The above statement is to the best of my recollection is not accurate. Looks more like an exaggerated conflation of posts on different subjects.
Let's take the last part first, "deaf or an EE." There are not a vast number of posts claiming that, yours was probably the first to put those two words together in that way. What people usually suggest as possible explanations of why someone doesn't hear something are: 1. Maybe there is some problem with your system, and 2. Maybe you don't know what to listen for. Rarely does it get to the level of name calling where people are throwing around terms like 'deaf.' That tends to happen after tempers start to flare.
Again IIRC the thing about EEs is not what people generally say in conversations about sighted listening. Maybe it could be a reference to some points I was recently trying to make about overreliance on and or misinterpretation of spectral analysis results. If so, that was not in the context of calling anyone deaf. I did suggest that looking a spectral peaks first might influence what someone listens for later.
Regarding the claim of, 'subjective sighted listening results as valid for everyone,' I don't see where that comes from. It has often been acknowledged that people sometimes experience errors of listening. I have tried to make a case that such errors, like many or most cognitive errors, are most likely not random. Human minds tend to make systematic errors. If so, it means we can work on learning to recognize when we are making such errors.
As to 'anyone questioning those results is...,' is another exaggeration. The point you recently raised about low phase noise clocks for ethernet equipment was worded fairly and not exaggerated.
Maybe you don't realize how you are coming across sometimes when you write things that sound emotional, mocking, exaggerated, inaccurate, and or contemptuous. Its something you do from time to time, maybe when you are tired and or feel fed-up with someone. Not sure. OTOH when you write like a professional it puts you in a better light for all readers. It makes people much more likely to see you as someone to be taken seriously.
Last edited:
Regarding "subjective sighted listening results as valid for everyone" forum search with "sound better" just from author Markw4 returns about 200 hits. Presumably all those are based on subjective sighted listening tests as you do not perform DBTs. If you make a "sound better" claim you presumably think that your listening result is valid to others as well or why make such a claim.
First search result is at: https://www.diyaudio.com/community/search/454348/?q="subjective+sighted+listening+results+as+valid+for+everyone"&c[users]=Markw4&o=relevance
Regarding use of "sound better," or "sounds better," I don't see a problem expressing such an opinion on SQ. Rather I would say its necessary since there is no equivalent measurement methodology. JBL, ESS, and Purifi all use listening tests, and for good reasons.
Regarding use of "sound better," or "sounds better," I don't see a problem expressing such an opinion on SQ. Rather I would say its necessary since there is no equivalent measurement methodology. JBL, ESS, and Purifi all use listening tests, and for good reasons.
Regarding audibility of dac filters, there is a thread on it: https://www.diyaudio.com/community/threads/i-cant-hear-the-difference-in-filters.335711/page-3
A few particular posts:
https://www.diyaudio.com/community/threads/i-cant-hear-the-difference-in-filters.335711/post-5740776
https://www.diyaudio.com/community/threads/i-cant-hear-the-difference-in-filters.335711/post-5740039
https://www.diyaudio.com/community/threads/i-cant-hear-the-difference-in-filters.335711/post-5740078
There have been similar expressions of inaudibility in other threads, the problem is figuring the right search terms for the exact wording. Do recalling a claim about the filters being only for marketing, that they don't sound different. Don't know how much time I want to spend searching for it though.
A few particular posts:
https://www.diyaudio.com/community/threads/i-cant-hear-the-difference-in-filters.335711/post-5740776
https://www.diyaudio.com/community/threads/i-cant-hear-the-difference-in-filters.335711/post-5740039
https://www.diyaudio.com/community/threads/i-cant-hear-the-difference-in-filters.335711/post-5740078
There have been similar expressions of inaudibility in other threads, the problem is figuring the right search terms for the exact wording. Do recalling a claim about the filters being only for marketing, that they don't sound different. Don't know how much time I want to spend searching for it though.
Conventional wisdom has it that anything above 20 kHz is always inaudible to humans. When you believe that and either neglect or deal with a couple of second-order effects, such as the possible impact on intersample overshoots and possible effects due to intermodulation between components above 20 kHz, then it follows logically that the filter response above 20 kHz can't have any effect on human listeners, and that any sample rate from 44.1 kHz to infinity sounds the same.
Conversely, if there is a difference audible to humans, then it is either one of those second-order effects or due to components above 20 kHz being audible under certain conditions.
Conversely, if there is a difference audible to humans, then it is either one of those second-order effects or due to components above 20 kHz being audible under certain conditions.
IME non-linear phase filters affect transient response in the pass band (e.g. crest factor). Slow rolloff filters may attenuate HF in the passband. Differences in alias suppression may also be a factor. For whatever reasons, different dac filters sound different to at least some people or the dac manufacturers wouldn't include them (unless maybe someone wants to believe its all snake oil).
I suspect what really happens is some of these guys listen, don't hear a difference, then post-hoc construct a technical theory to explain why nobody else should hear a difference.
I suspect what really happens is some of these guys listen, don't hear a difference, then post-hoc construct a technical theory to explain why nobody else should hear a difference.
Last edited:
A long time ago I had a piece of junk known as CD723. It sounded bad. I recorded its digital output with the SPDIF input of my soundcard. I noticed it had a buggy digital volume control which when set to maximum applied an undithered gain of 0.999...something instead of 1, screwing with the LSBs.
The difference between the original file and the file recorded digitally from the CD723 was not difficult to hear in ABX.
"Oh but you'll never hear it, it's in the noise floor" ->
The difference between the original file and the file recorded digitally from the CD723 was not difficult to hear in ABX.
"Oh but you'll never hear it, it's in the noise floor" ->
Good point, I forgot that filters with a poor phase response are in fashion nowadays. That probably shows my age; back in the 1980's, phase linearity was a selling point and minimum-phase behaviour was not, nowadays it appears to be the other way around.
A long time ago I had a piece of junk known as CD723. It sounded bad. I recorded its digital output with the SPDIF input of my soundcard. I noticed it had a buggy digital volume control which when set to maximum applied an undithered gain of 0.999...something instead of 1, screwing with the LSBs.
The difference between the original file and the file recorded digitally from the CD723 was not difficult to hear in ABX.
"Oh but you'll never hear it, it's in the noise floor" ->
You wouldn't believe how many digital designers involved in signal processing still don't even know basic dither theory. It's as fundamental to quantization as the sampling theorem is to sampling, there is even mathematically a very close correspondence.
Filter passband ripple is another second-order effect, by the way. For a phase-linear filter, it more or less corresponds to a pre- and a post-echo.
In fact we did a little blind test on that last year on this forum, see https://www.diyaudio.com/community/...makes-nos-sound-different.371931/post-6761338 The only test person who wasn't sure he heard any difference at all was the only one who scored better than chance.
In fact we did a little blind test on that last year on this forum, see https://www.diyaudio.com/community/...makes-nos-sound-different.371931/post-6761338 The only test person who wasn't sure he heard any difference at all was the only one who scored better than chance.
Returning to my original post, after the summer break, I delved into some aspects better, defining a new statistical indicator to quantify the distortions on transients. At the usual link all the details. Soon I should add more considerations on higher order distortions, feedback effects, and more.