Interesting article - "24-bit Audio Has More Resolution Than 16-bit Audio"

[rgpit]

http://benchmarkmedia.com/blogs/news/15121729-audio-myth-24-bit-audio-has-more-resolution-than-16-bit-audio?utm_source=Application+Notes&utm_campaign=f80542b9fc-AppNotes_15aug-+Audio+Myth+%28LINK%29&utm_medium=email&utm_term=0_e28f43b8aa-f80542b9fc-125366521&mc_cid=f80542b9fc&mc_eid=4758a337c5

[AdamG]

Good read!

I'm happy with my 24bit USB DAC, it's a huge improvement over the on-board audio I was using.

[Sir Thrift-a-Lot]

I am a firm believer in using 24 bit in the editing/mastering stage.   Any changes are much more transparent because the math used is so much more precise.   But I also agree with the article that if you properly dither your finished 24 bit file (I use isoTope RX), it is all but impossible to hear a difference.

[MacGeek]

Interesting article; 24 bit does it.  I particularly appreciate the last line.... "please do not underestimate the importance of a low-noise audio system"....this has been at least one primary audio goal from the beginning.

[bearjew]

There are many points in this that I agree, and I will not pretend to be an expert on DSD, but I know for a fact that DSD is not simply a 1-bit PCM digital recording standard.  DSD is a totally different system, where your bit-depth isn't necessarily going to determine your resolution.  24-bit is significantly better to listen to than 16-bit.  What he seems to fail to mention is that DSD is rarely used for recording (if it all), and only in post production (still not often).  Thrifty, I agree that 24 bit editing and mastering is very important, but I find it absurd that people don't just record in 24-bit as well; it's almost frustrating.  And also, Izotope RX is brilliant software.  It is the best restoration software out there next to CEDAR, which normally costs in the 10s of thousands of dollars.

[OldiesButGoodies]

I know, it struck me that the article refers to DSD as a "1 bit" format,  but it is accurate.  The technique for extracting the signal is way different,  so it is not fair (I think) to state that quantization noise in DSD is "huge".  I have tried to detect a difference in quality between the two formats but it is really hard for my tin ears.  Truth to me is that 44.1/16bit sounds pretty good and 44.1/24 bit sounds great. Anything above that gets hard to qualify as better,  since it can always depend on the sounds was mastered/mixed/messed with in addition to encoding at a higher rate.

P

[Sir Thrift-a-Lot]

I have tried to detect a difference in quality between the two formats but it is really hard for my tin ears. 

This quote opens up a whole other can of worms.   I firmly believe that many people who can hear the difference firmly believe that they can hear the difference.   If they are unwilling to DBT, I give them zero credence.

[F1nut]

DBT is not valid for the evaluation of stereo sound. Even Bell Labs, the inventors of stereophonic reproduction, knew that way back when.

[Sir Thrift-a-Lot]

DBT is not valid for the evaluation of stereo sound. Even Bell Labs, the inventors of stereophonic reproduction, knew that way back when.

Why?   

What scientific process do you instead favor for evaluation of claims?

[F1nut]

Some reading for you.

http://www.polkaudio.com/forums/showthread.php?104973-A-Historical-Overview-of-Stereophonic-Blind-Testing&highlight=bell

http://www.polkaudio.com/forums/showthread.php?111056-Further-Thoughts-On-ABX-Testing-Of-Stereophonic-Audio-Systems&highlight=bell

http://www.polkaudio.com/forums/showthread.php?98139-Questions-Regarding-Power-Cable-ABX-Test&highlight=test

Robert Harley's thoughts on the matter.

  Now back to the question of the blinded testing. Here is what the now publisher (Robert Harley) of one of the major magazines wrote a few years ago....


Quote:
Blind tests nearly universally appear to indicate that no differences exist between electronics, cables, capacitors, etc. In fact, one infamous test "revealed" that no sonic differences exist between power amplifiers. Mark Levinson, NYAL Futterman OTL tube monoblock, NAD, Hafler, and Counterpoint power amplifiers were all judged to be sonically identical to each other and to a $219 Japanese receiver (footnote 7). This very test, wielded by the objectivists as proof that all amplifiers sound alike, in fact calls into question the entire blind methodology because of the conclusion's absurdity. Who really believes that a pair of Futterman OTL tube amplifiers, a Mark Levinson, and a Japanese receiver are sonically identical? Rather than bolster the objectivist's case, the "all amplifiers sound the same" conclusion of this blind test in fact discredits the very methodology on which hangs the objectivist's entire belief structure.
 
If differences do exist between components, why don't blind tests conclusively establish the audibility of these differences? I believe that blind listening tests, rather than moving us toward the truth, actually lead us away from reality.

First, the preponderance of blind tests have been conducted by "objectivists" who arrange the tests in such a way that audible differences are more difficult to detect. Rapid switching between components, for example, will always make differences harder to hear. A component's subtleties are not revealed in a few seconds or minutes, but slowly over the course of days or weeks. When reviewing a product, I find that I don't really get to know it until after several weeks of daily listening. Toward the end of the review process, I am still learning aspects of the product's character. Furthermore, the stress of the situation—usually an unfamiliar environment (both music and playback system), adversarial relationship between tester and listener, and the prospect of being ridiculed—imposes an artificiality on the process that reduces one's sensitivity to musical nuances.

Going beyond the nuts and bolts of blind listening tests, I believe they are fundamentally flawed in that they seek to turn an emotional experience—listening to music—into an intellectual exercise. It is well documented that musical perception takes place in the right half of the brain and analytical reasoning in the left half. This process can be observed through PET (Positron-Emission Tomography) scans in which subjects listening to music exhibit increased right-brain metabolism. Those with musical training show activity in both halves of the brain, fluctuating constantly as the music is simultaneously experienced and analyzed. Forcing the brain into an unnatural condition (one that doesn't occur during normal music listening) during blind testing violates a sacrosanct law of science: change only one variable at a time. By introducing another variable—the way the brain processes music—blind listening tests are rendered worthless.







   

[Sir Thrift-a-Lot]

What scientific process do you instead favor for evaluation of claims?

[F1nut]

I believe you'll find the answer to your question in the links I provided.

Let me ask you one from the Robert Harley comments. Do you believe a test that results in a pair of Futterman OTL tube amplifiers and a Japanese receiver being sonically identical even remotely close to valid?

[Sir Thrift-a-Lot]

The fact that some people are deaf does not negate the validity of DBT.   In fact, it negates the validity of anecdotal evidence, which is what many seem to favor as the gold standard in audio methodology.   It seems to me that some people not hearing a difference between those two amplifiers proves that experience based rationale is utter hogwash.

[Sir Thrift-a-Lot]

PWK said it best.

[Sir Thrift-a-Lot]

I truly believe that people who scorn DBT are afraid that their own fallacies and shortcomings will be revealed.   I also believe that many "upgrades" in audio are placebo effect.   I don't really care what Joe Schmoe can and can't hear.   I care what I can and can't hear.   If I can't hear a difference, there isn't one (for me).   Anyone who says otherwise is likely doing a reach around for your wallet or is influenced by someone who is.