[castleofargh]

  I have tried and haven't found any difference between 16/44 cd and 16/44 from 24/96.
 
But i have a question..
is the effect of different downsampling algorithms used noticeable to a discerning ear..because of dithering?

some will tell you yes, some will tell you no. as always it's best to try abx yourself and make your own opinion.
I have a very hard time discerning stuff -60db under most playing musics, so I had no trouble making my opinion on dither. but that's me.

 

[kar13]

  Do you have detailed information about the origin of both versions that sound so obviously different ?
Maybe the 16 bit flac is a CD release version from the 80's and the 24/96 is a less than 5 year old remaster?
Just guessing of course

24/96 is a vinyl lp release from bruebeck's greatest hits..and the 16/44 is from cd version. vinyl is cbs records 1967 release.afaik the vinyl is not a remastered version.

If the dynamic range is different between the two versions, the master is different, and you aren't really just comparing the formats. Also, if you can hear the noise floor on the 16 bit recording, it's either noise from the original recording (tape hiss?), or it was mastered at far too low a level (like -40 or -50dBFS). It's also possible that your playback system has appallingly bad noise when playing 16 bit. For a fair comparison, take the 24 bit version and downconvert it to 16/44, then upconvert it back to 24/96, then compare those two.

 
I have tried it on pc's with a dac and hence Im sure its not the noise in my system and that is why i dared to post such controversial observations as it baffles me..
 

   
It's really pretty simple.
 
We're NOT talking about generations in terms of copies not being the same as originals. Of course, since it's just bits we're talking about, a copy will be identical to the original (if you don't do something wrong). The same is true if you convert from one lossless format to another - no change. However, when you convert a digital audio file to a different sample rate, new numbers must be calculated, and part of the process involves digital filtering. It is NOT a totally lossless conversion. Therefore, whenever you convert one sample rate to another, you slightly alter the content. (For example, if you convert a 16/44k file to 16/96k, then convert the 16/96k file back to 16/44k, you might expect to end up with exactly what you started with, but you won't.) Some conversion programs alter the signal more than others, and some offer several different options in terns of filtering, but none is "absolutely perfect". Therefore, if you have a file that was actually mastered at 24/192k, and you buy a copy at 24/192k, your copy should be identical to the original. However, if you convert it to 16/44k, even excluding any possible difference because of the different sample rate, you will also have differences due to the artifacts of the conversion process. (And so, if what you're buying was mastered at 24/192k, a 24/192k copy should be identical to the original, but, even excluding any possible audible difference because of the different sample rate itself, the 16/44k version will have been altered slightly by the conversion process itself.)
 
An easy to demonstrate this for yourself is to start with a 24/96k "master file" and convert it to 16/44 using several different "high end" programs; you will find that the results are similar, but not identical (and sometimes the differences are audible).

thanks for clarifying my doubt in my previous post!

   
Please do tell about these alleged listening experiences. Were they carefully controlled comparisons with adequate quality controls or were they the usual audiophile casual sighted evaluations?
 
The usual casual sighted audiophile evaluations are utter junk when it comes to reliable tests of sound quality.
 
They usually suffer from one or more of the following fatal problems, usually all of them. Any of these problems is fatal to the credibility of the evaluation for the reasons stated:
 
(1) Audiophile Casual Sighted evaluations are not reliable evidence because they are not even tests. That is, they do not involve comparison to a fixed, reliable standard.
 
(2) Audiophile Casual Sighted Evaluations are not  reliable evidence because they involve excessively long switchover times, which makes them highly susceptible to false negatives because they desensitize the listeners.
 
(3) Audiophile Casual Sighted Evaluations are not reliable evidence because the do not involve proper level matching, which makes them highly susceptible to false positives because people report the level mismatches as sonic differences.
 
(4) Audiophile Casual Sighted Evaluations are  reliable evidence because they do not involve listening to the identical same piece of music or drama within a few milliseconds, creating false positives because people report the mismatched music as sonic differences in the equipment.
 
(5) Audiophile Sighted Casual Evaluations are not  reliable evidence because they constantly reveal the true identity of the UUTs to the listener, creating false positives because people report their prejudices and preconceived notions as sonic properties of the equipment.

I use iems on any system where a/b ing is done.
I mentioned in my post that (1), (3) are already taken care of.
(2)Over many comparison tests that i have done and continue to do this effect is of course taken care of.
(4) is overcome as the 2 versions are ripped in random sequences by software and usually burned to a cd.
The environment is ensured to be quiet, and although I do understand your skepticism,i am not enforcing you to believe in my experience.
If there is anything else I can do, to prove this as false positive please tell me about it.
 
I would like to point out that this happens only with Take five and its gotten to the point where I look for sonic cues and differences and can distinguish almost effortlessly..
I have done similar A/b ing with tracks of edm electronic, and rock genres and have been unsuccessful at similar identification.

  I bought Mark Ronson - Uptown Funk in 88.2K/24 FLAC format it's nice and clean, loud without any mid-frequency issue. Whether the 44.1K CD version sounds the same is irrelevant to me.

+1 enjoyment of music always rules..

 

[interpolate]

There is a reason most modern mastering limiters support a 60dB dynamic range. The average RMS peak determines how loud the actual recording is and how it's been compressed.

 

[kar13]

 
some will tell you yes, some will tell you no. as always it's best to try abx yourself and make your own opinion.
I have a very hard time discerning stuff -60db under most playing musics, so I had no trouble making my opinion on dither. but that's me.

i will have to take this as my next experiment..thanks for your input!

 

[old tech]

Bear in mind this peer reviewed paper of some 10 years standing has not been refuted...

http://drewdaniels.com/audible.pdf

 

[arnyk]

Bear in mind this peer reviewed paper of some 10 years standing has not been refuted...

http://drewdaniels.com/audible.pdf

 
Moran, one of the two authors emailed me earlier this year pointing out a recent AES conference paper that some thought constituted a refutation:
 
http://www.realhd-audio.com/?p=3627
 
Mark Waldrep (well known High Rez advocate) wrote:
 
"I finished the afternoon by attending a few paper sessions. The first was titled, “The Audibility of Typical Digital Audio Filters in a High-Fidelity Playback System”. Although it may not be obvious from the title of the paper, this is the first AES publication that refutes the Meyer/Moran research that has been so often quoted as “proof” that CD specification PCM audio is enough for music reproduction (Meyer and Moran’s research has been widely discredited including by myself because of the lack of real high-resolution content used during the study)."
 
However, that was before the AES community commented on the alleged refutation:
 
https://secure.aes.org/forum/pubs/conventions/?ID=416
 
For example:
 
"
The conclusions in the abstract and in the introduction differ markedly from those offered at the end of the paper. They are also not adequately supported by the research presented in the paper.
The introduction contains a lengthy series of speculations which seem to be preoccupied with casting doubt on some past research, yet the paper essentially fails to substantiate the speculations. This comes dangerously close to being unfair.
"
Any lack of real high resolution content used in the M&M study is the fault of the promoters of the SACD and DVD-A formats, as it was recordings that were advertised as being high resolution audio that formed the basis of the Meyer and Moran study.
 
After the publication of the Meyer and Moran study it became known that about half or more of the SACD and DVD-A formatted recordings were sourced from low resolution masters, and once that resolution is gone, just putting it on media that is capable of high resolution performance does not correct the situation at all. 

 

[arnyk]

I use iems on any system where a/b ing is done.
I mentioned in my post that (1), (3) are already taken care of.
(2)Over many comparison tests that i have done and continue to do this effect is of course taken care of.
(4) is overcome as the 2 versions are ripped in random sequences by software and usually burned to a cd.
The environment is ensured to be quiet, and although I do understand your skepticism,i am not enforcing you to believe in my experience.
If there is anything else I can do, to prove this as false positive please tell me about it.
 

 
Here is your original post:
 
Unfortunately, whenever you convert a file, there is filtering involved, which introduces some (very slight) changes to the sound. If you don't hear a difference between the 24/96k original and the converted 16/44 version, then this will prove that neither difference is significant (in your test rig, with your sample content, and through your ears). But, if you do hear a slight difference, you won't know for sure how much of it is due to the difference in sample rate or bit depth, and how much is due to slight alterations introduced in the conversion process itself.
 
It's also fair to remind everyone that we are talking about subtle differences here... subtle to the point where they may only be audible at all with certain source material, or with certain speakers or headphones, and maybe not all of us can hear them at all... however that doesn't rule out the fact that they may be audible and significant under some circumstances. (Either way, they may well be less major than differences between speakers or headphones.)
 
Another post suggested that the differences may be due to other things - perhaps because a different master was used. While this is certainly true, I think its importance to end users is being overemphasized. It may matter a great deal to a music producer, or a streaming service, whether a high-def download really sounds better because it's high-def or because it was remastered. However, to the person buying and listening to it, all that matters is that the "high-def remaster" does in fact sound better. Most of the current crop of high-def reissues have been remastered, often in a way that is significantly better than the original CD version... in which case it's worth buying because of the better remastering. (I might even suggest that, once we've established that the 24/192 remaster of Album X sounds better than the CD version, we can also assume that the "master copy" of the remaster was in fact done at 24/192. Therefore, even if the sample rate itself doesn't make a significant difference, the 24/192 version will be "a copy of the master" while the 16/44 "CD quality" version will have been converted form that new master - and so will be "one generation out - and possibly slightly different". In that situation, even if we were to agree that the fact that one is at 24/192 didn't actually matter, it would still make sense to buy the 24/192 version that was a direct 1:1 copy of the new master, rather than the 16/44 version which had been converted from it.)
 
If you accept that many remasters sound better than the original - for whatever reason - then it simply doesn't make sense to agonize over the differences between a 16/44 version and a 24/192k version (the price difference is usually small, and the cost of storage has gotten so low that the size of the file itself simply doesn't matter all that much). A lot of people bought a lot of music in AAC or MP3 format, only to find out later that the difference was obvious on "their new stereo", and end up being disappointed, or end up spending a lot of money buying their collection all over again. I'd rather spend a few $$$ more and buy the best quality version that's available when I make my purchase, and so minimize the risk of having to buy it again later.
 
(I think it's kind of cool to be able to buy an actual copy of the 24/192k remaster; I used to just hate buying a vinyl album... and knowing that what I had in my hand wasn't nearly as good as the original version that was recorded on the master tape. Even if, in a particular case, I don't notice a difference, It makes me feel better to know that there isn't a better quality copy out there "that I'm missing".)

 
 
 
 
I don't see where in the above any of the points I raised were even touched on, so the entire above makes no sense to me.  Perhaps a little fact-based discussion of the points raised rather than the wave of a hand and flat dismissal would help?
 
For example, using IEMs addresses zero of the points I raised.  They are fine, I use headphones, earphones and loudspeakers about equally often, but so what?
 
One of the points - the first point about audiophile evaluations not even being tests, can't be "taken care of" because it is an inherent failing.
 
The casual apparent  attempt at sort-of randomization shouldn't impress anybody who even took just first year statistics.

 

[old tech]

 
Moran, one of the two authors emailed me earlier this year pointing out a recent AES conference paper that some thought constituted a refutation:
 
http://www.realhd-audio.com/?p=3627
 
Mark Waldrep (well known High Rez advocate) wrote:
 
"I finished the afternoon by attending a few paper sessions. The first was titled, “The Audibility of Typical Digital Audio Filters in a High-Fidelity Playback System”. Although it may not be obvious from the title of the paper, this is the first AES publication that refutes the Meyer/Moran research that has been so often quoted as “proof” that CD specification PCM audio is enough for music reproduction (Meyer and Moran’s research has been widely discredited including by myself because of the lack of real high-resolution content used during the study)."
 
However, that was before the AES community commented on the alleged refutation:
 
https://secure.aes.org/forum/pubs/conventions/?ID=416
 
For example:
 
"
The conclusions in the abstract and in the introduction differ markedly from those offered at the end of the paper. They are also not adequately supported by the research presented in the paper.
The introduction contains a lengthy series of speculations which seem to be preoccupied with casting doubt on some past research, yet the paper essentially fails to substantiate the speculations. This comes dangerously close to being unfair.
"
Any lack of real high resolution content used in the M&M study is the fault of the promoters of the SACD and DVD-A formats, as it was recordings that were advertised as being high resolution audio that formed the basis of the Meyer and Moran study.
 
After the publication of the Meyer and Moran study it became known that about half or more of the SACD and DVD-A formatted recordings were sourced from low resolution masters, and once that resolution is gone, just putting it on media that is capable of high resolution performance does not correct the situation at all. 

 


The irony is that the so called flaw is that none of the listeners could tell which of the SACDs or DVD-As were sourced from the lower resolution masters. If anything, this strengthened the Moran and Myer findings.

 

[StanD]

 

   
Moran, one of the two authors emailed me earlier this year pointing out a recent AES conference paper that some thought constituted a refutation:
 
http://www.realhd-audio.com/?p=3627
 
Mark Waldrep (well known High Rez advocate) wrote:
 
"I finished the afternoon by attending a few paper sessions. The first was titled, “The Audibility of Typical Digital Audio Filters in a High-Fidelity Playback System”. Although it may not be obvious from the title of the paper, this is the first AES publication that refutes the Meyer/Moran research that has been so often quoted as “proof” that CD specification PCM audio is enough for music reproduction (Meyer and Moran’s research has been widely discredited including by myself because of the lack of real high-resolution content used during the study)."
 
However, that was before the AES community commented on the alleged refutation:
 
https://secure.aes.org/forum/pubs/conventions/?ID=416
 
For example:
 
"
The conclusions in the abstract and in the introduction differ markedly from those offered at the end of the paper. They are also not adequately supported by the research presented in the paper.
The introduction contains a lengthy series of speculations which seem to be preoccupied with casting doubt on some past research, yet the paper essentially fails to substantiate the speculations. This comes dangerously close to being unfair.
"
Any lack of real high resolution content used in the M&M study is the fault of the promoters of the SACD and DVD-A formats, as it was recordings that were advertised as being high resolution audio that formed the basis of the Meyer and Moran study.
 
After the publication of the Meyer and Moran study it became known that about half or more of the SACD and DVD-A formatted recordings were sourced from low resolution masters, and once that resolution is gone, just putting it on media that is capable of high resolution performance does not correct the situation at all. 

 


The irony is that the so called flaw is that none of the listeners could tell which of the SACDs or DVD-As were sourced from the lower resolution masters. If anything, this strengthened the Moran and Myer findings.

I doubt that's going to fly since the common source master's resolution was already limited. To be valid I would think that all test materials be made from the same high resolution source.

 

[arnyk]

  I doubt that's going to fly since the common source master's resolution was already limited. To be valid I would think that all test materials be made from the same high resolution source.

 
I totally agree with both points.  
 
The use of resolution-limited sources invalidates the experiment. For the test to be valid as proposed, high resolution sources had to be used.
 
There are a number of valid defenses of Meyer & Moran:
 
(1) Their mistake was to take vendor claims at face value. They and the rest of the audio world were hoodwinked by the publicity and marketing of the SACD and DVD-A media. 
 
(2) In fact nobody seems to have noticed the absence of actual high resolution source material for 5-6 years after the introduction of the SACD and DVD-A.
 
(3) The people who finally brought this issue to light seem to have based their knowledge on technical tests, not listening tests.
 
Putting these facts together, we find evidence of a large, unintentional DBT providing solid evidence that there are no audible benefits to high resolution media because the legacy media format (audio CD) is entirely adequate for accurate distribution of music as it is recorded to this day. This conclusion is also supported by all extant formal testing to this day.

 

[old tech]

Hey Arnyk

Why not ask Moran to redo the study and settle this once and for all? Apart from the vendor supplied music issue, the time that has elapsed since that paper (particularly with all the marketing claims around downloadable 24/96 and 24/192 files, it would be very prescient.

 

[arnyk]

Hey Arnyk

Why not ask Moran to redo the study and settle this once and for all? Apart from the vendor supplied music issue, the time that has elapsed since that paper (particularly with all the marketing claims around downloadable 24/96 and 24/192 files, it would be very prescient.

 
 
I actually haven't heard much of anything from those guys since. Brad Meyer's public email address is: EBradMeyer@att.net . 

 

[RRod]

Hey Arnyk

Why not ask Moran to redo the study and settle this once and for all? Apart from the vendor supplied music issue, the time that has elapsed since that paper (particularly with all the marketing claims around downloadable 24/96 and 24/192 files, it would be very prescient.

 
I would surprise me if any hi-res vendor would now willingly donate material to such a test, especially if results were again destined for a professional publication. Too much to lose; nothing to gain.

 

[old tech]

 
I would surprise me if any hi-res vendor would now willingly donate material to such a test, especially if results were again destined for a professional publication. Too much to lose; nothing to gain.

 

But why is a vendor required? Could they not just download commercially available hi res files, purchase some DVD A's and SACD's and do the test themselves?

 

[arnyk]

   
I would surprise me if any hi-res vendor would now willingly donate material to such a test, especially if results were again destined for a professional publication. Too much to lose; nothing to gain.

 
 
The last instance of such a thing that I know of involved a listening test that was conducted on the AVS forum by a representative of a high def recording vendor: AIX.
 
The best test files he could provide included a relatively large sliding inter-recording time delay that was easy enough to ABX all by itself. He was made aware of this promply but when I downloaded the latest versions of the files some months later the delay was still there and I could ABX it  quite quickly 16/16.