Dolby Elevates the Quality of Lossless Audio on Blu-ray - Page 3 - AVS Forum
Forum Jump: 
 
Thread Tools
Old 05-25-2012, 11:45 AM
AVS Special Member
 
RobertR's Avatar
 
Join Date: Dec 2000
Location: California
Posts: 6,244
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 235 Post(s)
Liked: 423
Quote:
Originally Posted by PeterTHX View Post

They did that with DD/DTS and that didn't stop people from proclaiming DTS was automatically better no matter what, even though the blind test showed even at 448kbps vs 1536kbps they couldn't readily tell the difference between the two.

The perception about DTS involved different mixes than the Dolby mixes, which is a different issue.
RobertR is offline  
Sponsored Links
Advertisement
 
Old 05-25-2012, 01:43 PM
AVS Special Member
 
Roger Dressler's Avatar
 
Join Date: Jul 2000
Location: Oregon
Posts: 8,967
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 820 Post(s)
Liked: 364
Quote:
Originally Posted by RobertR View Post

People don't have to prove a negative ("there isn't a difference"), nor does it make sense for them to try to do so. It's the advocate for the technology that needs to show people hear a difference under controlled (ie double blind) conditions.

Which audio companies have done double blind A/B tests to prove the merit of the products and technologies? Ayre, Meridian, Krell, Classe, Theta, Yamaha, Sony, Apple, Bose, Boulder? None that I have seen. Hasn't stopped anyone making claims of superiority for their products.

Quote:


People might be skeptical, but if the double blind methodology was thoroughly laid out and appeared to be sound, that would make for a strong case for the technology, much more powerful than testimonials or descriptions of it.

I guess that pretty much puts audio and automobile review magazines out of business. Shame no one told them.

Deadwood Atmos theater
AV7702 Atmos 7.4.4, SSP-800 PLIIx 7.4
Aerial Acoustics 7B/CC3B fronts, B&W CWM8180 surrounds, Tannoy Di6 DC heights, Hsu ULS-15 subs
Roger Dressler is online now  
Old 05-25-2012, 02:14 PM
AVS Special Member
 
RobertR's Avatar
 
Join Date: Dec 2000
Location: California
Posts: 6,244
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 235 Post(s)
Liked: 423
Quote:
Originally Posted by Roger Dressler View Post

Which audio companies have done double blind A/B tests to prove the merit of the products and technologies? Ayre, Meridian, Krell, Classe, Theta, Yamaha, Sony, Apple, Bose, Boulder? None that I have seen. Hasn't stopped anyone making claims of superiority for their products.

All that shows is that making claims without sound scientific support is commonplace in the audio industry. I was urging Dolby to be different in this regard. QSC Audio does make an ABX device, by the way.

Quote:


I guess that pretty much puts audio and automobile review magazines out of business.

It's no surprise that a company wouldn't want to conduct such a test, if they're not confident of a favorable outcome.
RobertR is offline  
Old 05-25-2012, 02:59 PM
AVS Special Member
 
AndreYew's Avatar
 
Join Date: Jan 2000
Location: Southern California
Posts: 1,670
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Roger Dressler View Post

Which audio companies have done double blind A/B tests to prove the merit of the products and technologies? Ayre, Meridian, Krell, Classe, Theta, Yamaha, Sony, Apple, Bose, Boulder? None that I have seen. Hasn't stopped anyone making claims of superiority for their products.

I guess that pretty much puts audio and automobile review magazines out of business. Shame no one told them.

Harman does so with their speakers. Also cars are so apparently different, I don't think they really need a controlled test to prove that different cars are different.

Dolby also did tests on jitter audibility in the past. It was a good thing, and they should have continued to do such things.
AndreYew is offline  
Old 05-25-2012, 10:58 PM
cwt
AVS Special Member
 
cwt's Avatar
 
Join Date: Jul 2006
Location: nsw australia
Posts: 1,229
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 32
Quote:
Originally Posted by pokekevin View Post

Dts is way better.... lol jk

Hands up everyone who will be drawn to an extent to these 96k discs rather than the ''ordinary'' 48k variety ? A chance to sample meridian apodising technology without spending $20000 ; wheres the drool icon Its the human condition to be curious

Anyone questioning these discs now before hearing them; you know your not allowed to heap praise on these discs afterwards ; credibility is at stake [j/k] Ime sure you will give them a fair hearing Look forward to buying some ; I get the impression a good % will be 7.1 not 5.1 noticing the satriani disc and dolbys atmos predilection
cwt is offline  
Old 05-26-2012, 12:44 AM
AVS Special Member
 
Roger Dressler's Avatar
 
Join Date: Jul 2000
Location: Oregon
Posts: 8,967
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 820 Post(s)
Liked: 364
Quote:
Originally Posted by AndreYew View Post

Harman does so with their speakers.

Their amps, too? And Surround processors? Rare exceptions aside, I think it is fair to say that double-blind A/B tests are not how products are sold to consumers.

Quote:


Also cars are so apparently different, I don't think they really need a controlled test to prove that different cars are different.

I thought Robert was saying that products should not need testimonials or descriptions to make the case to consumers. Yes, cars are so obviously different, yet they use testimonials and descriptions in their marketing. Was my point.

Quote:


Dolby also did tests on jitter audibility in the past. It was a good thing, and they should have continued to do such things.

I'm sure they conduct all sorts of tests on a ongoing basis. As do Meridian, and the studios who use these technologies.

Deadwood Atmos theater
AV7702 Atmos 7.4.4, SSP-800 PLIIx 7.4
Aerial Acoustics 7B/CC3B fronts, B&W CWM8180 surrounds, Tannoy Di6 DC heights, Hsu ULS-15 subs
Roger Dressler is online now  
Old 05-26-2012, 12:54 AM
AVS Special Member
 
Roger Dressler's Avatar
 
Join Date: Jul 2000
Location: Oregon
Posts: 8,967
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 820 Post(s)
Liked: 364
Quote:
Originally Posted by RobertR View Post

All that shows is that making claims without sound scientific support is commonplace in the audio industry.

That was my point.

Quote:


I was urging Dolby to be different in this regard.

What did they say?

Quote:


QSC Audio does make an ABX device, by the way.

I have not seen them publish double-blind A/B tests of their products.

Deadwood Atmos theater
AV7702 Atmos 7.4.4, SSP-800 PLIIx 7.4
Aerial Acoustics 7B/CC3B fronts, B&W CWM8180 surrounds, Tannoy Di6 DC heights, Hsu ULS-15 subs
Roger Dressler is online now  
Old 05-26-2012, 03:34 AM
AVS Special Member
 
RBFilms's Avatar
 
Join Date: May 2006
Location: New York / California
Posts: 1,267
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 19
Sound quality can be subjective in my opinion. Double blind testing would provide repeatable and accurate results with a trained ear or an educated ear. The listener needs to understand what to listen for. Once they do ... which is easy to teach them in 3 minutes or less ... they they can hear the difference every time. I have educated listeners in minutes on how to identify & appreciate note accurate bass, subtle details, proper imaging, a good soundstage, and most important of all ... the organic quality of vocals, instruments, and percussion. You can get amazingly consistent results from everyday folks with just a little bit of listener education.

With today's young society believing that ultra compressed lossy MP3 sounds better than a CD ... this education is necessary. Of course, most CD Discs sound terrible as well ... but they do sound better than an MP3 file.

I worked at DTS, and their commitment to quality was second to none. I have seen them perform double blind A/B tests numerous times using Golden Ears, Audiophiles, and Others. They always use Stax Headphones and reference quality Players, DAC's. and Line Amps for all of their testing.

I happen to use Stax Headphones for reference as well. There is no better headphone in my opinion for listening without the effect of room interaction with acoustics.

In fact, one of the best Golden Ears I know was responsible in part for DTS tweaking and improving their 96/24 CODEC before it was released.

Not sure what other companies do that....but I would agree with you that it would be very few.

Quote:
Originally Posted by Roger Dressler View Post

Which audio companies have done double blind A/B tests to prove the merit of the products and technologies? Ayre, Meridian, Krell, Classe, Theta, Yamaha, Sony, Apple, Bose, Boulder? None that I have seen. Hasn't stopped anyone making claims of superiority for their products.

I guess that pretty much puts audio and automobile review magazines out of business. Shame no one told them.


Richard J. Casey



Disney WOW - World of Wonder


Producers Guild of America, New Media Council
(BD Industry Insider)
RBFilms is offline  
Old 05-26-2012, 03:47 AM
AVS Special Member
 
RBFilms's Avatar
 
Join Date: May 2006
Location: New York / California
Posts: 1,267
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 19
The biggest issue I have with most companies is that they rely on science and not their ears to determine what supposedly sounds better. This is why I respect companies that still use their ears make improvements that actually sound better prior to releasing a product.


Quote:
Originally Posted by Roger Dressler View Post

That was my point.

What did they say?

I have not seen them publish double-blind A/B tests of their products.


Richard J. Casey



Disney WOW - World of Wonder


Producers Guild of America, New Media Council
(BD Industry Insider)
RBFilms is offline  
Old 05-26-2012, 04:49 AM
AVS Special Member
 
RBFilms's Avatar
 
Join Date: May 2006
Location: New York / California
Posts: 1,267
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 19
Great and very correct statement:

"I don't want my disc sounding "better" than the master it came from I want it sounding the same."

Thank you for that ... I agree with this statement.

-----

No offense to any of the pros or enthusiasts here, but up-sampling, dithering, and converting audio is one of the worst possible things anyone can do to sound and music recordings.

As a Producer/Director and an obsessed Audiophile/Videophile, I try to avoid converting audio in the digital domain. There are several reasons why and this was a theory that was also supported by DTS when I worked there.

Manipulation of audio in the digital domain creates audible distortion and artifacts due to dithering and other digital atrocities such as limited bit depth and sampling rates that wreak havoc when altering or modifying the signal.

In fact, when we master original audio we take everything OUT of the Digital Domain using a hand made Tube Based Digital to Analog Converter. We master in a Tube Analog chain where none of these digital atrocities can affect the signal, then convert back to digital using EMM DAC's to create a final digital master.

We also try to minimize any EQ & Signal Processing during production, mixing, and mastering. All of this is simply distortion of the audio signal and we always try to avoid all forms of signal manipulation and signal processing as much as realistically possible.

This is how we master all audio for products produced to our own Digital Reference Standard (DRS). This is exactly what we did with all of the demo content & narration on our Disney WOW - World of Wonder Home Theater Calibration disc.

Digital processing is generally harmful in many ways to analog sound. Jitter in particular is the worst enemy and biggest detriment to sound quality whenever it is recorded, mixed, processed, transferred, authored, encoded, or compressed.

Some engineers using "science" have determined that the human ear is incapable of "hearing" differences between real world organic sounds and a digitally processed signal. This is simply not true and all the test tones, analyzers, and equipment in the world cannot outperform the best test instrument ever made ... the human ear.

Several people, including several former engineers at Microsoft, have made claims that there is no audible difference between 48/16, 48/24, 96/24 or even 192/24. That is simply not true.

Also, it is simply not true that people cannot hear the difference between DD 448 lossy compression and 1.54. Anyone can hear the difference ... there is a seriously noticeable difference in fact ... even my non-audiophile friends can hear it without any education on what to listen for.

It is statements like these coming from industry professionals ... especially those involved in influencing technology or standards ... that has me most concerned. None of these statements are true so it is a worry that this thinking actually exists and that it may have a negative influence on technology ... as it already has. It is in fact, very sad that people promote this "good enough" technology and worse that they believe it is good ... or in this case of up-converting and dither a signal ... better.

In reality, it is easy to perform blind testing with any able bodied person that consistently demonstrates their ability to EASILY hear and see the difference in sound & picture quality between different technologies and formats. It is also possible to demonstrate the detrimental effect of dithering & converting signal.

448 verses 1.54 ... lossy versess lossless .... 48/16 verses 48/24 ... 48/24 verses 96/24 ... 96/24 verses 192/24 ... and most important of all ... digital verses analog.

Every one of these comparisons can be proven by demonstration to anyone ... especially if you take 60 seconds to explain what to listen for when demonstrating audio or what to look for in the case of video.

I know I will be flamed for my statements, but I would not make them if I had proven all of this repeatedly to myself and my peers.

My point as it relates to the original post? I do not support as a professional or a consumer the up-conversion of audio or any processing that claims to sound "better than the master" .... so unless you are removing pops and ticks without affecting the signal quality ... or you have master that is afflicted with noise and garbage ... then this is simply a false claim as far as I am concerned.

Richard J. Casey



Disney WOW - World of Wonder


Producers Guild of America, New Media Council
(BD Industry Insider)
RBFilms is offline  
Old 05-26-2012, 06:09 AM
AVS Special Member
 
Roger Dressler's Avatar
 
Join Date: Jul 2000
Location: Oregon
Posts: 8,967
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 820 Post(s)
Liked: 364
Quote:
Originally Posted by RBFilms View Post

No offense to any of the pros or enthusiasts here, but up-sampling, dithering, and converting audio is one of the worst possible things anyone can do to sound and music recordings.

Manipulation of audio in the digital domain creates audible distortion and artifacts due to dithering and other digital atrocities such as limited bit depth and sampling rates that wreak havoc when altering or modifying the signal.

And another mechanism that wreaks havoc on audio is brickwall filters, as used for 44.1 or 48 kHz audio. You appear to dispute claims that sample rates make no audible difference, so it seems fair to include that as one of the enumerated atrocities.

A lot of DACs upsample (oversample) and a lot of disc players and AV processors upsample. It's a fact of life, and not all of it is done to optimal standards. By applying a highly evolved upsampling algorithm in an offline computer without the processing limitations of consumer DSP products, the foibles of the consumer device can largely be bypassed.

Quote:


I do not support as a professional or a consumer the up-conversion of audio or any processing that claims to sound "better than the master" .... so unless you are removing pops and ticks without affecting the signal quality ... or you have master that is afflicted with noise and garbage ... then this is simply a false claim as far as I am concerned.

The goal is to more closely approach the sound quality of the audio in the mics or the console, canceling the atrocities of "1x" sample rates as used downstream.

Deadwood Atmos theater
AV7702 Atmos 7.4.4, SSP-800 PLIIx 7.4
Aerial Acoustics 7B/CC3B fronts, B&W CWM8180 surrounds, Tannoy Di6 DC heights, Hsu ULS-15 subs
Roger Dressler is online now  
Old 05-26-2012, 09:28 AM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 19,886
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 1429 Post(s)
Liked: 1027
Quote:
Originally Posted by RBFilms View Post

With today's young society believing that ultra compressed lossy MP3 sounds better than a CD ...

Just an FYI: that myth was recently debunked.

AVS thread about the topic.

More here, with recently updated data.

Sanjay
sdurani is offline  
Old 05-27-2012, 06:41 PM
AVS Special Member
 
RBFilms's Avatar
 
Join Date: May 2006
Location: New York / California
Posts: 1,267
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 19
Up-Converting from a lower sample rate to a higher sample rate is not good ... correct.

Recording analog with modified equipment does not have the "brick wall" filter issues that you mentioned. Recording digital at 192/24 using an excellent Analog to Digital Converter also eliminates this concern.

When recording or transferring analog to 96/24 and especially 192/24 ... the analog to digital converters make a huge difference. There are only a handful of Analog to Digital Converters I would find acceptable.

EMM makes the best sounding Analog to Digital Converters I have heard. This is especially true at 192/24 where lesser Digital to Analog converters add artifacts due to their inability to properly handle the very high sampling rates.

Yes, some consumer gear does convert ... and most do it badly. However, as a producer, I would not do this. Why destroy signal integrity at the source when I can give the end user the best quality possible. Maybe three (3) years from now some manufacturers will build consumer equipment that plays back the signal better than it odes today. One that signal is touched in any bad way by the Producer, it is corrupt forever.

However, it is jitter that really does the most harm. This is why we work diligently to eliminate the introduction of jitter all the way through the digital chain that is within our control.

Bottom Line - No processing is best but that is not entirely possible when it comes to delivering the signal in a consumer format. We instead try to avoid all processing up to the point of authoring a master for authoring.

Alos, you cannot "more closely approach the sound quality of the audio in the mics or the console" with unnecessary processing. The best way to achieve this goal is to keep your hands off the signal and pass it through as unscathed as humanely possible.

Listen to any Mapleshade / Wildchild CD for a perfect sonic example of this theory.

http://www.mapleshaderecords.com/


Up-Converting will degrade, not improve the sound quality of the original native source. After 38 years in this industry, I have proven this more times than I can remember to pros, enthusiasts, and consumers with simple demonstrations. Unnecessary processing does not improve sound quality ... this is not a theory, it is a simple fact.


Quote:
Originally Posted by Roger Dressler View Post

And another mechanism that wreaks havoc on audio is brickwall filters, as used for 44.1 or 48 kHz audio. You appear to dispute claims that sample rates make no audible difference, so it seems fair to include that as one of the enumerated atrocities.

A lot of DACs upsample (oversample) and a lot of disc players and AV processors upsample. It's a fact of life, and not all of it is done to optimal standards. By applying a highly evolved upsampling algorithm in an offline computer without the processing limitations of consumer DSP products, the foibles of the consumer device can largely be bypassed.

The goal is to more closely approach the sound quality of the audio in the mics or the console, canceling the atrocities of "1x" sample rates as used downstream.


Richard J. Casey



Disney WOW - World of Wonder


Producers Guild of America, New Media Council
(BD Industry Insider)
RBFilms is offline  
Old 05-27-2012, 06:45 PM
AVS Special Member
 
RBFilms's Avatar
 
Join Date: May 2006
Location: New York / California
Posts: 1,267
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 19
I am not sure that thread debunks the theory. When was the last time you spoke to a Best Buy Blue SHirt or a kid on the street listing to an I-Pod?

I am a technology evangelist at heart, so I am alway stalking to folks about the benefits of higher quality picture & sound ... including a better emotional connection to the content. Most consumers I talk to are VERY unaware .. they are even surprised to hear that their systems require calibration for optimal performance. There is a serious lack of education for consumers regarding technology ... amongst other things. I assume it is all part of the "dumbing down of america" .... and for sure art of the "good enough" theory embraced & promoted by big name corporations ... which I will not name ...


Quote:
Originally Posted by sdurani View Post

Just an FYI: that myth was recently debunked.

AVS thread about the topic.

More here, with recently updated data.


Richard J. Casey



Disney WOW - World of Wonder


Producers Guild of America, New Media Council
(BD Industry Insider)
RBFilms is offline  
Old 05-27-2012, 07:42 PM
AVS Special Member
 
Roger Dressler's Avatar
 
Join Date: Jul 2000
Location: Oregon
Posts: 8,967
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 820 Post(s)
Liked: 364
Quote:
Originally Posted by RBFilms View Post

Up-Converting from a lower sample rate to a higher sample rate is not good ... correct.

Correct? Who are you agreeing with?

Quote:


Recording analog with modified equipment does not have the "brick wall" filter issues that you mentioned.

Ok. And that's how all movies were/are made?

Quote:


Recording digital at 192/24 using an excellent Analog to Digital Converter also eliminates this concern.

Nice. Except it is not relevant since no one makes movies that way.

Quote:


When recording or transferring analog to 96/24 and especially 192/24 ... the analog to digital converters make a huge difference. There are only a handful of Analog to Digital Converters I would find acceptable.

That is not relevant in a discussion of movie production, is it, since they use 48 kHz.

Quote:


EMM makes the best sounding Analog to Digital Converters I have heard. This is especially true at 192/24 where lesser Digital to Analog converters add artifacts due to their inability to properly handle the very high sampling rates.

Irrelevant. See above.

Quote:


Yes, some consumer gear does convert ... and most do it badly. However, as a producer, I would not do this. Why destroy signal integrity at the source when I can give the end user the best quality possible.

By all means. But other than you, no one is talking about your content. This thread is about Blu-ray movie sound.

Quote:


Maybe three (3) years from now some manufacturers will build consumer equipment that plays back the signal better than it does today. Once that signal is touched in any bad way by the Producer, it is corrupt forever.

Happily, I don't see any advocating that.

Quote:


Bottom Line - No processing is best but that is not entirely possible when it comes to delivering the signal in a consumer format. We instead try to avoid all processing up to the point of authoring a master for authoring.

That's nice to know. But irrelevant in this thread as already mentioned.

Quote:


Also, you cannot "more closely approach the sound quality of the audio in the mics or the console" with unnecessary processing.

Probably not. Happily, no one is advocating unnecessary processing.

Quote:


The best way to achieve this goal is to keep your hands off the signal and pass it through as unscathed as humanely possible.

That would best be achieved by using sample rates not in use by movies.

Quote:


Up-Converting will degrade, not improve the sound quality of the original native source.

That assumes the original native source was digital. Maybe it was analog, sampled at 48 kHz with a less than ideal converter.

Quote:


After 38 years in this industry, I have proven this more times than I can remember to pros, enthusiasts, and consumers with simple demonstrations. Unnecessary processing does not improve sound quality ... this is not a theory, it is a simple fact.

That's the definition of unnecessary.

Not all upsamling is the same, to put it mildly. You have not proved it with the upsampling at hand.

Deadwood Atmos theater
AV7702 Atmos 7.4.4, SSP-800 PLIIx 7.4
Aerial Acoustics 7B/CC3B fronts, B&W CWM8180 surrounds, Tannoy Di6 DC heights, Hsu ULS-15 subs
Roger Dressler is online now  
Old 05-28-2012, 09:05 AM
AVS Addicted Member
 
sdurani's Avatar
 
Join Date: Oct 2001
Location: Monterey Park, CA
Posts: 19,886
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 1429 Post(s)
Liked: 1027
Quote:
Originally Posted by RBFilms View Post

I am not sure that thread debunks the theory. When was the last time you spoke to a Best Buy Blue SHirt or a kid on the street listing to an I-Pod?

The thread discusses an AES paper by Sean Olive (slidepack linked earlier), which used double-blind testing of teenagers rather than talking to a Best Buy blue-shirt or a kid on the street listening to to an iPod.

Your position (youngsters prefer ultra compressed lossy MP3 to CD sound) gained popularity when Jonathan Berger (professor of music at Stanford University) was quoted in the New York Times based on an "informal study" he had done with his students over the course of seven years.

Rather than take this 'MP3 effect' at face value, someone decided to test the claim. Hence the double-blind study and AES paper, which was nicely summed up by Geoffrey Morrison earlier this month in Sound & Vision magazine.

Sanjay
sdurani is offline  
Old 05-28-2012, 09:10 AM
AVS Special Member
 
RBFilms's Avatar
 
Join Date: May 2006
Location: New York / California
Posts: 1,267
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 19
Well, lets hope that thinking spreads....it is not evident in my neck of the woods.

I would like to see better quality picture & sound promoted to consumers. There are certain companies involved in the industry that pre-determine the level of quality that they believe is "good enough" for the general consumer.

They even go as far as to say they cannot hear a difference ... which is absolutely not true.

So lets hope that knowledge of the fact that there is better ... spreads like wildfire.

Quote:
Originally Posted by sdurani View Post

The thread discusses an AES paper by Sean Olive (slidepack linked earlier), which used double-blind testing of teenagers rather than talking to a Best Buy blue-shirt or a kid on the street listening to to an iPod.

Your position (youngsters prefer ultra compressed lossy MP3 to CD sound) gained popularity when Jonathan Berger (professor of music at Stanford University) was quoted in the New York Times based on an "informal study" he had done with his students over the course of seven years.

Rather than take this 'MP3 effect' at face value, someone decided to test the claim. Hence the double-blind study and AES paper, which was nicely summed up by Geoffrey Morrison earlier this month in Sound & Vision magazine.


Richard J. Casey



Disney WOW - World of Wonder


Producers Guild of America, New Media Council
(BD Industry Insider)
RBFilms is offline  
Old 05-28-2012, 10:07 AM
AVS Special Member
 
RBFilms's Avatar
 
Join Date: May 2006
Location: New York / California
Posts: 1,267
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 19
Sorry for any confusion in my reply.

I will keep this brief, simple, and to the point so it is easy to understand my POV:

I am responding primarily to this:

Dolby TrueHD with advanced 96k upsampling integrated into Dolby Media Producer delivers enhanced studio-quality surround sound

These are the real facts:

The OP talks about Producers using Dolby 96/24 Up-Sampling to improve sound using a 48/24 source.

I am a Producer/Director of Movies, Documentaries, and Music Concerts so I am specifically commenting on the Dolby PR in this thread. None of what I am saying is "irrelevant" by any means.

Up-Sampling is a detriment to Audio Quality.

Not ALL Movies, Documentaries, and Music Concerts are recorded Analog. There are plenty that are recorded using Digital Capture. Most are produced, mixed and output at 48/24 using who knows what converters.

96/24 and 192/24 are very relevant to production of programming since the Blu-ray Format supports these audio formats but only some movies are utilizing this technology.

96/24 us being used by some Producers today and the entire OP was about the benefit of higher sampling rates.


Based on what I stated above, my conversation about Analog to Digital Converters seems to be very relevant. Why aren't more Producers transferring their analog recordings to 96/24 or 192/24?

If "even" Dolby states 96/24 is better, something most engineers understand ... (and others supported way before Dolby ever did) ... and Blu-ray Disc supports 96/24 and 192/24 ... then why are my comments about Analog to Digital Converters "irrelevant."

I am also stating how I believe Movie Soundtracks are being mishandled now and how I believe they could be handled to achieve better sound quality. This seems to be the topic of discussion here.

Also, why Up-Convert when the Movie Makers can transfer their Analog recording using High Quality Analog to Digital Converters like EMM to begin with and avoid the Unnecessary Processing that Dolby seems to promote on a regular basis. Why isn't anyone promoting that approach which is a far superior technique and the right thing to do to begin with.

By Dolby Up-Sampling prior to the mastering / authoring process, they have destroyed the integrity of the original native source. The correct way to handle this is to give the consumer the best quality soundtrack as native and transparent to the original source as possible. Once Dolby touches that signal by up-sampling it, the quality has been impacted forever.

My point is that 3 years from now there may be better up-sampling technology in a consumer gear than anything Dolby is promoting now for professional use. By giving the consumer the native format, they are better served by future improvements in technology instead of locked in to old and outdated up-sampling methods, techniques, and technology.

Dolby is promoting unnecessary processing by up-sampling 48/24 soundtracks to 96/24. I guess that is what I should have stated. I was trying to be nice.

I am happy to prove my point. Send me a Dolby Up-Sampling Box I will be happy to perform a demonstration using analog content transferred to 48/24, 96/24, and 192/24 and compare it to 48/24 up-sampled to 96/24 using the Dolby box.

Case in point. The original Lake Systems box that dealt with Head Related Transfer Functions (HRTF) featuring Analog inputs, which I have, sounds far superior to the Dolby Headphone Surround Box that replaced it once Dolby purchased that technology. The limit of 48/24 and the digital processing in the box are the prime culprits for dumbing down the potential quality of the final output. Not that adding distortion ever sounds great, but Lake did it better.

Major companies like Dolby that have influence over the industry to improve sound quality should support and promote improvements that utilize the ability of current consumer formats to achieve a better entertainment experience. I understand that is what Dolby may think they are doing by promoting Up-Sampling of 48/24 sources, but I am sorry to say I do not agree nor do I see this approach as a viable solution. Promoting the use of 96/24 and 192/24 transfers from Analog sources using quality Analog to Digital Converters and maintaining the integrity of that signal throughout the chain with minimal processing is a far better solution in my opinion.

Last but not least, I do not see nor do I recognize you as having any authority to dismiss my comments as irrelevant when the entire thread is about improving movie soundtracks and mentions 96/24 to begin with.


Quote:
Originally Posted by Roger Dressler View Post

Correct? Who are you agreeing with?

Ok. And that's how all movies were/are made?

Nice. Except it is not relevant since no one makes movies that way.

That is not relevant in a discussion of movie production, is it, since they use 48 kHz.

Irrelevant. See above.

By all means. But other than you, no one is talking about your content. This thread is about Blu-ray movie sound.

Happily, I don't see any advocating that.

That's nice to know. But irrelevant in this thread as already mentioned.

Probably not. Happily, no one is advocating unnecessary processing.

That would best be achieved by using sample rates not in use by movies.

That assumes the original native source was digital. Maybe it was analog, sampled at 48 kHz with a less than ideal converter.

That's the definition of unnecessary.

Not all upsamling is the same, to put it mildly. You have not proved it with the upsampling at hand.


Richard J. Casey



Disney WOW - World of Wonder


Producers Guild of America, New Media Council
(BD Industry Insider)
RBFilms is offline  
Old 05-28-2012, 01:14 PM
Advanced Member
 
welwynnick's Avatar
 
Join Date: Jul 2005
Location: Welwyn, Herts, UK
Posts: 976
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 13
I definitely agree about the need to minimise ALL digital audio signal processing. I don't think that audio is "safe" just because its in the digital domain.

A few years ago I was privileged to work with an exceptional electronics engineer, the kind of guy who could seemingly design anything from the ground-up, without using cookbook design (ie: an integrated solution and application notes). A rare man; and he told me an interesting story. He designed a digital audio mixing desk for a recording facility, and he tried hard to preserve audio quality. There were necessarily several processing stages, and testing showed that too many stages degraded digital audio quality. He investigated this, and found an interesting rule:

Every time you apply a digital process to data, on average, you lose half a bit of data.

Therefore, the data pathway has to be much wider than the minimum audio requirement, so he had to specify apparently over-the-top DSP precision to everything he did. Lots of important and experienced people didn't believe what he said, so he sat them all down and reasoned his way through everything he had done, until after a very long time, everyone was forced to agree with him. So if you want 16-bits worth of dynamic range, and you want to apply 16 processes to the digital audio (which isn't that much) then you need at least 24 bits just to preserve those 16 bits.

And what is a "digital process"? That's simple - its just any operation that causes the number to change. So that doesn't include any recording or switching functions, but it does include level change, filtering, equalization, conversion, etc. Any process that takes one number and converts it to another number. When you think about it, it stands to reason: when there's a calculation, the LSB will be either rounded or truncated, so there's an equal chance that it will be right or wrong. On average, it will be wrong half the time, so you lose that much precision or depth of modulation with every process. What's interesting is that its a linear loss, and there are no root-sum-squares in there.

So the short of it is that you need a much wider data pathway than the desired result. People like Meridian, who really understand digital audio, use 48-bit DSP in their recent processors, which sounds like overkill, but may be a necessity.

Nick
welwynnick is offline  
Old 05-28-2012, 01:52 PM
AVS Special Member
 
Roger Dressler's Avatar
 
Join Date: Jul 2000
Location: Oregon
Posts: 8,967
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 820 Post(s)
Liked: 364
Quote:
Originally Posted by welwynnick View Post

He investigated this, and found an interesting rule:

Every time you apply a digital process to data, on average, you lose half a bit of data.

Therefore, the data pathway has to be much wider than the minimum audio requirement, so he had to specify apparently over-the-top DSP precision to everything he did. Lots of important and experienced people didn't believe what he said, so he sat them all down and reasoned his way through everything he had done, until after a very long time, everyone was forced to agree with him. So if you want 16-bits worth of dynamic range, and you want to apply 16 processes to the digital audio (which isn't that much) then you need at least 24 bits just to preserve those 16 bits.

When you think about it, it stands to reason: when there's a calculation, the LSB will be either rounded or truncated, so there's an equal chance that it will be right or wrong. On average, it will be wrong half the time, so you lose that much precision or depth of modulation with every process. What's interesting is that its a linear loss, and there are no root-sum-squares in there.

In a proper DSP calculation, the results are dithered. When two dithered processes are cascaded, there is that half bit of loss in dynamic range, which results from two noise sources adding together. The sum is 3 dB higher noise.

If you add another 2 processes, it goes up another 3 dB. Add 4 more, 3 dB again, 8 more, 3 dB again. That's 16 total cascaded processes, with a total noise increase of 12 dB, or 2 bits, not 8 bits.

Quote:


So the short of it is that you need a much wider data pathway than the desired result. People like Meridian, who really understand digital audio, use 48-bit DSP in their recent processors, which sounds like overkill, but may be a necessity.

Yes, their expertise and perfectionist approach is embodied in their 96kHz upsampling apodizing technology.

Deadwood Atmos theater
AV7702 Atmos 7.4.4, SSP-800 PLIIx 7.4
Aerial Acoustics 7B/CC3B fronts, B&W CWM8180 surrounds, Tannoy Di6 DC heights, Hsu ULS-15 subs
Roger Dressler is online now  
Old 05-28-2012, 03:21 PM
AVS Special Member
 
Roger Dressler's Avatar
 
Join Date: Jul 2000
Location: Oregon
Posts: 8,967
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 820 Post(s)
Liked: 364
Quote:
Originally Posted by RBFilms View Post

These are the real facts:

Up-Sampling is a detriment to Audio Quality.

Perhaps it is, but there is no evidence to prove ALL upsampling is detrimental. That, sir, is a real fact.

Quote:


Not ALL Movies, Documentaries, and Music Concerts are recorded Analog. There are plenty that are recorded using Digital Capture. Most are produced, mixed and output at 48/24 using who knows what converters.

The point is that one must protect against aliasing when creating 48 kHz digital audio. That is a fact.

Quote:


96/24 and 192/24 are very relevant to production of programming since the Blu-ray Format supports these audio formats but only some movies are utilizing this technology.

Exactly. That's why we are talking about movies that are not recorded in 96 or 192 kHz.

Quote:


96/24 is being used by some Producers today and the entire OP was about the benefit of higher sampling rates.

Not higher sample rates in general, but applying higher sample rates to 48 kHz movies in particular.

Quote:


Based on what I stated above, my conversation about Analog to Digital Converters seems to be very relevant. Why aren’t more Producers transferring their “analog” recordings to 96/24 or 192/24?

Why not ask them? Probable answer: time and cost. They probably create the home video transfers starting with the same digital masters made for the D-Cinema release.

Quote:


If "even" Dolby states 96/24 is better, something most engineers understand ... (and others supported way before Dolby ever did) ... and Blu-ray Disc supports 96/24 and 192/24 ... then why are my comments about Analog to Digital Converters "irrelevant."

It's O/T. This thread is about upsampling 48 kHz movies, not about general techniques and tools for audio recording. I'm sure there are other threads for that discussion.

Quote:


I am also stating how I believe “Movie Soundtracks” are being mishandled now and how I believe they could be handled to achieve better sound quality. This seems to be the topic of discussion here.

Exactly!

Quote:


Also, why Up-Convert when the Movie Makers can transfer their “Analog” recording using High Quality Analog to Digital Converters like EMM to begin with and avoid the “Unnecessary Processing” that Dolby seems to promote on a regular basis. Why isn't anyone promoting that approach which is a far superior technique and the right thing to do to begin with.

There are no analog masters these days. Analog sources, yes. But the mixing is done in DAWs and consoles at 48 kHz. The master is digital.

Quote:


By Dolby “Up-Sampling” prior to the mastering / authoring process, they have destroyed the integrity of the original native source.

a) The upsampling is done only at the very end, in the Media Producer prior to TrueHD encoding, to create the disc file. The source master in the salt mine is not altered. b) Speaking of integrity, may I point out that you have shown no evidence whatsoever on which to claim anything was destroyed, yet you persist in such proclamations.

Quote:


Once Dolby touches that signal by up-sampling it, the quality has been impacted forever.

Yes, for the better in this case, IMHO.

Quote:


My point is that 3 years from now there may be better up-sampling technology in a consumer gear than anything Dolby is promoting now for professional use.

That would be in direct contradiction to your own "real fact": >>Up-Sampling is a detriment to Audio Quality.<< So, you actually concede that proper upsampling is not so bad after all?

Quote:


The correct way to handle this is to give the consumer the best quality soundtrack as native and transparent to the original source as possible.

By giving the consumer the native format, they are better served by future improvements in technology instead of locked in to old and outdated up-sampling methods, techniques, and technology.

So we consumers should not have purchased VHS tapes as they failed to deliver the best quality? Or DVDs? Or AppleTV/Vudu/Amazon movie downloads? Or watch HBO? You seriously think Blu-ray is the end of the line? Seems we are perpetually locked in to old and outdated technology, at least until the next one comes along.

Quote:


Dolby is promoting unnecessary processing by up-sampling 48/24 soundtracks to 96/24. I guess that is what I should have stated. I was trying to be nice.

You are trying to be clairvoyant. See next point.

Quote:


I am happy to “prove” my point. Send me a Dolby Up-Sampling Box … I will be happy to perform a demonstration using analog content transferred to 48/24, 96/24, and 192/24 and compare it to 48/24 up-sampled to 96/24 using the Dolby box.

You know the outcome but have not actually done the test? I'm sure that with your public disdain for Dolby and its technologies, they would be quite eager to set you up with a test unit. Give Craig a call.

Quote:


Case in point. The original Lake Systems box that dealt with Head Related Transfer Functions (HRTF) featuring Analog inputs, which I have, sounds far superior to the Dolby Headphone Surround Box that replaced it once Dolby purchased that technology. The limit of 48/24 and the digital processing in the box are the prime culprits for dumbing down the potential quality of the final output. Not that adding distortion ever sounds great, but Lake did it better.

How does this bear any relevance to this thread?

By the way, since you also seem misinformed on this Dolby topic, I will remain O/T to address it. Dolby never made or sold a Dolby Headphone box to replace the Lake unit -- which was a studio processor. Lake, however, designed a new, alternative embodiment of their headphone technology, working closely with Sanyo Semiconductor, that could fit in a small, low cost chip. That chip is used in a number of standalone headphone processors such as from Pioneer. However, the full DSP-based solution in the Tag McLaren AV32RDP and Denon processors always sounded better to me.

Lastly, Dolby (who now owns Lake) have developed successor Dolby Headphone technology, expanded to handle 7.1 channels.

Quote:


Major companies like Dolby that have influence over the industry to improve sound quality should support and promote improvements that utilize the ability of current consumer formats to achieve a better entertainment experience. I understand that is what Dolby may “think” they are doing by promoting “Up-Sampling” of 48/24 sources, but I am sorry to say … I do not agree nor do I see this approach as a viable solution. Promoting the use of 96/24 and 192/24 transfers from Analog sources using quality Analog to Digital Converters and maintaining the integrity of that signal throughout the chain with minimal processing is a far better solution in my opinion.

Your opinions are duly noted.

Quote:


Last but not least, I do not see nor do I recognize you as having any authority to “dismiss” my comments as “irrelevant” when the entire thread is about improving movie soundtracks and mentions 96/24 to begin with.

I was merely stating my opinions on your comments. There is no need to recognize me whatsoever.

Deadwood Atmos theater
AV7702 Atmos 7.4.4, SSP-800 PLIIx 7.4
Aerial Acoustics 7B/CC3B fronts, B&W CWM8180 surrounds, Tannoy Di6 DC heights, Hsu ULS-15 subs
Roger Dressler is online now  
Old 05-28-2012, 05:01 PM
AVS Special Member
 
Fanboyz's Avatar
 
Join Date: Dec 2007
Location: Minnesota
Posts: 2,431
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 27 Post(s)
Liked: 19
All upsampling may not be detrimental, but it will undoubtedly make the resultant soundtrack measurably different from the master,

This would betray the concept of lossless audio.
Not probably anymore than a 24bit track being encoded as 16bit- but another technical wrinkle disquiet all our minds.

The proper setting for sharpness is always0.
Also my Oppo BDP-103D is region free.
That makes me awesome.
Fanboyz is offline  
Old 05-28-2012, 05:31 PM
AVS Special Member
 
AndreYew's Avatar
 
Join Date: Jan 2000
Location: Southern California
Posts: 1,670
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by welwynnick View Post

Every time you apply a digital process to data, on average, you lose half a bit of data.

Roger already addressed this with addition, but this is well-known to computer programmers. Every time you add two numbers, you need one more bit of precision, because you could potentially double the maximum value of the numbers being added. And every time you multiply, you could need twice as many bits to hold the result.

The Motorola DSP56001 series of DSP chips, available back in the late 80s, had a 56-bit fixed-point accumulator for this reason. State of the art audio DSP today uses double-precision floating point with 52 bits of mantissa and 11 bits of exponent.

But your point is taken, because there are some very complicated DSP processes (like high quality EQ or filtering, especially at high sample rates) that require many, many more bits (like hundreds or thousands) to perform accurately because there are so many computation steps, and the people who implement them do have to do the proper math in order to get those DSP routines working.

And before the results are truncated down to 24 bits or whatever the delivery media will hold, dither is added. Dither, contrary to what has been said before, is not detrimental to audio quality. Lack of dither can be severely detrimental.

As for upsampling's accuracy, using a very high quality upsampler (Weiss Saracon), people have obtained bitwise null results going from 96 kHz to 44.1 kHz back up to 96 kHz, using linear phase filters. The files did not null when using minimum phase filters (like the ones Dolby's trying to push). Don't like accuracy? Use minimum phase, no-ringing filters.

A null test means subtracting one file from another and seeing what remains. The test above had zero difference in the files after one file had been through a downsampling and an upsampling. Here is a link to the claims by Bob Katz, someone with not inconsiderable audiophile cred:

http://bach.pgm.com/pipermail/proaud...ay/015334.html

Quote:


Well, I performed a "single variable" test on minimum phase versus
linear phase low pass filters and can now state to my satisfaction that
the minimum phase filter alters the original sound in a very interesting
way, but it is clearly an alteration or aberration, and not an authentic
representation of the source. I think I've also eliminated pre-echo as
any influence, but you tell me based on my report below.

I started with two excellent 2496 master recordings of different
acoustic music that I had made. I then downsampled these two recordings
to 3244 using R8Brain Pro in two modes, one linear phase, the other
minimum phase.

I then upsampled the lp vs. mp results back to 3296 using Saracon as it
is the most transparent sample rate converter, so as to provide the
least sonic loss.

Then I did some null tests at 96 kHz on the results. The mp (as would be
expected) produced a decreasing null in the high frequency range as the
frequency increased. It was also surprising to hear a less than perfect
null between the linear phase version and the original, and I have to
put this at the feet of R8Brain as I have previously gotten 100% perfect
nulls using Saracon for a round trip. The world's best sample rate
converter has to be used if you want to make a judgment of filter
quality, guys! For audible transparency you need extraordinary quality
of DSP.

Next I did some listening, comparing the 96 original versus the
96-44mp-96 versus the 96-44LP-96. Yes, I applied 24 bit dither on the
audition of all of these, even where it was not necessary (when
listening to the 2496 original no dither need be applied going to the
DAC, but at least this way there is a constant amount of 24 bit noise
being applied, not that this is audible, but what the hey).

The immediate effect: the minimum phase low pass filter produces a very
pleasant and interesting added "holographic" depth effect. But since
this depth effect is clearly not in the original or in the LP round trip
either, it can be concluded that the minimum phase filter produces a
very strong, pleasant but inaccurate coloration. I definitely have to
file this effect in my "I could use this" folder! 41 years in this
business and I'm still learning something new.

Also interesting is that I can hear a barely audible loss, less "open"
sound listening to the LP round trip compared to the original. Again I
attribute this to R8Brain's low pass not being as good as Saracon's,
which is so transparent that I have never been able to hear a loss going
96-44-96 before. The best 22.05 kHz filter is audibly transparent, but
anything less is not.

That's the news. Interesting, eh? So, minimum phase DACs go out the
door. Laws of physics remain immutable.

And yes, you have to be very careful when performing null tests, lining both files up exactly, which can sometimes be impossible if the DSP introduces sub-sample delay: http://bach.pgm.com/pipermail/proaud...ay/015347.html
AndreYew is offline  
Old 05-28-2012, 07:02 PM
AVS Special Member
 
Roger Dressler's Avatar
 
Join Date: Jul 2000
Location: Oregon
Posts: 8,967
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 820 Post(s)
Liked: 364
Quote:
Originally Posted by AndreYew View Post

As for upsampling's accuracy, using a very high quality upsampler (Weiss Saracon), people have obtained bitwise null results going from 96 kHz to 44.1 kHz back up to 96 kHz, using linear phase filters. The files did not null when using minimum phase filters (like the ones Dolby's trying to push). Don't like accuracy? Use minimum phase, no-ringing filters.

Well, if the files null, then the artifacts of the 44.1 kHz anti-alias filter remain present. By definition, the apodized version should not null, or else the defect could not have been addressed.

As to Mr. Katz experiments, who knows to what purpose the MP filter he used was designed. However it sounded, it is a different filter than Dolby is using.

Deadwood Atmos theater
AV7702 Atmos 7.4.4, SSP-800 PLIIx 7.4
Aerial Acoustics 7B/CC3B fronts, B&W CWM8180 surrounds, Tannoy Di6 DC heights, Hsu ULS-15 subs
Roger Dressler is online now  
Old 05-28-2012, 07:07 PM
AVS Special Member
 
Roger Dressler's Avatar
 
Join Date: Jul 2000
Location: Oregon
Posts: 8,967
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 820 Post(s)
Liked: 364
Quote:
Originally Posted by Fanboyz View Post

All upsampling may not be detrimental, but it will undoubtedly make the resultant soundtrack measurably different from the master,

This would betray the concept of lossless audio.
Not probably anymore than a 24bit track being encoded as 16bit- but another technical wrinkle disquiet all our minds.

The audio is still delivered by a lossless codec. So your concern is this upsampling pre-process. Ask yourself how "lossless" the audio remains once it enters the AVR and hits various post-processing stages on the way to the amplifiers.

Deadwood Atmos theater
AV7702 Atmos 7.4.4, SSP-800 PLIIx 7.4
Aerial Acoustics 7B/CC3B fronts, B&W CWM8180 surrounds, Tannoy Di6 DC heights, Hsu ULS-15 subs
Roger Dressler is online now  
Old 05-28-2012, 10:12 PM
Senior Member
 
fuzz092888's Avatar
 
Join Date: Feb 2012
Posts: 491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 64
Quote:
Originally Posted by RBFilms View Post

With today's young society believing that ultra compressed lossy MP3 sounds better than a CD ... this education is necessary. Of course, most CD Discs sound terrible as well ... but they do sound better than an MP3 file.

This is not necessarily true.

http://www.soundandvisionmag.com/blo...ds-are-alright

http://seanolive.blogspot.com/2012/0...-japanese.html
fuzz092888 is online now  
Old 05-29-2012, 11:53 AM
AVS Special Member
 
AndreYew's Avatar
 
Join Date: Jan 2000
Location: Southern California
Posts: 1,670
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Roger Dressler View Post

Well, if the files null, then the artifacts of the 44.1 kHz anti-alias filter remain present. By definition, the apodized version should not null, or else the defect could not have been addressed.

I'm not sure I understand the reasoning here. If a 96 kHz-native signal (ie. originally recorded in 96 kHz) nulls against a 44.1 kHz signal that was downsampled from that 96 kHz signal with a linear phase (preringing) filter, then that means the linear phase filter did not introduce any extraneous artifacts.

Quote:


As to Mr. Katz experiments, who knows to what purpose the MP filter he used was designed. However it sounded, it is a different filter than Dolby is using.

This is a fair point, but the difference he found with the MP filter (no preringing) is entirely consistent with what MP filters do by their nature: the null decreased with increasing frequency, which means that the frequency response changed with higher frequency. It is well-known that MP filters alter both phase and amplitude of the original signal increasingly with higher frequency, just like any analog lowpass filter. He's posted frequency response graphs of the null elsewhere that show that behavior.

As for Dolby's filter, is it pretty different than what Meridian's used in the 808 and the HDMI adapter?
AndreYew is offline  
Old 05-29-2012, 01:14 PM
AVS Special Member
 
Fanboyz's Avatar
 
Join Date: Dec 2007
Location: Minnesota
Posts: 2,431
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 27 Post(s)
Liked: 19
Quote:
Originally Posted by Roger Dressler View Post

The audio is still delivered by a lossless codec. So your concern is this upsampling pre-process. Ask yourself how "lossless" the audio remains once it enters the AVR and hits various post-processing stages on the way to the amplifiers.

I do ask myself this all the time, the answer filtered and distorted...

The proper setting for sharpness is always0.
Also my Oppo BDP-103D is region free.
That makes me awesome.
Fanboyz is offline  
Old 05-29-2012, 03:21 PM
AVS Special Member
 
Roger Dressler's Avatar
 
Join Date: Jul 2000
Location: Oregon
Posts: 8,967
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 820 Post(s)
Liked: 364
Quote:
Originally Posted by AndreYew View Post

I'm not sure I understand the reasoning here. If a 96 kHz-native signal (ie. originally recorded in 96 kHz) nulls against a 44.1 kHz signal that was downsampled from that 96 kHz signal with a linear phase (preringing) filter, then that means the linear phase filter did not introduce any extraneous artifacts.

But he did not say it nulled in this test.
Quote:


Bob Katz: Then I did some null tests at 96 kHz on the results. The mp (as would be expected) produced a decreasing null in the high frequency range as the frequency increased. It was also surprising to hear a less than perfect null between the linear phase version and the original, and I have to put this at the feet of R8Brain as I have previously gotten 100% perfect nulls using Saracon for a round trip.

He does not describe the Saracon test other than "round trip" so I cannot know how he did it.

If indeed he has produced null tests for LP filters that show a perfect match, what does that say about the validity of such a test as a means for judging sonic purity when it ignores pre-ringing? Not even Voxengo claim that their LP filter avoids pre-ringing, only their MP filter.

Quote:


This is a fair point, but the difference he found with the MP filter (no preringing) is entirely consistent with what MP filters do by their nature: the null decreased with increasing frequency, which means that the frequency response changed with higher frequency.

Yes. Frequency and probably phase, too.

Quote:


As for Dolby's filter, is it pretty different than what Meridian's used in the 808 and the HDMI adapter?

It is certainly a new embodiment, but as to how it differs, I have no idea, nor would I say so if I did.

Deadwood Atmos theater
AV7702 Atmos 7.4.4, SSP-800 PLIIx 7.4
Aerial Acoustics 7B/CC3B fronts, B&W CWM8180 surrounds, Tannoy Di6 DC heights, Hsu ULS-15 subs
Roger Dressler is online now  
Old 05-30-2012, 05:43 AM
AVS Special Member
 
RBFilms's Avatar
 
Join Date: May 2006
Location: New York / California
Posts: 1,267
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 19
I got a kick out of this line in your post:

"Lots of important and experienced people didn't believe what he said, so he sat them all down and reasoned his way through everything he had done, until after a very long time, everyone was forced to agree with him."

I too have proven many things to many non-believers. The problem is, even after proving it ... there really was no impact as they revert back to their old way of thinking ... and I do not have the time or energy to convince everyone ... unfortunately.

The problem is, most "Engineers" do not use or trust their ears. I can easily demonstrate the Flaws & Negative Effects of what most Engineers consider sound & accepted "Engineering Principles" for both Analog & Digital Audio & Video. However, the ears and the eyes tell me a different story.

I trust my eyes and my ears over any instruments. If someone tells me that 48/16 is "Good Enough" ... because they have determined what I can or cannot hear ... and I can prove differently ... well ... then I have a problem with that.

If an engineer tells me that 18mps VC1 is superior to MPEG or AVC at twice the bit rate, I want to know what he is smoking because on my monitors it is obvious that this nowhere near the truth.

If an engineer tells me that dithering audio and up-converting improves sound when I know for a fact the opposite is true, then I have a problem with that as well.

All of these examples are from real experiences with engineers on AVS Forum by the way.

What bothers me most is how these "Engineers" ... when challenged ... dismiss my comments as ridiculous and unfounded and refuse to believe anything but their own science ... for lack of a better term.

All of the engineers I work with understand the theories & practices for best sound & picture that I advocate in these forums. What frightens me most is that the engineers driving technology do not ... or won't admit it if they do.

We listen to sound and we watch images. I am sorry if I am bothered by Engineers who pre-determine and make decisions on what is "Good Enough" for me. It is kind of like ... shut-up, sit down and like it because I am telling you what is best.

Wrong ... I am in the studio ALL the time and see some of the crap that is output and fed to consumers on a regular basis. It is part of why the I-Pod Generation does not know any better. People are forced to accept less than they should when it comes to picture & sound quality ... which has a direct impact on their entertainment experience in the home.

Personally, I try to advocate care when it comes to producing quality entertainment products within the limited capabilities of today's technology. Sorry if that upsets some folks, but I believe people should do their best when it comes to any job they do.


Quote:
Originally Posted by welwynnick View Post

I definitely agree about the need to minimise ALL digital audio signal processing. I don't think that audio is "safe" just because its in the digital domain.

A few years ago I was privileged to work with an exceptional electronics engineer, the kind of guy who could seemingly design anything from the ground-up, without using cookbook design (ie: an integrated solution and application notes). A rare man; and he told me an interesting story. He designed a digital audio mixing desk for a recording facility, and he tried hard to preserve audio quality. There were necessarily several processing stages, and testing showed that too many stages degraded digital audio quality. He investigated this, and found an interesting rule:

Every time you apply a digital process to data, on average, you lose half a bit of data.

Therefore, the data pathway has to be much wider than the minimum audio requirement, so he had to specify apparently over-the-top DSP precision to everything he did. Lots of important and experienced people didn't believe what he said, so he sat them all down and reasoned his way through everything he had done, until after a very long time, everyone was forced to agree with him. So if you want 16-bits worth of dynamic range, and you want to apply 16 processes to the digital audio (which isn't that much) then you need at least 24 bits just to preserve those 16 bits.

And what is a "digital process"? That's simple - its just any operation that causes the number to change. So that doesn't include any recording or switching functions, but it does include level change, filtering, equalization, conversion, etc. Any process that takes one number and converts it to another number. When you think about it, it stands to reason: when there's a calculation, the LSB will be either rounded or truncated, so there's an equal chance that it will be right or wrong. On average, it will be wrong half the time, so you lose that much precision or depth of modulation with every process. What's interesting is that its a linear loss, and there are no root-sum-squares in there.

So the short of it is that you need a much wider data pathway than the desired result. People like Meridian, who really understand digital audio, use 48-bit DSP in their recent processors, which sounds like overkill, but may be a necessity.

Nick


Richard J. Casey



Disney WOW - World of Wonder


Producers Guild of America, New Media Council
(BD Industry Insider)
RBFilms is offline  
 
Thread Tools


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off