am also not arguing that room effects, electronics, recording quality, sampling during mixing, and speaker performance don't all have an affect. I don't believe that $1200 interconnects supply a tangibly better sound than $30 ones.

What I do know is that uncompressed and lossless tracks sound much better on a decent system than the lossy compression we've heard so far. Whether higher sampling rates on uncompressed tracks would be noticeably better is (for the third time this thread!) something I am not at all convinced of, but I still want the audio mix to always exceed what my system has the ability to resolve, rather than have the software be the weak link. Hence I would prefer the higher sampling rate, while freely admiting that I have no data that it will sound better on my HT setup.

A) harmonic interaction has already occured to the pressure waves by the time your organic "low pass filter" engages,...

The key here is that the "organic" low pass filter will create an additional layer of interference to out-of-band harmonic interaction, which will trivialize any intereference that would have come from the "insufficient" sampling frequency.

...abd B) it's always the same, and as a constant will not effect apparent differences in imaging.

It may not even be the same between time 1 vs time 2 or even from left ear to right ear, if threshold sensitivity is induced.


What I do know is that uncompressed and lossless tracks sound much better on a decent system than the lossy compression we've heard so far.

It's certainly possible (and personally, I am in the pcm/lossless bandwagon, as a preference, as well), but the reason for the difference will have to come from something other than the mathematics/physics/whatever of sampling rate and bit depth. There is just no support for it.

Why is this hard to explain? You do not hear the higher frequencies; instead, because they amplify or partially cancel certain audible frequencies, it's this modification of the audible harmonic frequencies that you are hearing. Alteration of harmonically reinforced (or cancelled) sounds changes their timbre.
There is only one column of air molecules residing in your auditory canal, and those molecules are vibrating at the rates of the frequencies impinging on them. The inner ear is detecting the component frequency pressure variations. It doesn't matter how many frequencies are present, individual frequencies are detected, as per the pressure fronts.

Now if there is a fixed phase relationship between certain frequencies you will hear a psychoacoustical beat. There is no pressure front with the beat frequency. If every frequency affected every other, in terms of pressure front, you would have just noise.

So the ear has to be able to detect both frequencies, and if one is supersonic, it doesn't matter. IMO

Edit: left in a sentence about harmonics on a discussion I took out

There is only one column of air molecules residing in your auditory canal, and those molecules are vibrating at the rates of the frequencies impinging on them. The inner ear is detecting the component frequency pressure variations. It doesn't matter how many frequencies are present, individual frequencies are detected, as per the pressure fronts.

Now if there is a fixed phase relationship between certain frequencies you will hear a psychoacoustical beat. There is no pressure front with the beat frequency. If every frequency affected every other, in terms of pressure front, you would have just noise.

So the ear has to be able to detect both frequencies, and if one is supersonic, it doesn't matter. IMO

Edit: left in a sentence about harmonics on a discussion I took out

The fact that this discussion has reached the molecular level, is indicative of the incredibly minute, theoretical, alleged "differences" we're now debating. Color me skeptical, but relatively speaking, humans have really lousy hearing.......now, if dogs and hamsters were into AV.......and could type.....

The best way to tell the difference:

Listen to Soprano HD DVD 1.5mpbs DD+
Listen to Soprano Blu-ray 4.6mbps PCM

The best way to tell the difference:

Listen to Soprano HD DVD 1.5mpbs DD+
Listen to Soprano Blu-ray 4.6mbps PCM

Sure, but do it in a double blind, scrupulously level matched procedure with scientific controls. The results would be far more convincing to objectivists, and subjectivists alike, wouldn't you presume? :)

The best way to tell the difference:

Listen to Soprano HD DVD 1.5mpbs DD+
Listen to Soprano Blu-ray 4.6mbps PCM

The problem is that you can't be sure of the improvements if the volume levels aren't precisely matched. Just like when the review of The Departed came out and the reviewer compared the BD PCM track with the HD-DVD TrueHD track. He claimed that the PCM track was fuller and more robust sounding. Since the TrueHD track is likely bit-identical to the PCM, the reviewer was probably hearing a difference in player output volume or a variance in the mastering volume on the PCM encode.

it represents exactly what is in there maybe if you knew how to read you would see this

I only have trouble reading when your frustration results in numerous spelling, punctuation, and syntax errors. That graphic is clearly a fictional product of your imagination.

I hope you will stop repeating that, since it isn't true. Two Universal titles that came out before "King Kong" and had TrueHD were already discussed before your post. And now Universal has put lossless on "Alpha Dog", which comes out May 1st, according to the highdefdigest review. Kind of goes against the argument that Universal is against TrueHD when the do it for 2 titles, announce it for 2 more and cancel it, then do it for another title.
--Darin

Universal clearly doesn't have a vested interest in releasing HD DVD movies with TrueHD audio. How many Universal films have TrueHD? 3? maybe 4? Out of 97? Come on, Universal is clearly not big in releasing TrueHD tracks. Just like Warner Bros. Sure they could of released TrueHD or PCM on Happy Feet, Superman Returns, The Fountain, etc on Blu-Ray, but they didn't.

You Blu-Ray fanboys just need to get over it.

Universal clearly doesn't have a vested interest in releasing HD DVD movies with TrueHD audio. How many Universal films have TrueHD? 3? maybe 4? Out of 97? Come on, Universal is clearly not big in releasing TrueHD tracks. Just like Warner Bros. Sure they could of released TrueHD or PCM on Happy Feet, Superman Returns, The Fountain, etc on Blu-Ray, but they didn't.

You Blu-Ray fanboys just need to get over it.

If it's not a bandwidth issue, it might be a "planned obsolescence" "business opportunity" for Universal, Warner, (and Sony for that matter) to squeeze another purchase out of y'all with the re-release of existing HD DVD/Blu-Ray movies with "Improved" TrueHD/DTS HD MA/uncompresssed 24 bit LPCM sound.

24 Bits is better.

Even if current systems (and future!) can't resolve it there are sufficient reasons to want the full 24 bits.

The real world practise in studio's and during mastering is that everything is going 24 bits.
Why let the signal suffer from a downconversion to 20 or 16 bit?

Although 16 bit and 20 bit are sufficient for excellent sound quality both can be considered 'brittle'. Additional digital processing in an AV amp can degrade the signal quality. 24 bits will give more headroom to play with with only a minimal loss in signal quality.


Although we can't really record below 20 bits resolution without noise and other issues, processed signal output can drop below the 20 bits level without signal degradation.

Another much overlooked fact is that 24 bits does not only improves noise level and signal resolution, it also improves transient/timing response. Roughly speaking adding one bit gives the same improvement in timing/and transient response as a doubling of the sampling frequency.

Um, harmonic reinforcement and cancellation is extremely basic physics, generally taught to science majors in their second semester of physics. Not exactly rocket science (although some of that is covered too :D )

Because of this, clipping of "inaudible" notes and other forms of compression (including limitations in the original recording, and the range of sounds your speakers and electronics can reproduce) are largely what are responsible for making recorded music sound different from live sound.
I keep on seeing these "inaudible note" declarations. Give us a source, or stop it.

Another much overlooked fact is that 24 bits does not only improves noise level and signal resolution, it also improves transient/timing response. Roughly speaking adding one bit gives the same improvement in timing/and transient response as a doubling of the sampling frequency.
Source? Transient response is frequency and phase based. If the frequency isn't in the sample range, adding dB's to it won't help.

Agreed.

Heh, I'd like to see the math on how the addition of 1 bit affects the timing/transient response, equivalent to the doubling of sampling frequency. :D

It's hard to ignore that this topic seems to have attracted the "green marker pen" crowd, big time.

I keep on seeing these "inaudible note" declarations. Give us a source, or stop it.

Testing of the audibility of ultrasonic harmonics has been done in the academic world. A simple google search brings up an IEEE paper on the subject, but you have to pay to get it, or be an IEEE member.

While the audibility of ultrasonic harmonics is not yet fully accepted science, that is the nature of the science business. It is possible to be a skeptic, and yet also keep an open mind to all the interactions of physics.

Here is an article about upsampling, that takes a different view of ultrasonics. It does not subscribe to ultrasonics being audible, but rather that their presence affects the distortions inherent in DACs. Advancement in science is often found by looking deeper into a subject, rather than accepting a few basic axioms as the only truth. Okay -- we can't in general hear ultrasound. But, can its presence or lack of it affect any part of the playback system? Can its time smear affect (puting energy before a transient, for instance -- which never happens in nature) be noticed by the brain? Is the ear really such a perfect filter that it never has nonlinearities of its own, which can, under some circumstances, change the perceived tonal qualities of sound when ultrasonic overtones (which exist in nature) are present?

http://www.mlssa.com/pdf/Upsampling-theory-rev-2.pdf

The question is not whether 24/96 music sounds better than CD (or 48/16). It does. The real science is to find out why -- not to keep thinking that because you don't see a simple explanation for it, it must not happen. (I mean you in the collective sense -- I'm not attacking your posts specifically).

I've said before that 18 or 19 bits is enough for the final cut, as long as more bits are used in mastering. But I firmly believe that 96 kHz sampling is needed (or at least more than 44, and better than 48 is desired). Whether its the in-band distortions that the anti-aliasing filters cause during recording, the ultrasonic time-smear, or positive (or negative) nonlinearities in the playback system or our ears -- that's the question.

Agreed.

Heh, I'd like to see the math on how the addition of 1 bit affects the timing/transient response, equivalent to the doubling of sampling frequency.
.

I agree, it can't.

It's just as possible that why you think it sounds better may not have anything to do with "24" or "96", altogether. Who knows?...maybe 24/96 is "not enough", either. Maybe these mystery qualities need 10x those specs? This specs race could go on forever, once you eliminate the basic math/physics from the phenomenon.

If it's not a bandwidth issue, it might be a "planned obsolescence" "business opportunity" for Universal, Warner, (and Sony for that matter) to squeeze another purchase out of y'all with the re-release of existing HD DVD/Blu-Ray movies with "Improved" TrueHD/DTS HD MA/uncompresssed 24 bit LPCM sound.

There is an extended edition of King Kong on SD DVD which is 28 minutes longer than the original. It hasn't been announced for HD DVD, but my guess is Universal is going to double dip that title eventually.

It's just as possible that why you think it sounds better may not have anything to do with "24" or "96", altogether. Who knows?...maybe 24/96 is "not enough", either. Maybe these mystery qualities need 10x those specs? This specs race could go on forever, once you eliminate the basic math/physics from the phenomenon.

Yes, you could be right, but I don't think we are divorced from basic physics here. If you want to contend that 16/48 is enough, provided we fix all other distortions in the playback and recording system, that is certainly one theory. It sticks to the simple axioms that we can't hear ultrasonics nor high-frequency phase distortions, and blames various other system bugaboos for any reduction in sound quality over 24/96. However, there is another camp that states that the simple axioms are not complete.

Either way, we don't have perfect recording or playback systems, so 24/96 is better. Should we just have a sampling-rate war? No -- that is likely not justified. But, with the 80 db/octave brick-wall anti-aliasing filters used during CD recording, it is amazing that it gives as good sound as it does (but its taken 25 years to get to reasonable sound).

As I read the paper, the theory applies to encoder error issues. Once those errors are avoided they aren't introduced again in the decoder are they? In any case, the higher frequencies haven't been preserved in the decode (have they?)

As I read the paper, the theory applies to encoder error issues. Once those errors are avoided they aren't introduced again in the decoder are they? In any case, the higher frequencies haven't been preserved in the decode (have they?)

I wouldn't take the paper as proof of anything, but it makes a good argument. It is also written in 2000 or so, which is ancient by the rate of technology advancement ;)

No, the paper applies to reconstruction via upsampling. It states that by doing a particular type of upsampling, that images the in-band signal up to the ultrasonic, and then only slowly rolls off this ultrasonic signal, the sound quality is increased in DACs. It claims this is not due to reduced time smear, because time-smear was already introduced by the anti-aliasing recording filter, and cannot be removed. It also claims time smear is present in loudspeakers to a larger degree (which is true, but anti-aliasing and brick-wall reconstruction puts energy before a transient -- something that never happens in nature, nor in speakers).

The article claims that sound quality improvement is due to an ultrasonic dithering of the amplitude level, which reduces audible distortions from DAC nonlinearities. In this theory, the content of the ultrasonic energy is not important. It is just important that there be more of it during loud passages, and less during quiet passages.

My bad, should read instead of skimmed. But still, it doesn't say that the supersonic frequencies are audible, directly or indirectly, just that they might avoid codec errors that can be introduced by filtering. There may be other solutions, and IMO the effect must be unnoticeable in most applications anyway, or else we would not have doubts.

Could you paraphrase the IEEE paper?

anti-aliasing and brick-wall reconstruction puts energy before a transient -- something that never happens in nature, nor in speakers).

Lowpass filters w. associated effects are endemic to one degree from the electronics, to the speakers, and even nature (even your own ears). Oversampling is also employed to address aliasing/brickwall filter issues. So it is a bit too "blanket statement" to say x never happens in nature, nor in speakers.

My bad, should read instead of skimmed. But still, it doesn't say that the supersonic frequencies are audible, directly or indirectly, just that they might avoid codec errors that can be introduced by filtering. There may be other solutions, and IMO the effect must be unnoticeable in most applications anyway, or else we would not have doubts.

Could you paraphrase the IEEE paper?

I wish I could paraphrase it -- I'm not an IEEE member any more. However, the google search did give enough lines to see that they did notice an audible affect with high frequency harmonics. However, like all things, you have to ask "why".

Well, upsampling is "de rigeur" these days, and it clearly makes a difference in sound quality. There may be something to this article. However, it doesn't sound as good as a "real" 24/96 encode. To me, the 96 kHz encodes are smooth and natural in the highs. Very few, if any, CDs are.

Agreed.

Heh, I'd like to see the math on how the addition of 1 bit affects the timing/transient response, equivalent to the doubling of sampling frequency. :D

It's hard to ignore that this topic seems to have attracted the "green marker pen" crowd, big time.


If you guys understood shannons sampling theorem you would have understand why that is.

That's not snake oil or green marker pen, it's basic science.

Quantization introduces two errors: amplitude domain *and* time domain.
16 bit has get's you bigger errors and 24 bit gives you smaller errors.

Simply put:
Imagine two sample points.
Pass a signal that has it's 1st sample point with amplitude 3.49 and the second is -3.
With 2 bit resolution you get 3 and -3. The reconstructed signal passes zero exactly between the two samples. In this example it's easy to spot the amplitude domain error but also the time domain error. The original signal did not pass zero exactly between the two samples.

Add sufficient bits so that the two samples can be fully represented and there is no amplitude error but also no time domain error.

You can verify this easily with a piece of paper and a green marker. :p

However, it is a presumption that the accuracy is "not enough" to get to the proper zero point within an error window that happens to be many orders of magnitude smaller than other links in the chain which could cause you problems. Once you account for typical noise floors, it doesn't matter if you have "unlimited accuracy". It is essentially crossing the zero-point as close as you could possibly determine within the randomness of noise contributions. For all you know, that point is happening at 13-14 bits, let alone 24-bits.

Quantization introduces two errors: amplitude domain *and* time domain.
16 bit has get's you bigger errors and 24 bit gives you smaller errors.

Simply put:
Imagine two sample points.
Pass a signal that has it's 1st sample point with amplitude 3.49 and the second is -3.
With 2 bit resolution you get 3 and -3. The reconstructed signal passes zero exactly between the two samples. In this example it's easy to spot the amplitude domain error but also the time domain error. The original signal did not pass zero exactly between the two samples.

Add sufficient bits so that the two samples can be fully represented and there is no amplitude error but also no time domain error.

You can veryfy this easily with a piece of paper and a green marker. :p

Interestingly enough, your explanation only makes sense on a signal which is already filtered to be within the audible band. Perhaps you and the ultrasonic skeptics should agree here ;)

By doubling the sampling frequency, the original waveform does not have to filtered as heavily during recording. This is what improves the time-domain. If you want to filter your data to < 20 kHz, and then do a more accurate time-domain sampling of hacked-up waveform, be my guest. However, the increased accuracy you get will not improve the signal quality beyond even 1 bit of S/N. It can't because the inaccuracy of the amplitude can't simply be wished away.

The problem of sampling rate on the timing of zero crossings is why dithering is added to a playback (and pre-recorded) signal. Without the dithering, audible distortions are heard, because of systematic errors that can occur with the sampling. With proper dithering, the distortions become white noise, on the order of a fraction of the lowest bit.

No, the bit-depth determines noise. The bit rate determines the frequency response that you can achieve with the recording. It is this frequency response, and how you do your anti-aliasing recording filters (phase changes, etc.), that determine the theoretical limit on your time-domain accuracy.

....

No, the bit-depth determines noise. The bit rate determines the frequency response that you can achieve with the recording. It is this frequency response, and how you do your anti-aliasing recording filters (phase changes, etc.), that determine the theoretical limit on your time-domain accuracy.

I do understand that doubling the sampling frequency doubles the passband too. It also gets you lower time domain errors. It also gives a lower quantization noise on average as well as it's spread across a wider bandwidth. It goes both ways.

However doubling the sample rate is more waistfull than adding a single bit.

You really should revisit the sampling theorem basics.

In my example I assumed a legal signal that does not exceed nyquist. (Do I need a filter then? ;) ) The zero crossing was intended to symplify the example.
I hope you do understand where the time domain error is in the example and how it is reduced by adding just one bit.

Is bit depth scaled? 16 bits covers the same dynamic range as 24, or just 16 parts of the 24 bit range?

P.S. in that IEEE study I would like to know the level of the supersonic sounds. At a high level they may load the ear to the point where it responds differently to sonics.

16 bits covers one dynamic range
24 bits covers a larger dynamic range

the 16 bits contained within 24-bits covers the same dynamic range as a native 16-bit dynamic range

BTW, I'm always impressed with the good information we can get here. :)

Where are these ultrasonic sounds coming from? Most mics are rated @ 20hz - 20khz. Some are rated up to 30khz but generally are noisier than a high quality 20-20 mic.

While sounds >20kHz exist when listening to live music it is debatable if they add anything to the music either directly or indirectly. For recorded music there should be no debate. If the mic didn't record the sound then for all intents and purposes it never existed. Virtually all mics even professional grade only have a frequency response of 20Hz - 20kHz. Some push out to 30kHz but they have a higher noise floor and may not be applicable to all recording environments. There is no "perfect mic", in the world of phsyics ther are no free lunches. While you could design a mic that picked up sounds even further you would also boost the noise floor of the mic. Some mics are intentionally <20khz for the purpose of trading high frequency sounds for and even lower noise floor.

Maybe someone in the industry can chime in but I would assume that when dealing with movie soundtracks low noise would trump ultrasonic sounds and thus the mics in use would be 20-20 range and a mix of low noise floor or high dynamic range depending on the purpose.

When you combine the limitations of the mics with the limitation on consumer grade DAC (<120db dynamic range) the need for 24/96 start to look more and more for some consumer's to justify their $20K setups. 20/48 is more than enough. Sound reproduction is something that can't be solved by simply throwing more bits at it. Until we are able to design DACs that run at cryogenic tempuratures and mics that can break laws of physics by having extreme frequency response AND high dynamic range AND low noise most of those bits in a 20/48 soundtrack are simply "empty".

The best way to tell the difference:

Listen to Soprano HD DVD 1.5mpbs DD+
Listen to Soprano Blu-ray 4.6mbps PCM

The Sopranos on HD DVD is DD+ 640 kbps :(

I do understand that doubling the sampling frequency doubles the passband too. It also gets you lower time domain errors. It also gives a lower quantization noise on average as well as it's spread across a wider bandwidth. It goes both ways.

However doubling the sample rate is more waistfull than adding a single bit.

You really should revisit the sampling theorem basics.

In my example I assumed a legal signal that does not exceed nyquist. (Do I need a filter then? ;) ) The zero crossing was intended to symplify the example.
I hope you do understand where the time domain error is in the example and how it is reduced by adding just one bit.

Sorry, I misread your post.

However, your example of "time domain error" is not what I am talking about. I am not talking about increasing accuracy at the sub-bit level. We have way more bits than we need in 20 or 24 bits. The time domain error I'm talking about, is the error inherent in the filtering and reconstruction process. There are phase shifts in the passband, and time-smear ringing of frequencies close to the Nyquist frequency. The more the Nyquist frequency can be separated from the audible band the better. I'm sure there is a practical limit of usefulness, but 44.1 kHz sampling is just nuts. 48 allows twice the stop band above 20 kHz. 96 provides a whole octave of stop-band. This hugely reduces these time smears.

While some people point to speakers as having "huge time smears and distortions" as reasons why almost all digital music should be "good enough", I can say these people don't seriously listen to music on a good home system. It is possible to have distortions 1/100 the distortion power of those from speakers, and have it be audible. It's about the type of distortion, not just its magnitude, that makes our brains say "that sounds good" or "that sounds wrong". Time smear that puts high-frequency energy before a transient is never seen in nature. Creating it in a digital playback system is not a good idea. Increasing the sampling rate fixes these problems. Yes, it takes double the space, but I believe it is worth it. 20 bits is more than enough. I haven't had enough experience with 48 kHz to know if it is enough. I just know that 44.1 is certainly NOT enough.

Any sound that gets recorded by a microphone is inherently going to exhibit "pre-transient energy", due to all things having a finite lowpass filter function. It's going to happen, regardless of how high you sample the signal, unless you are using some kind of infinite bandwidth microphone coupled to a flux capacitor.

Any sound that gets recorded by a microphone is inherently going to exhibit "pre-transient energy", due to all things having a finite lowpass filter function. It's going to happen, regardless of how high you sample the signal, unless you are using some kind of infinite bandwidth microphone coupled to a flux capacitor.

The time smear by a low-pass filter is not the same as what we reconstruct in a digital system. An analog system can't create energy before a transient. It can only delay certain components of the transient. This delay is frequency-dependent, and is caused by the phase shift of the filter.

Before you say these result in the same thing, the biggest difference is there is no Nyquist frequency in an analog system. The reconstruction process for frequencies near the nyquist frequency is a ringing to both sides of the transient. While this is all nice in the frequency domain, it is something totally outside of nature in the time domain.

Nature is plagued with instances of "ringing" or more specifically, resonances. Microphones have resonances. Tweeters have resonances. It doesn't much matter if the ringing is coming from a brickwall filter or a spring-mass system. Ringing is ringing and they are no stranger to nature. This is largely a moot point, anyway, once you factor in the impact of digital oversampling as part of the reconstruction process.

This is largely a moot point, anyway, once you factor in the impact of digital oversampling as part of the reconstruction process.

Nope, oversampling can't solve the problem. One can design different oversampling algorithms, and try to find one that sounds the best, but the problem is within sampling theory itself. The closer you get to the Nyquist frequency, the less time-domain information you have. Correct reconstruction (one that does not create spurious high frequency junk) creates ringing before the transient.

Some designers have tried different oversampling algorithms, which are in fact wrong frequency-wise, but attempt to minimize this time-domain ringing. I don't know if the attempts sound better or not.

Understandably, there will be people who will never be satisfied with whatever oversampling or anything, for that matter, other than infinite bandwidth. For the rest of us, it is far better than we could ever bother to be concerned about. Microphones, mastering equalizations, perceptual data compression, amplifiers, speakers, room effects, your own ears, etc. still manage to mangle the original waveform far, far worse than even the most mediocre ad/da conversion.

Nope, oversampling can't solve the problem. One can design different oversampling algorithms, and try to find one that sounds the best, but the problem is within sampling theory itself. The closer you get to the Nyquist frequency, the less time-domain information you have. Correct reconstruction (one that does not create spurious high frequency junk) creates ringing before the transient.

Some designers have tried different oversampling algorithms, which are in fact wrong frequency-wise, but attempt to minimize this time-domain ringing. I don't know if the attempts sound better or not.

44.1 ks is not used with film. (The 44.1ks sampling frequency was selected for compliance with tv signal frequencies)
For film it's 48ks. This gives about 8khz bandwidth for filtering.
The fear for time smear is unjustified since most ad converters are oversampling anyway. The nyquist frequency is much higher as a consequence and pre filtering is far less intrusive than with the 44.1ks in the early days.

In addition the sampling theorem can be demonstrated fully in the analog domain.
If the sample is not digitized but thrown at a reconstruction filter the signal gets restored fully. There is no quantization error in that case.
However the steep reconstruction filter introduces the ringing that is a very important aspect of the reconstruction process. In the filter the ringing, that results from throwing the pulses at the steep filter, convulates in the filter and the 'summing' if this ringing results in the reconstruction of the signal between the samples. This process is all analog by nature.

Digital comes into play as it is a conveniant method for transport and storage without the signal loss associated with analog storage systems.

Sampling frequencies near the nyquist frequency with 20..24 bits lowers the high frequency distortion of the system. (My example shows why.)

...your own ears, etc. still manage to mangle the original waveform far, far worse than even the most mediocre ad/da conversion.

I think on that point, we hit an impasse at which we should agree to disagree. All "mangling" is not created equal (long live vinyl :D )

Is 24bit important for movies/ are you kidding, yes it is. think of the music score in the movie. Surrounds effects are brief, but atleast the front Left, center, and right channels should be 24/48 or 24/96. Since this is were 90% of the movie occurs, bulk of dialog and music track.

If you look up 16bit vs 24bit, you will see the major difference in the mapping of sounds and how a more accurate bitrate improves the logic and make a track feel more like analog.

~Bobby

16 bit vs. 24 bit is like bringing 2 measuring sticks to a water fountain to measure the depth. The first stick is 1 meter long with demarcations down to 1 mm. The second stick is 1 meter long with demarcations down to 1 micrometer. Either measuring stick can be used to get a depth reading, but the shear noise in the depth measurement (waves and ripples caused by the water fountain falling back into the pool) is going to be larger than 1 mm or 1 micrometer. So the lsb is a moot point, since the best deterministic reading you can take will be on the order of cm's, at the very least. So the "extra accuracy" of the micrometer measuring stick gains you nothing for the actual depth reading, except possibly encoding the noise more accurately. The mm measuring stick will give you a depth measurement that is just as good.

using an analogy as obsurd as the above is pointless, we are measuring something with pin point fluctuations in change of tone, pitch, stage, FS, and so on. Yes there is a better noise floor, but with that the perks of added Dynamic range and SNR also exist. Its a lot more complicated then you make it out to be.

Anyone interested in the difference read this site (http://www.soundblaster.com/resources/read.asp?articleid=53936&cat=1)

~Bobby

It's only absurd, because you refuse to acknowledge the implications. In fact, it gets to the very heart of the matter. The measuring stick (whether a physical reference or a mathematical reference) is only as good as the ultimate noise floor. Finer measurements beyond that are utterly wasted.

Tone, pitch, whatever have nothing to do with it. The only thing that it can measure is a specific amplitude at a specific point in time.

Which does a better job of reproducing cymbals or the overtones of horns/winds/strings etc ? 16 bit/44.1k PCM or 24/96 PCM ? Why do so many people prefer the sound of vinyl or reel-to-reel tapes vs. CD (despite the inconvenience and snap/crackle/pop). Why do recording studios bother with high resolution PCM/analog tape masters if 16/44.1k is "good enough" to capture the original music ? Why does anyone who listens to master tapes thinks it thoroughly trashes any CD playback ?

Or is everyone agreed 24/96 PCM beats the pants off of 16/44.1k with any well-recorded instrumental/vocal music ?

It appears to me we just don't have a good enough handle on how the human ear processes audio information and how to duplicate that processing with manufucturable monitoring equipment.

Which does a better job of reproducing cymbals or the overtones of horns/winds/strings etc ? 16 bit/44.1k PCM or 24/96 PCM ? Why do so many people prefer the sound of vinyl or reel-to-reel tapes vs. CD (despite the inconvenience and snap/crackle/pop). Why do recording studios bother with high resolution PCM/analog tape masters if 16/44.1k is "good enough" to capture the original music ? Why does anyone who listens to master tapes thinks it thoroughly trashes any CD playback ?

Or is everyone agreed 24/96 PCM beats the pants off of 16/44.1k with any well-recorded instrumental/vocal music ?

It appears to me we just don't have a good enough handle on how the human ear processes audio information and how to duplicate that processing with manufucturable monitoring equipment.

Studio recordings obviously ise PCM or analog masters to have a great dynamic range to work with. The question is does the average consumer have the equipment to hear the difference.

.....The question is does the average consumer have the equipment to hear the difference.
This is the right question. And average doesn't mean J6P or JQP, it also means [average] conditions (program, gear, environment, expectations, and ear). I would love to see tests and even "certification" that if I buy this, and do this, studies show I will think the difference is worth it. But if it were obvious, I think manufacturers would rush to produce the gear at mass consumption prices.

In the absence of comparative tests and mass-market attempts at better gear I am inclined to believe in snake oil. :( Bose rules, xxxx (name your favorite) drools.

Had Bose in my motorhome, tho' :)

Studio recordings obviously ise PCM or analog masters to have a great dynamic range to work with. The question is does the average consumer have the equipment to hear the difference.

I'd suspect your "average" HT enthusiast (e.g. AVSforum posters) who spends a pretty penny on decent audio/video gear will hear a difference.

How do you prove it ? My ears tell me subjectively higher resolution recordings (SACD, DVD-A, DAD, vinyl) give me more "musically-relevant feedback" than lower resolution CDs. That's all subjective; I can't say I've done any ABTs because they would be extremely difficult to set up. Ideally, I'd like to have the live performers in the room to establish a reference. but failing that... CDs vs SACD/DVD-A of same music. Oops, try to find suitable audio material which has been given the same TLC with the original masters and subsequently converted to CD and DAD/DVD-A/SACD media. And of course, with hybrid CDs/SACDs, people will accuse the disc authors of deliberately screwing up the CD layer so that the SACD sounds better (which in general they do).

Also, you are more likely to pickup the differences with well recorded instrumental music (e.g. jazz/classical) and less likely to hear differences with your highly compressed, poorly recorded (typically) rock/pop music .

I've had the privilege of hearing a master tape (1 1/2" reel to reel) on someone's (admittedly above average) high end system and the music was incredible, as close to live as I've ever heard. Why can't I even come close to that sound on CD if 16/44.1 was "good enough" ?

I'd suspect your "average" HT enthusiast (e.g. AVSforum posters) who spends a pretty penny on decent audio/video gear will hear a difference.

How do you prove it ? My ears tell me subjectively higher resolution recordings (SACD, DVD-A, DAD, vinyl) give me more "musically-relevant feedback" than lower resolution CDs. That's all subjective; I can't say I've done any ABTs because they would be extremely difficult to set up. Ideally, I'd like to have the live performers in the room to establish a reference. but failing that... CDs vs SACD/DVD-A of same music. Oops, try to find suitable audio material which has been given the same TLC with the original masters and subsequently converted to CD and DAD/DVD-A/SACD media. And of course, with hybrid CDs/SACDs, people will accuse the disc authors of deliberately screwing up the CD layer so that the SACD sounds better (which in general they do).

Also, you are more likely to pickup the differences with well recorded instrumental music (e.g. jazz/classical) and less likely to hear differences with your highly compressed, poorly recorded (typically) rock/pop music .

I've had the privilege of hearing a master tape (1 1/2" reel to reel) on someone's (admittedly above average) high end system and the music was incredible, as close to live as I've ever heard. Why can't I even come close to that sound on CD if 16/44.1 was "good enough" ?

The reason why I bring this up is because I don't know if it is worth the capacity usage for 24/96 PCM or even 24/48 PCM for the average consumer. One of the hallmarks of DVD was bonus features. HD DVD and Blu-Ray obviously have the space for lossless, with HD DVD useing TrueHD. However I don't want it to get to a point that the next gen HD format supporters get so focused on the niche groups that they start slashing special features from releases or release HD DVD players and Blu-Ray players with extremely high end outputs and features that most of the consumers in the market place doesn't have a use for.

HDMI 1.3, 7.1 analog outputs, 24/96 PCM, 24/48 Dolby TrueHD is all great stuff. But it raises cost, and uses up alot of memory.

I see a similarity between the HD formats and politics. If both play to their base too much, then neither will get support of the majority. I know that is kind of a poor comparison, but I hope it gets the point across.

HDMI 1.3, 24/96 PCM, 24/48 Dolby TrueHD is all great stuff. But it raises cost, and uses up alot of memory. But if the new formats can't provide space for it all, then cut the extras out. WHo has time for 35 hours of Matrix featurette garbage?


For some of us who spend a considerable sum on the audio side and construct a room with the quietest of noise floors, this is the Holy Grail. This is what it is all about. Cost? who cares...I'll pay...Heck, I used to pay up to $120 for LD special editions.

Audio matters. You don't know what you don't know...and if you did, you'd be impatiently waiting for hi res audio tracks.

HDMI 1.3, 24/96 PCM, 24/48 Dolby TrueHD is all great stuff. But it raises cost, and uses up alot of memory. But if the new formats can't provide space for it all, then cut the extras out. WHo has time for 35 hours of Matrix featurette garbage?


For some of us who spend a considerable sum on the audio side and construct a room with the quietest of noise floors, this is the Holy Grail. This is what it is all about. Cost? who cares...I'll pay...Heck, I used to pay up to $120 for LD special editions.

Audio matters. You don't know what you don't know...and if you did, you'd be impatiently waiting for hi res audio tracks.

Yeah 35 hours is an overkill, but none is a disgrace.

Are there any music videos with 16/48 soundtracks which were derived from 24/96 (or better) Masters and which also has been recorded on DVD-A with either 24/48 or 24/96 tracks ?

The closest (and not very) comparison I have is the Pat Metheny Blu Ray with a "DTS HD" soundtrack (whatever that is) played back on my "DTS-HD capable" (?) player with a different Pat Metheny DVD-A with 24/96 soundtrack; there might some of the same material recorded, but different recordings, different microphones/venues etc. Also the playback DVD-A hardware is different.

For some of us who spend a considerable sum on the audio side and construct a room with the quietest of noise floors, this is the Holy Grail. This is what it is all about. Cost? who cares...I'll pay...Heck, I used to pay up to $120 for LD special editions.



This describes less than 1% of the home video market (a generous estimate). The large majority DO care about cost and value, therefore the studios will pack the disk full of extras and exclude the 24-bit sound when they can.

...

HDMI 1.3, 7.1 analog outputs, 24/96 PCM, 24/48 Dolby TrueHD is all great stuff. But it raises cost, and uses up alot of memory.

....

Actually it doesn't raise cost. Since most studios work with 24bits anyway it's actually more 'costly' to do an additional downconversion to 16 bits...

Memory isn't an issue anymore.

Actually it doesn't raise cost. Since most studios work with 24bits anyway it's actually more 'costly' to do an additional downconversion to 16 bits...

Memory isn't an issue anymore.

It appears he was speaking of player costs, not the cost to the studio.

It appears he was speaking of player costs, not the cost to the studio.

Doubt it, most if not all DAC's and DSP hardware is processing 24 bits.

Going from 16 bit to 24 bits is a free lunch these days.

Doubt it, most if not all DAC's and DSP hardware is processing 24 bits.

Going from 16 bit to 24 bits is a free lunch these days.

In the days where manufacturers will make an engineering change to eliminate a $0.10 item from a BOM, you think that the things MichaelHDDVD mentioned (HDMI plugs, RCA analog plugs, licensing fees for TrueHD, additional memory, etc...) tranlsate to zero cost?

I disagree.

In the days where manufacturers will make an engineering change to eliminate a $0.10 item from a BOM, you think that the things MichaelHDDVD mentioned (HDMI plugs, RCA analog plugs, licensing fees for TrueHD, additional memory, etc...) tranlsate to zero cost?

I disagree.

Who claimed zero cost?
24bit processing doesn't translate into higher cost these days.

Cost discussion about hardware is pointless anyway. Per spec hardware must be able to handle 24bit audio.

It's not the dsp costs, it's the new hdmi receiver, high quality speakers, the room treatment costs & noise envirnonment. Lotta bucks to hear the subtleties, even if you are under 30 and can still hear them.

Who claimed zero cost?
24bit processing doesn't translate into higher cost these days.

Cost discussion about hardware is pointless anyway. Per spec hardware must be able to handle 24bit audio.

OK, I disagree that it's zero ADDITIONAL cost.

As WayneL has pointed out, there's more to it then simply assuming that 24bit processing is "a free lunch" because one component in the chain can do it.

This describes less than 1% of the home video market
but this is what it means to be an AVSer

(a generous estimate). The large majority DO care about cost and value, therefore the studios will pack the disk full of extras and exclude the 24-bit sound when they can.


and that is why it is important that the next format has enough to make both happy.


PS the thing is that the 1% is 90% of who is buying now. If we don't fight for quality now, we will definetly not get it when J6P joins in

I have a copy of the Chinese movie "The Banquet" with DTS 96/24 audio and it rocks! It easily beats the best True HD titles I have on HDDVD. But then this may be the case because I am using the 5.1 analog outs as my receiver does not have an HDMI input. Anyway, you guys who are into awesome movie audio should check this one out. It will not disappoint!

What I do know is that uncompressed and lossless tracks sound much better on a decent system than the lossy compression we've heard so far.

What have 'we ' heard so far? Msot of the data so far suggests that lossy compression, done well, becomes inaudible to the vast majority of people on the vast majority of music, once the bitrates rise above 160 kpbs or so. Have you actually blind-tested yourself to find that this is true or false for you?

I've ABX's mp3s that I've made (VBR, ~190 kbps target), using the recommended LAME encoder and settings from hydrogenaudio.org, playing them on good Audio-Technica headphones from an M-Audio 2496 soundcard, using foobar 2000 as the playback software in kernal streaming mode -- and I have been unable to successfully ABX source from mp3 in 16-trial test runs (using WinABX), except when using certain 'killer signals' used for LAME codec development. This is perhaps not unexpected, as my hearing is pretty poor above 16 kHz -- typical for my age group.

Actually it doesn't raise cost. Since most studios work with 24bits anyway it's actually more 'costly' to do an additional downconversion to 16 bits...

Memory isn't an issue anymore.

I meant putting analog outs and HDMI 1.3 on HD DVD and Blu-Ray players raises the costs. Which could possibly alienate people who want to get a HD DVD or BR player due to the cost, especially for a feature most households don't have.

You talking about all those audio recording engineers and audio listening enthusiasts who'd swear 24/96 PCM audio is a huge improvement over 16/44.1k CD ? If so, there are a heck of lot of people who are delusional; including myself. In my experience, 24/96 playback thoroughly trashes 16/44.1 with any well-recorded music in my system.


The only proven benefit of higher bitrates -- and the REAL reason it's used in pro audio -- is because you can do intensive digital processing during production stages, without fear of accumulating audible errors. The need for it at the delivery end, in a home listening environment, is highly questionable, as long as proper dithering was used int he final production 24-->16bit step.


Believe what you want. A lot of us are convinced high resolution audio is desirable, even with movie soundtracks. If you ever get a chance, try to audition the Eagle's "Hotel California" DVD-V w/DTS soundtrack vs. the DVD-A 24/96 soundtrack on a decent system.

For this comparison to eb at all fair, you'd have to ensure that the two mixes are identical in level in all channels coming off the disc all the way through to loudspeakers. That may not be true either because they were mixed differently, or because the rig processes DTS and DVD-A differently (e.g., outputting correct levels for LFE for one format but not the other).

And of course the comparison would HAVE TO BE BLIND.

I meant putting analog outs and HDMI 1.3 on HD DVD and Blu-Ray players raises the costs. Which could possibly alienate people who want to get a HD DVD or BR player due to the cost, especially for a feature most households don't have.

These 'costly' features you mention have nothing to with the 24 bit digital audio.

Absolutely true, but not really relevant in this case, as 1) the PCM tracks are not louder, and 2) there are several very different listening levels used in my home theater (depending on audience composition...e.g. fiance = quieter), and the quieter PCM still invariably demonstrates better localization and imaging than louder compressed tracks. Still, it's worth a test when I get some free time.

The thing about the 'loudness effect' is that the listener does not perceive the louder presentation as being louder. And judging relative loudness by 'ear' is not particularly reliable if the differences are small. Level matching needs to be done with equipment, to be accurate. So I'd say you don't know that the PCM tracks weren't louder in one or more channels. And too there's also the pervasive effect of bias, which requires that comparisons be done blind to get reliable answers.

It has been proven that when you teach a person what to listen for, their ability to tell the difference improves. A lot of folks go into dbt without the knowledge of how to listen for differences. Small details like reverberation tails, transient attacks, and ambience are not what people are listen for, they always listening to things that are easy to code. This is why dbt are so difficult for most folks.

It is absolutely the case that training in audio discrimination improves one's ability to hear the differences trained for. That would likely improve performance on a DBT.

However, the vast majority of subjectivists ALREADY CLAIM TO HEAR the 'small differences' in question; several in this very thread. If they then fail to pass a DBT, what does that say about the claims they have ALREADY made about their discriminative ability?

Why does anyone who listens to master tapes thinks it thoroughly trashes any CD playback ?


It's not true that everyone does, in the audio production world.

And since we're asking questions, how many audio 'engineers' who make such claims as above, actually test them with a properly performed, statistically significant blind test? (And how many actually have any engineering/science/digital information processing training?)


Or is everyone agreed 24/96 PCM beats the pants off of 16/44.1k with any well-recorded instrumental/vocal music ?


No, everyone is not agreed on that. THere's general agreement that >16bits is a smart idea for production work, if the delivery format is 16 bit. THere's good, sound technical reasons behind that belief.

It's not the dsp costs, it's the new hdmi receiver, high quality speakers, the room treatment costs & noise envirnonment. Lotta bucks to hear the subtleties, even if you are under 30 and can still hear them.

:rolleyes: Obviously some of take better care of physical ability to hear....

Room treaments in a world of DIY are easier then you state, I could build my own acoustic panels at less then $20ea... Speakers are getting better by the day, so its easy to find accurate low cost solutions, for instance Definitive, paradigm, B&W, Rocket, etc.

Equipment, electronics with new receivers and preamps capable of 110-120dB noise floor and 100dB+ SNR and low THD+N can you honestly say there is a tremendous gap between studio equipment and home consumer equipment in terms of playback quality? I think its ironic, but Jeffs probably got as much if not more money invested in his home theater then most studios do! Sure I am not up there in terms of his system but I know I can hear the difference, even my ather who is 61 can hear it too, so..

~Bobby

I wish I could paraphrase it -- I'm not an IEEE member any more.

Please at least cite the author/title, so others can look it up, and see, for example, if it's a paper (peer-reviewed) or a preprint (not). I'm thinking it can't be Oohashi's paper because that wasn't published in JAES. If it's one of (Meridian) RObert Stuart's papers, I've read them and IIRC they contain no DBT data, or bibliographic references to same -- only anecdotes and theoretical reasons why higher sampling rates could be audibly better.

...

No, everyone is not agreed on that. THere's general agreement that >16bits is a smart idea for production work, if the delivery format is 16 bit. THere's good, sound technical reasons behind that belief.

I'm all for 24bit. Simply because the 24bits allow for better processing quality in my AV Amplifier. 48Ks or 96ks is less benificial.

And it sounds much better than 16 bit. There is simply no contest.

As long as no one produces credible tests showing 24 bit is better, it remains an area of opinion.

BTW, turn off those equipment cooling fans too

As long as no one produces credible tests showing 24 bit is better, it remains an area of opinion.

BTW, turn off those equipment cooling fans too

Anyone can drum up plots from measurements that proofs this showing plots with lower noise floor and less distortion.

This is not a matter of opinion or blind faith.
>16 bits is better this is fact. Professional audio engineers now about this for quite some time now.

Reports about credible tests for this can be found if you are willing to search for it.

...and 32 bits is better than 24 bits, right? How, oh how, do professional audio engineers bare the task of compromising at only 24-bits?

I believe 20/48 is probably the best an individual is going to be able to make out, but I'm all for 24/48 whenever possible.

24 bits is preferred to 16 bits. Period.

As enthusiasts, its another ideal to dream about for movie soundtracks.

As equipment and room acoustic / DSPs improve, it'll make significant improvements to modest systems.

It is absolutely the case that training in audio discrimination improves one's ability to hear the differences trained for. That would likely improve performance on a DBT.

However, the vast majority of subjectivists ALREADY CLAIM TO HEAR the 'small differences' in question; several in this very thread. If they then fail to pass a DBT, what does that say about the claims they have ALREADY made about their discriminative ability?

You have an entire cottage industry of music enthusiasts who'll swear vinyl and reel-to-reel provide far more "musical information" than any 16/44.1k CD. Can they ALL be wrong ? I don't even play the vinyl game, but I understand where they coming from simply from my own subjective experience base.

I'd love to be able to participate in objective DBT audio tests but these have to be very difficult to set up properly and I am not willing to exert the effort to do so. I doubt I have the capability to pull it off even if I were serious about trying it. I'd have to have access to the master tapes, the ability to play them back, the ability to transfer to 24/96 and/or 24/48 and/or 20/48 or DSD and/or 24/192 and/or 16/44.1 with equal amounts of the TLC in the transfer process, then get more than a few volunteers to subject themselves to listening tests; which volunteers having been trained to appreciate the vocal/instrument overtones, soundstaging, timbre, transient responses, "air/ambience" etc...

As it is, I try to support live performances to establish the reference, then compare to "similar" recordings on consumer media (e.g. CD or SACD) to see how close the consumer media gets to the live performance. CD ain't even close, but SACD does get me marginally closer.

:rolleyes:

Jeff,

We all know that saying "You can lead the horse to water..."

Some people will argue an argue but never give. Apparently these people also think that 16bit PC color is more the adequete and 24bit/32bit color isn't noticible change. Or that a Porche 911 isn't a noticible improvement in handling over a Saturn or Kia.....

I guess some people just settle and forget the finner things in life, you and I we persue evolving happiness of IMPROVED technology...

~Bobby

The scale is a key context here.

If it were really down to a comparison between a Porsche and a Kia, there wouldn't be much to dispute (and the math and physics would be there to support you, as well).

What we are really down to here, is comparing a land-speed jet car to a MacLaren F1 on a freeway with typical traffic load. The jet car is certainly superior to the F1 for ultimate topspeed, but you will NEVER get a chance to see top speed in either car on a public highway, anyway. Both are beyond the needs and ultimate performance of the typical driving enthusiast and probably most experienced drivers, as well. Attributing the top speed capabilities to any driving benefits while on that hwy is utterly moot, similar to attributing the 16 vs. 24 bit depth capabilities to any audio benefits on even a premier listening room.

This is not to say that you aren't hearing some sort of difference. It's just that there is no basis, so far, that bit depth is responsible for it.

As far as horse analogies, you can also say that a horse lacks the mental faculty to distinguish between a mirage and a lake of water. It's pointless to reason with one, at that level. Mirage may as well be as "real" as real water.

As a car analogy maybe a Hummer carrying one passenger is apropos to bigger is better.

:rolleyes:

Jeff,

We all know that saying "You can lead the horse to water..."

Some people will argue an argue but never give. Apparently these people also think that 16bit PC color is more the adequete and 24bit/32bit color isn't noticible change. Or that a Porche 911 isn't a noticible improvement in handling over a Saturn or Kia.....

I guess some people just settle and forget the finner things in life, you and I we persue evolving happiness of IMPROVED technology...

~Bobby

So what about 32bit sound? That would allow 192db of dynamic range. Most mastering is now moving to a 32 bitdepth and some highend solution are now 64bit floating point. Should we move to that also?

If bigger is always better why haven't we moved to 64bit color depth on computers? Why settle for a crappy 16.7 million colors when we can have billions or even trillions of colors we can't see?

Studios don't use 24, 32, or 64 bit in mastering because any audio sample has that much dynamic range. They use it because when they apply effects (equalization, stereo imaging, noise shaping) it allows a finer level of control. They can also worry less about clipping or the effect that any changes have to the noise floor.

I know you don't want to believe it but there is no improvement between 20/48 and 24/96. There also won't be any improvement in 10 years when you are convinced that 32 or 64/192 is needed because the "pros are using it".

So what about 32bit sound? That would allow 192db of dynamic range. Most mastering is now moving to a 32 bitdepth and some highend solution are now 64bit floating point. Should we move to that also?

If bigger is always better why haven't we moved to 64bit color depth on computers? Why settle for a crappy 16.7 million colors when we can have billions or even trillions of colors we can't see?

Studios don't use 24, 32, or 64 bit in mastering because any audio sample has that much dynamic range. They use it because when they apply effects (equalization, stereo imaging, noise shaping) it allows a finer level of control. They can also worry less about clipping or the effect that any changes have to the noise floor.

I know you don't want to believe it but there is no improvement between 20/48 and 24/96. There also won't be any improvement in 10 years when you are convinced that 32 or 64/192 is needed because the "pros are using it".

Poor, failed analogies....Or if you must use analogies lets look at things in reverse....When lowly cable was 240i...would anyone have conceived of the improvement 1080P would bring? 12X greater resolution??? Well, its here and I can see that even 1080P can be improved upon....!!

Let's talk about here and now. 24 bit is where it is at. Equipment, DSPs, room design, etc are all so much improved relativeto 5 years ago... It makes these new HD audio formats quite useful for the enthusiast who has built his system towards excellence. As khellandros66 says, the hobby is about pursuing excellence... If you you are content in your 16 bit world, so be it..But there is better..and it is already here.

Tell me about your system? Is it that your system lacks the resolution to even have 24 bit recordings make a difference?

Nearly all systems (with perhaps the exception of something sitting in a laboratory) are woefully inadequate to resolve genuine 24-bit audio. At the least, you need something with 140 dB s/n ratio, preamp levels approaching 100 volts, 1000's of W of amplification/channel (that's plural, not A thousand), speakers that exhibit zero thermal compression under 1000's W of input, and the ability to convert that power into a max spl over any given point in the audible spectrum. Now couple that with a listening room that can absorb 140 dB spl's w/o coloring the sound unnaturally and ambient C-weighted noise levels remaining in the single digits.

No doubt, your system is impressive, but I doubt it even remotely meets those specs to properly render "24-bit sound".

Poor, failed analogies....Or if you must use analogies lets look at things in reverse....When lowly cable was 240i...would anyone have conceived of the improvement 1080P would bring? 12X greater resolution??? Well, its here and I can see that even 1080P can be improved upon....!!

Let's talk about here and now. 24 bit is where it is at. Equipment, DSPs, room design, etc are all so much improved relativeto 5 years ago... It makes these new HD audio formats quite useful for the enthusiast who has built his system towards excellence. As khellandros66 says, the hobby is about pursuing excellence... If you you are content in your 16 bit world, so be it..But there is better..and it is already here.

Tell me about your system? Is it that your system lacks the resolution to even have 24 bit recordings make a difference?

Please.....Technology hasn't really changed all that much in the last five years, certainly not to the point where you'd be able to consistently distinguish a variety of dithered 16 bit music samples from their 24 bit sources. You'd have a very good chance using test tones, and/or program material at ear-splitting SPLs to discern the lower noise floor of 24 bits, but music at real-world levels?.....I think you're kidding yourself.

Nearly all systems (with perhaps the exception of something sitting in a laboratory) are woefully inadequate to resolve genuine 24-bit audio. At the least, you need something with 140 dB s/n ratio, preamp levels approaching 100 volts, 1000's of W of amplification/channel (that's plural, not A thousand), speakers that exhibit zero thermal compression under 1000's W of input, and the ability to convert that power into a max spl over any given point in the audible spectrum. Now couple that with a listening room that can absorb 140 dB spl's w/o coloring the sound unnaturally and ambient C-weighted noise levels remaining in the single digits.

No doubt, your system is impressive, but I doubt it even remotely meets those specs to properly render "24-bit sound".

Well, I am pretty much there.... as are many serious hobbiests who have such systems.

But that is besides the point...Is it your contention that a 16 bit recording would sound identical to an identical 24 bit recording in such a well equipped room??

The answer is 'no'. The 24 bit will sound better on a capable system..Now will the average listener be able to resolve all that is there?? No..But they will resolve more than a 16 bit recording. It may not be a huge improvement but it is a significant one.

So the specifications you cite give you, what, a 100 dB window of dynamic range? Very good, but well short of 24-bit dynamic range. A 100 dB window is just short of 17 bits of audio resolution. It may even be worse than that if your preamp level voltage range isn't optimized to match up with your NC30 room.

A non-truncated 24 bit source is ideal to draw from...mathematically, it is still an advantage but in terms of subjectivity, it will be a significant improvement.

I suppose you are going to tell me that vinyl is inferior to 16 bit as well..?

A non-truncated 24 bit source is ideal to draw from...mathematically, it is still an advantage but in terms of subjectivity, it will be a significant improvement.

It won't matter, if you are truncating noise bits. It's a moot point, if you are doing proper normalization and dither, anyway.

I suppose you are going to tell me that vinyl is inferior to 16 bit as well..?

It's off-topic, but 96 dB s/n is better than 75 dB s/n (which typically drops with every successive playback of that vinyl disc), last time I checked.

As has been stated numerous times already, no doubt you do hear a difference. It's just not the bit depth that is responsible for the observation.

Let's talk about here and now. 24 bit is where it is at. Equipment, DSPs, room design, etc are all so much improved relativeto 5 years ago...
DACs haven't. No consumer grade DAC has dynamic range >120db. 20bit resolution = 120db dynamic range. So why again do we need more? To waste space? A 24bit DAC does not mean it has 24bit resolution. It simply means that it can accept a 24bit signal. There are 32bit DAC with lower SNR than 24bit DACs. Should we replace high end high SNR 24bit DAC with lower SNR 32bit DAC just because 32 > 24 so it must be better?

It makes these new HD audio formats quite useful for the enthusiast who has built his system towards excellence. As khellandros66 says, the hobby is about pursuing excellence... If you you are content in your 16 bit world, so be it..But there is better..and it is already here. I never said 16 bit is enough.
16bit = 96db dynamic range
18bit = 108db dynamic range
20bit = 120db dynamic range
22bit = 132db dynamic range
24bit = 144db dynamic range

Most high end consumer DACs are in the 105-110db SNR. When playing back anything >18bit samples the least significant bits are below the noise floor. You can hope for magic if you want but unless your DAC has >120db SNR you there is no difference between a 20bit sample and a 24bit sample. There is no consumer gear at any price point that >120db SNR. Period.

Tell me about your system? Is it that your system lacks the resolution to even have 24 bit recordings make a difference?

So it has come down to "my toy is bigger than your toy". Sorry to ruin your fantasy world but I don't care if you spent $100K on your audio equipment you receiver can not convert 24bit samples with any more precision than it can 20bit samples. Even if it could there is no mic on the planet that has a low enough noise floor to capture said sound to begin with.

The truth is when DVD-A & SACD came out there was no evidence or theory that would even suggest 24/192 would offer anything better than 18/48 would. Want to know why DVD-A was designed to support these "amazing" resolutions? Consumers would never have paid more money for a spec that is so slightly improved over CD (16/44.1 vs 18/48). Anything higher is simply empty bits. You can pump it as high as you want. I have no doubt eventually they will be advocating 32bit / 192kHz as the only serious HD audio format but it will be just as much snake oil as 24/96 is.

Well, I am pretty much there.... as are many serious hobbiests who have such systems.

No your not. Your DAC does not have <110db SNR. It can "accept" a 24bit sample and due to internal noise it essentially loses at least 6 of those bits. You are ready for 18bit audio. 20bit audio would make the recording future proof. By the time you can properly resolve 20bit audio HD DVD and BD will have long since went the way of VHS.

Is it your contention that a 16 bit recording would sound identical to an identical 24 bit recording in such a well equipped room??

You keep asking the same wrong question.

The real question is will a 24bit and 20 bit sample sound identical? Yes they will. Will a 18bit and 24bit sample sound identical in anything less than a lab? Yes they will.

Just because 16 bits is not good enough it doesn't mean that 24bits is needed.
24bits has 256x as many values. Likely there is a value in between (18-20bit) that is more than enough and doesn't waste space just because 24/192 sounds cool and expensive.

If they had applied your logic to the video side of BD/HD DVD it have gone something like this:
1) 480p is not enough resolution (1/3MP)
2) The only acceptable resolution is 256X as much = aprox 72MP
3) So we need 6480P (11520 x 6480).
4) There are no 6480P sets available so most of that resolution is lost.
5) The only 2 choices are 480P and 6480P. 480P isn't good enough so we NEED 6480P.
6) All consumers will watch it on a 1080P set so most (80%) of that increased resolution is wasted ......
BUT 6480P looks better than 480P so it's worth it.

However cooler heads prevailed and guess what they went with 1080P which offers about 6x the resolution and looks very good compared to DVD without wasting space.

You can argue about 24 bits being too much and for delivery it actually is.

There is no argument about 16 bit being just to little. It can be very good for delivery but lacks the headroom for further processing.
Now we are at the point for audio that we have real headroom in delivery formats.

A very good system is at 18..19 bits resolution so we can argue that 20 bits would be enough. And it would be enough.
24 bits are a bonus, a free lunch, coming for free because in computerland everything is a multiple of 8 bits these days.

Why keep arguing with flawed logic like:

1) 480p is not enough resolution (1/3MP)
2) The only acceptable resolution is 256X as much = aprox 72MP
3) So we need 6480P (11520 x 6480).
4) There are no 6480P sets available so most of that resolution is lost.
5) The only 2 choices are 480P and 6480P. 480P isn't good enough so we NEED 6480P.
6) All consumers will watch it on a 1080P set so most (80%) of that increased resolution is wasted ......
BUT 6480P looks better than 480P so it's worth it.

You can argue about 24 bits being too much and for delivery it actually is.

There is no argument about 16 bit being just to little. It can be very good for delivery but lacks the headroom for further processing.
Now we are at the point for audio that we have real headroom in delivery formats.

A very good system is at 18..19 bits resolution so we can argue that 20 bits would be enough. And it would be enough.
24 bits are a bonus, a free lunch, coming for free because in computerland everything is a multiple of 8 bits these days.

Why keep arguing with flawed logic like:

1) 480p is not enough resolution (1/3MP)
2) The only acceptable resolution is 256X as much = aprox 72MP
3) So we need 6480P (11520 x 6480).
4) There are no 6480P sets available so most of that resolution is lost.
5) The only 2 choices are 480P and 6480P. 480P isn't good enough so we NEED 6480P.
6) All consumers will watch it on a 1080P set so most (80%) of that increased resolution is wasted ......
BUT 6480P looks better than 480P so it's worth it.


Ditto, Frank..

24 bit will give improvement today but as equipment advances, the software will be ahead of the curve in providing us with an even better soundtrack. I can't imagine wanting 16 bit when 24 bit is available. Few might be able to resolve it to an audible improvement in their systems TODAY, but if this format, like DVD, sticks around 10 years..the equipment will catch up and sound quality will continue to evolve. It is like the 1080P coming on the seen. Some of us have had 1080P displays for a couple years..now the video software has finally caught up with 1080P24 encodes. With audio, 24 bit is here before mass produced technology can fully take advanatage of it. That is a good thing and progress we should all commend as enthusiasts. Perhaps, 24 bit is all we'll ever need?

Only recently there where some 'shootouts' between 720p and 1080p projectors where folks were setting themselves up so that they were not able to tell the difference. In following discussion it turned out that some 720 owners where just scratching the upgrade itch.

In my setup the improvement from 720p to 1080p was very obvious. SDE alone was worth it. Not to mention the improved sharpness and better detail.

Now people are advocating that 16 bits should be 'good enough'.
The 24 bit hardware is there already and now that we have 24bit content becoming more and more available why complain/argue against about it?

Only recently there where some 'shootouts' between 720p and 1080p projectors where folks were setting themselves up so that they were not able to tell the difference. In following discussion it turned out that some 720 owners where just scratching the upgrade itch.

In my setup the improvement from 720p to 1080p was very obvious. SDE alone was worth it. Not to mention the improved sharpness and better detail.

Now people are advocating that 16 bits should be 'good enough'.
The 24 bit hardware is there already and now that we have 24bit content becoming more and more available why complain/argue against about it?

And some people will suggest the reason they are seeing a better picture with 1080p is not because of the higher resolution, but because the projector technology was better/newer with the 1080p display. I'm sure there are crap implementations for 1080p displays which look worse than good implementations of 720p displays.

That hasn't stopped the upgrade itch; I've seen a 1080p display I much prefer over my 720p display.... must..... resist.....

You can argue about 24 bits being too much and for delivery it actually is.

There is no argument about 16 bit being just to little. It can be very good for delivery but lacks the headroom for further processing.
Now we are at the point for audio that we have real headroom in delivery formats.

A very good system is at 18..19 bits resolution so we can argue that 20 bits would be enough. And it would be enough.
24 bits are a bonus, a free lunch, coming for free because in computerland everything is a multiple of 8 bits these days.

Computers don't process anything in 8bits anymore. Not since about the 386 days. Word size has moved from 8bit to 16bit to 32bit. If you load 8 bits into memory it's stored in a 32bit memory location. By the logic that 24bits is free 32bit is even more free. So why not move to 32bits?

Glad some people finally agree that 20bits is good enough. There are many people on this thread who either believe that 24 > 20 or they avoid the issue by always arguing 16 vs 24 like there are no options in between. TrueHD support any bitdepth. If every title was released with 18/48 or 20/48 lossless would you still be arguing for 24bit? You might not but a lot would and it would have no more information just more noise and take up more space.

Audio processing is moving to 32bit integer and 64bit floating point processing. Eventually all mastering will be done at either 32bit or 64bit then dithered to the needed bitdepth. If it has to be downsampled to 24 or 20 I would want 20. Wasting 4 extra bits (that only has noise encoded) just because marketing says it's better if foolish. Eventually the same people who "need" 24bit will be crying out for 32bit because they are "losing 99% of the dynamic range when downsampling to 24bit". The CE companies, studios, and companies like Dolby/DTS will be happy to sell everything all over again so you can get the extra 8 bits.

It's off-topic, but 96 dB s/n is better than 75 dB s/n (which typically drops with every successive playback of that vinyl disc), last time I checked.

As has been stated numerous times already, no doubt you do hear a difference. It's just not the bit depth that is responsible for the observation.

Its off topic a bit, but the noise in vinyl is not white noise. It has a spectrum. I saw a writup that claimed that in frequencies important for things like imaging, it actually equaled or exceeded CD noise (on a good pressing).

Regardless, the wcs s/n on vinyl hangs around 75 dB. It matters little if it is confined to a "spectrum" or not, and is even less relevant in real practice that includes wear'n'tear, embedded dust, and imperfect tracking.

16-bit digital is 96 dB across the entire design bandwidth, today, tomorrow, as many times as you desire to play it.

Now 18bit/48Khz I can agree with to a degree. I used to have harman/kardon HDCD player and the HDCDs I have are an improvement in imaging, but if a 24bit track only translates to 20bit then atleast thats better the 16bit, its the fact that we find someone claiming a lower bitrate is satisfactory thats frustrating.... And these are multimillion dollar movie productions, top notch everything should be used and available to whats the most import, the audience, and how it views and hears the actors and scenery are the importance.

~Bobby

24 translates to 20 in the DAC, but that is only one link in a chain with many links. Each link can further impact the nullification of further bits. As already demonstrated with theBland's system, the speakers and room end up reducing any 24-bit track to barely 17-bits of playback resolution. That's not even accounting for distortion and thermal compression at peak levels, yet.

You have an entire cottage industry of music enthusiasts who'll swear vinyl and reel-to-reel provide far more "musical information" than any 16/44.1k CD. Can they ALL be wrong ?


Sure. All the people who buy copper bracelets for arthritis can be wrong too. So can all the people who believe in lucky numbers.

But then again, much might depend on how technically they define 'musical information'. Vinylphiles are notoriously wooly when it comes to things like that.
The word 'information' has a specific meaning in the digital realm, however.


I don't even play the vinyl game, but I understand where they coming from simply from my own subjective experience base.


It's not unlikely that LPs will tend to sound different from CDs. They will tend to have been mastered rather differently at key stages. Then there's the inherent colorations of the medium to factor in. It's not unlikely that when two things sound different, some people will prefer one over the other. And when there are lots of ancillary factors -- the tactile and visual differences between LP and CD play, and quite possibly, for baby boomers, the nostalgia factor, to name a few -- who can say for sure what teh preferred sound will be?


I'd love to be able to participate in objective DBT audio tests but these have to be very difficult to set up properly and I am not willing to exert the effort to do so.

I doubt I have the capability to pull it off even if I were serious about trying it. I'd have to have access to the master tapes, the ability to play them back, the ability to transfer to 24/96 and/or 24/48 and/or 20/48 or DSD and/or 24/192 and/or 16/44.1 with equal amounts of the TLC in the transfer process, then get more than a few volunteers to subject themselves to listening tests; which volunteers having been trained to appreciate the vocal/instrument overtones, soundstaging, timbre, transient responses, "air/ambience" etc...

Indeed. Reason for doubt definitely exists, but it's often hard to set up the tests necessary to dispel it. So instead of taking the intellectually honest path and qualifying their claims, why do audiophiles tend to make such *definite* claims about what they heard? Like your claim about CD versus SACD, below.

As it is, I try to support live performances to establish the reference, then compare to "similar" recordings on consumer media (e.g. CD or SACD) to see how close the consumer media gets to the live performance. CD ain't even close, but SACD does get me marginally closer.

Then again, you could be wrong. And you seem to understand why. So, why are you so sure your belief about CD vs SACD accurately reflects reality?

Tell me about your system? Is it that your system lacks the resolution to even have 24 bit recordings make a difference?

And is it that yours *doesn't*? The system includes the room, you know. In the vast majority of situation, the 'system resolution' is determined by the ambient noise levels when the system is idling. Then there's the noise characteristics of the recording to consider -- if it's an analog source, or a 'live' recording, chances are that its required 'resolution' is already capably encompassed by 16 bits. Finally there's your own hearing to consider -- have you tested your 'just noticeable difference' abilities lately?

And even disregarding all this, almost certainly, you do not have access to the proper materials to determine whether the same release in 16 bit and 24 bit sounds better *due to the bitdepth*. You'd need a way to show that ONLY the bitdepth was changed.

PS.. I work for ToddAO.
Marc, now that you’ve “revealed” yourself, given the conversation we had …..what, about a year ago ?, I thought you would appreciate this link…………….

http://www.youtube.com/watch?v=qDjcWBUkVz8&mode=related&search=

P.S.
And, if I recall correctly, the CBS segment never even got into all his CAS nominations. :eek:
I think he did win an Emmy way back when for something.

How about that 24 bit resolution room? I doubt any room is balanced to within 10dB (if so, tell me how you detect it), and how you've managed to keep ambient noise levels below 30 dBA. So Pain 120-30 = 90 (-10 = 80 dB resolution), with about 1kW audio power. LOL

I am extending an open invitaion to anyone in the Los Angeles area to come and do a blind A/B test in a professional environment..

I have access to uncompressed, multichannel 24/48 masters of many major motion picture mixes, and have the proper equipment to do a 16 bit down rez, and then after the down rez, sync the two masters up in a sample accurate manner. play them in sync together, and compare away.

I will also make this a local group AVS gathering, and we can go through the mixing process and the like.. we did it last year, and had a grea time..

Any takers?

I am extending an open invitaion to anyone in the Los Angeles area to come and do a blind A/B test in a professional environment..

I have access to uncompressed, multichannel 24/48 masters of many major motion picture mixes, and have the proper equipment to do a 16 bit down rez, and then after the down rez, sync the two masters up in a sample accurate manner. play them in sync together, and compare away.

I will also make this a local group AVS gathering, and we can go through the mixing process and the like.. we did it last year, and had a grea time..

Any takers?

I would certainly love it! In orange county myself.

And, if I recall correctly, the CBS segment never even got into all his CAS nominations. :eek:
I think he did win an Emmy way back when for something.

I met an Emmy winner? That's what the IMDB entry seems to indicate... I'm towards LA but would certainly be interested in the blind A/B test.

I am extending an open invitaion to anyone in the Los Angeles area to come and do a blind A/B test in a professional environment..

I have access to uncompressed, multichannel 24/48 masters of many major motion picture mixes, and have the proper equipment to do a 16 bit down rez, and then after the down rez, sync the two masters up in a sample accurate manner. play them in sync together, and compare away.

I will also make this a local group AVS gathering, and we can go through the mixing process and the like.. we did it last year, and had a grea time..

Any takers?
Any idea of when and where? I would definitely be interested depending on the date. (Inland Empire dweller)