Digital Music Files: 16/44.1, 24/88, 24/96, 24/192 - Page 4

Nonsense. Dynamic range is limited in CDs for purely marketing (the "loudness wars") reasons, NOT technical ones. Again, you've ignored all the CDs that have NO dynamic range compression, and you've FAILED to name a SINGLE sacd with dynamic range exceeding that of CD.

You just refuse to get it do't you?

Here you go, actual test proving minimum of at least 3dB extra dynamic range for SACD. The actual measurements showed SACD > DVD-A > CD in dynamic range. Also note the comment that SACD recorded level is slightly lower than others then if volume is not adjusted in ABX comparisons, listeners may get the wrong idea.
http://www.audioholics.com/education...-part-1-page-2

And here is another - analyzing pink floyd dark side of the moon CD and SACD recording from the same master. "There are no clipped samples in the DSD data, but a whopping 362 in the CD data." and "More importantly, while the peaks of each file are the same, the average RMS power of the CD data--15.29dB average of both channelsis significantly higher than that of the SACD layer (17.7dB). No wonder the CD layer sounded louder. This suggests compression or peak limiting was used in the mastering to reduce the song's dynamic range compared with the SACD." and "The solo, however, has the squared-off shape that results when the music has been run through a peak limiter. This is a device that literally chops off the transient peaks, allowing the average level to be higher, hence louder."

So there you go - this is typical of 90%+ of CDs where mastering engineers boost the average level at the expense of dynamic peaks - resulting in clipping. I have not seen this done in SACDs - if you have provide proof.

http://www.stereophile.com/news/11649/

But I am sure you will invent another BS excuse to deny the truth.

All you've done is show that recording engineers like to compress the dynamic range (for PURELY COMMERCIAL REASONS, NOT TECHNICAL ONES), and a paltry 3 dB at that. You have NOT shown it's necessary. By the way, your own words demonstrate that dynamic range compression is not necessary on CD, else you would have said it happens 100% of the time.

And you STILL are woefully incapable of citing a SINGLE solitary SACD with a dynamic range greater than what CD is capable of. Not one....none at all...

You are a lost cause and obviously can't read.
First link shows you that the SACD has at least 3 dB higher dynamic range.
That is twice the volume if you understand the scale. And that CD version did not have clipping.

And obviously you reasoning is challenged - I have stated several times, but it does not penetrate your thick skull - that he few CDs that do not apply compression are recorded at low levels exactly because you have the dynamic limit. if they don't turn down the level you will exceed the dynamic range and clip naturally, rather than through technology. Obviously you have never been in a live recording session and have NO clue of how much dynamic variation there is.

And I could care less if the compression is commercial or technical.
The music you buy on CD is damaged because of it and it is very audible.

So go on and continue to play ostrich and keep sticking your head in the whole and deny REALITY.
People talk about global warming deniers.
Well you are a denier of another sort.
I am sure that even if someone dragged you into a studio and showed you live music peaks exceed both dynamic range and frequency range of CD you would still deny it until your last dying breath.

What I see is you insistent denial of reality of business practice with recording CDs, deny wider dynamics and frequency response of high res recordings, actual facts shown to you, just to justify to yourself spending few$ less per CD and justify your purchase of inadequate reproduction equipment.

As I have said before, if you are happy with just CDs, be happy.
Now go stick your head in the sand so that reality does not hit you in the face.

Peace

So now you're claiming that a 3 dB difference is twice as loud? I'm amused by your confusion of power levels with loudness levels.


In other words, having failed to show that there's a technical reason to compress CDs, you're attempting to claim that commercial reasons somehow support your argument.


A further concession by you that CDs are compressed for commercial reasons, not technical ones.

3db isn't twice as loud. You'd need two times the power to play 3db louder but it isn't twice as loud. If 3db more was twice as loud having two subs would be serious spl (I wish)

Yep. Twice as loud needs more like 10 dB.

All easily achievable with 16 bits.

I have personally played CDs at 120 dB SPL levels. What's your point?


Why would I do such a thing?

What limitation of the CD format prevents me from playing it any SPL that my speakers can achieve?

Are you telling me that amps or speakers can somehow *know* that the source is a CD and automagically reduce their gain or clip their output at some predetermined SPL level?

Aw, come on! ;-)

That's a possibility, but not as you seem to be claiming a necessity.

There are low efficiency speakers that can take monumental amounts of power and reproduce it with reasonable linearity.


Wrong, there are low efficiency speakers that can reproduce the peaks present in live music. There have been speakers like this for decades.

I'll ignore your qualification of "full range" because its your own invention. It's also irrelevant and gratuitous because subwoofers are well-known and widely used as separate components.

Here is just the tip of the iceberg in large Xmax subwoofers:

http://www.soundstream.com/AudioProd...oofers/xxx.htm

Soundstream XXX-18 18 10000 Watt RMS Quad 1.2 Ohm Competition Drag Subwoofer - Retail price < $1000, Xmax=30 mm

Soundstream XXX-15 15 8500 Watt RMS Quad 1 Ohm Competition / Db Drag Subwoofer - Retail price < $1000, Xmax 38 mm

http://www.parts-express.com/pe/show...number=293-666

TC Sounds LMS Ultra 5400 18" DVC Subwoofer Tetail price < $1000, Xmax 38 mm

https://ssl.perfora.net/www.ficaraud...iew.shopscript

SP415 Dual 2 < $500 Xmax = 33 mm

etc., etc., etc.

In practice, yes!




Ironically you just destroyed your own argument. If setting the peak levels of CD playback would result in average levels that were too low for most people that would mean that the CD has too much dynamic range for most people. And you would be right, because recordings that fully exploit the CD format do have too much dynamic range for many people - because when you turn up the peaks, the quiet passges may be too hard to hear in your typical listening room due to its background noise in the 30-40 dB SPL range.

I can't believe that you haven't ever played a CD that loud - with peaks at 105 dB! I surely have and with low efficiency speakers (NHT 2.5i )



Again your post shows a complete and utter lack of proper understanding of dynamic range. Dynamic range is not limited by maximum levels since anybody with a volume control can turn them up as much as their ears and equipment can bear.

The aspect of playback that separates poor dynamic range from good dynamic range is the quiet passages. With high dynamic range you can play the peaks loud and while the quiet passages might be barely audible, but they are composed of music not noise.

With poor dynamic range you can still play the peaks loud, but when you listen for quiet passages, they either aren't there or they are noisy.

We lived this way all the time in the days of LPs.

It turns out that if you go into a concert hall or a good recording studio and record music with the best microphones and recorders, you end up with recordings that have maybe 65-75 dB dynamic range. The widest dynamic range on any commercial recording I've ever inspected, and I've inspected a lot, was about 85 dB. I've personally made recordings with a little more than 85 dB dynamic range, but it was a real pain. The problem isn't the equipment, its the room and the people in it.

BTW I'm sure that I own more 24 bit converters than you do, and some of them bench check out at very close to 120 dB dynamic range. They can't help with the room and the people.

Pick a number - 65, 75, or 85 dB dynamic range - none of them tax the CD format.

What you haven't shown is that the difference was necessary.



Compressing CDs is commonly done, but its done for commercial and artistic reasons, not technical reasons. There are no doubt many SACDs and DVD-A that are also compressed because it is part of the legacy of the classic master recordings that they were made from.



Again a complete misunderstanding of dynamic range. Avoiding clipping is easy, you just turn down the gain. What you can't easily avoid is when the music gets so faint that it gets lost in the noise that is inherent in the medium. Thing is, that doesn't have to happen because 16 bits used optimally is more than enough to deal with every commercial recording that is known to exist.



With more than 1,000 live recording sessions behind me I have no clue of how much dynamic variation there is?

LOL!


This comment doesn't seem to relate to the Meyer and Moran JAES paper which was mentioned earlier in this thread:

http://www.aes.org/e-lib/browse.cfm?elib=14195

http://www.bostonaudiosociety.org/explanation.htm

The essence of their experiment was playing so-called hi-rez recordings and listening for the audible effects of reducing them to CD format sample size (16 bits) and sample rate (44.1 KHz). There were no audible effects.

This experiment at the very worst shows that while the hi-rez formats have greater theoretical potential, in actual listening it vanishes due to other practical considerations.

I don't understand your point here. Are you talking frequency response or dynamic range?
Dynamic range is the ratio of the loudest sound to softest sound you can record. It is dependent on word size.
Sampling frequency determines frequency response of the system.

Regardless, a millisecond is a relatively long time for CD format.

I've proven that so much of what you have said is wrong, and you have provided no effective defense.


How can you seriously say that about someone who has made over a thousand live recordings?


Ironically, more than half of those thousand recordings were made on a CD recorder. I've examined 100s of those recordings in a wave editor and found no clipping, and lowest level signals that were 20 dB or more above the noise floor.


I've also tested dozens of CD players using test equipment running @ 24 bits with > 110 dB dynamic range and 192 KHz sampling with > 90 KHz bandwidth.



Two words - Meyer and Moran.

"With the help of about 60 members of the Boston Audio
Society and many other interested parties, a series of
double-blind (A/B/X) listening tests were held over a period
of about a year. Many types of music and voice signals
were included in the sources, from classical (choral,
chamber, piano, orchestral) to jazz, pop, and rock music.
The subjects included men and women of widely varying
ages, acuities, and levels of musical and audio experience;
many were audio professionals or serious students of the art."

This isn't about my personal hearing limitations, its about tests involving large groups of individuals, including dozens of serious audiophiles.


Still waiting for a relevant response based on other than your personal unsupported assertions.

I've sat on the sideline for this, but thought i'd jump into the frey. There is some speculation and possible flaws on those results over at head-fi.org quoting sacd.net. Here is the quote and link to the post.


http://www.head-fi.org/t/486598/test...0#post_7553406


In reference to item no. 13 of the initial post of this thread, there was a rather extensive discussion over the methodology and validity of these tests over at SA-CD.net (http://sa-cd.net/showthread/42987/42987/y#42987). It seems that the authors unknowingly used mostly SACDs sourced from low-resolution recordings in the tests (as per their own list at http://www.bostonaudiosociety.org/explanation.htm), thus making the results statistically flawed. In other words, if you're testing someone with low-rez on one side and low-rez on the other, the result is obviously irrelevant. Maybe the post should be amended to include this information.

OTOH, there is this AES paper (below) on the same topic, albeit limited to comparisons between 44.1 and 88.2 kHz. That in turn, leads one to wonder what their findings would be with higher resolution formats such as DXD (352.8 kHz 24 Bit PCM) and/or DSD. Regardless, this blind ABX test should be included in the initial post as well...

Sampling Rate Discrimination: 44.1 kHz vs. 88.2 kHz
(http://www.aes.org/e-lib/browse.cfm?elib=15398)
It is currently common practice for sound engineers to record digital music using high-resolution formats, and then down sample the files to 44.1kHz for commercial release. This study aims at investigating whether listeners can perceive differences between musical files recorded at 44.1kHz and 88.2kHz with the same analog chain and type of AD-converter. Sixteen expert listeners were asked to compare 3 versions (44.1kHz, 88.2kHz and the 88.2kHz version down-sampled to 44.1kHz) of 5 musical excerpts in a blind ABX task. Overall, participants were able to discriminate between files recorded at 88.2kHz and their 44.1kHz down-sampled version. Furthermore, for the orchestral excerpt, they were able to discriminate between files recorded at 88.2kHz and files recorded at 44.1kHz.
Authors: Pras, Amandine; Guastavino, Catherine
Affiliation: McGill University, Montreal, Quebec, Canada
AES Convention:128 (May 2010) Paper Number:8101 Import into BibTeX
Subject:Audio Coding and Compression
E-Library Location: (CD 128Papers) /128/8101.pdf

This issue has already been discussed in this thread. The statement that Meyer and Moran's results are "statistically flawed" is incorrect. Their test is not flawed, and it shows what it shows. The statement above merely a hypothesis that a different test would produce a different result.


Haven't seen this yet, but the results are not unprecedented. There appear to be circumstances in which the reproduction of ultra-high frequencies can cause distortion in the audible range. Not sure how strong an argument that is for maintaining the higher sampling rate.

I'm sure it was unintentional, but you kinda misstated the actual nature of the problem. It has subsequently come to light from several sources that an incredibly high proportion (about 50%) of all SACDs and DVDs are sourced from low-resolution recordings, not just the ones chosen by Meyer and Moran.

This problem was detected by technical means, not listening tests of any kind. Information that was actually evidence of this problem was presented publicly by David Greisinger as early as 2003 at an AES conference in Baniff Alberta. http://www.davidgriesinger.com/surro...om_stereo2.ppt

"All commercial recordings tested by the author as of 6/1/03 contained either no ultrasonic information, or ultrasonic harmonics at levels more than 40dB below the fundamentals."

If there was a significant audible difference, why weren't the fakes called out by the golden eared reviewers when they reviewed them after their release?



That paper has been criticized for its (mis)use of statistical methods. There's an international standards group publication that describes both the testing methodology and recommended statistical methods for doing this kind of test.

http://www.itu.int/dms_pubrec/itu-r/...0-I!!PDF-E.pdf

Why aren't people following it? Are they saying that generally recognized statistical analyses have a built-in blind eye to tests involving bandwidth and bit depth? ;-)

You have proven nothing - every time I point to real evidence you go finding another excuse or stating that you will ignore that fact.

And just an FYI for those accompanying this thread these wonderful fellows who do not like to be inconvenienced with evidence contrary to their opinion, went crying to the board moderator that I have insulted them.

Must be because they have such a thin skin and do not like to see proof debunking their story. They had my last post deleted, even though there was absolutely nothing insulting in it.

Thought you ought to know.
And as I said in my deleted post - I am done wasting my time discussing an issue with a brick wall.

Anyways

Your deleted post had a political reference which was probably the reason it was removed. Forum rules are clear on this.

Good one! Gushing reviews of "hires superaudio" recordings that turn out to have been made from NON "hires" sources.

Yes we saw that the other 3 times you posted it.

I've found the moderation here quite fair and honest. Personally, I would prefer you continue posting, but rather then continued bluster, stick to the verifiable facts. That goes for all sides.

More then a few people seem to disagree with that.


Real evidence? You mean your false claim that the SACD format was 24 bit, or that it had 144 dB dynamic range? Did you ever get around to reading the independent source that I quoted to correct that?

Excuses? The only thing you posted that I said I would ignore was a gratuitous and irrelevant condition that you falsely added - that only woofers that were built into full-range speakers could be considered in a discussion of Xmax. Why did you add that irrelevant restriction?


I know nothing about that, only what I read here. I don't recall ever seeing the post that was apparently deleted.


You appear to have an odd idea about debunking. Repeated unsupported claims do not constitute debunking.


Ahh, the mysterious "they"... ;-)


Given the quality of the arguments you've posted here lately, no great loss!

I think he's half way to China by now.

I'd prefer he stay at the center.

These questionable reviews can be found all over the high end world.

Those low res SACD and DVD-A recordings may form one of the largest unintentional DBTs in the history of audio. ;-)

No surprise, all of the golden ears were found to be guessing randomly (about 50% right).

Why do you believe that the volume extremes will be reproduced differently (acoustically) when the source is 24/96 versus 44/16? Can you provide an example for each format showing where the level differences occur? I'm not talking about processing audio with volume compressors. Just show us the math and sequence of level changes that cause the volume extremes to vary based on delivery format.

--Ethan

Still hasn't responded about 3db increases being double the volume

Thanks Arny for summarizing this so clearly. I should have read further before posting above.

--Ethan

Well, it is rather embarrassing for him.

Everything worked together to give our side of the discussion a lot of weight.

It is kinda counter-intuitive that wide dynamic range material generally plays softer at the same level setting, but this is generally true. Don't people notice this?

Anybody who switches from playing broadcast TV to playing DVD or Blu Ray movies on the same system should be encountering this. The desired volume setting for broadcast TV is generally lower than that for a BD or DVD movie, most other things being equal.