I asked over there about its origin and was told that it came from some long-ago writings of Amir. Nobody provided a link though, despite me expressing interest in reading it.
I'll start by saying that I don't want to get involved in your age-old discussion of the two of you, but I genuinely was wondering the following: IF it were true that what each person hears is different from what everyone else hears, why in the world would you suggest a testing methodology?
Don't you think this is illogical?
Maybe you'll answer that it depends on the quantity (in absolute value or in percentage?) of participants and the related scores they assigned.
So, let's say that there are a hundred participants (which is an absurdly and intentionally high number, I guess) and that 95% (!) of them report the same type of experience, then that same experience would also apply to you too?
You could never be sure, so what?
Assuming that someone could perform a controlled and restrictive test following exactly the rules that you (or whoever you want) prescribe, that you were present at that test without participating (or even participating, if you want) and that you approved the validity of the various sessions, what would change in your stubborn and apparently very convinced position?
Doesn't that seem illogical to you?
IF you have a hearing system that no one else in the world has the same as you, even if the percentage were 95%, you would have no certainty that the results heard by 95% of the participants would also be valid for you.
At the end of the day, and without any provocation, but with a serious question, what the exact purpose is for which you recommend doing controlled tests?
Thanks.
Last edited:
Did you read what I wrote? Markw4 said that "all it takes is the existence of two dacs that measure well according to Amir's measurements, yet sound even slightly different to falsify his hypothesis". I'm quite sure his or your subjective opinion will convince no one. That is why I recommended controlled test.
Brief excerpt from the paper: “ Digitally synthesized sinusoids were used as signals. The signal was generated by a D/A con- verter (EDIROL UA-1000) at a sampling rate of 96 kHz and 16 bit resolution. The signals at 16 kHz and above were presented by a super-tweeter (PIONEER PT-R100) via a high-pass filter (PIONEER DN-100). A loudspeaker (DENON SC-A33) was used for the signals at 250 Hz…”
Perhaps, make the effort to read a paper before verbally ripping it?
Last edited:
I think it is Amir who needs to make a serious effort disprove the alternate hypothesis, including following ITU recommendations. OTOH, if I, or someone like me, did a serious controlled study they would simply reject it as non-credible. Maybe at best a one-off outlier. And I would be out the time, effort, and monetary cost with little or nothing gained.
As one member already told you at ASR it is just their current working hypothesis and I would assume they have no need to change that unless evidence to the contrary is shown. Besides AFAIK ASR is partly funded by fees from manufacturers having their devices tested. So their claim that well measuring DACs are transparent (hence making measurements futile) is potentially hurting their cashflow. And if somebody would provide solid evidence to the contrary they would just look foolish by sticking to their guns.
Who is going to pay for the dacs to be compared?
Also, what equipment is going to be used to verify the test dacs measurements? What I have access to now is okay for most stuff but definitely not SOA. More cost and complexity involved for that.
Who is going to flip the switches, who is going to be a qualified independent 3rd party observer, etc.?
BTW, I know it is a hypothesis, but its stated like a fact in the reviews that a particular dac is "audibly transparent." Its not, "we think this dac is probably audibly transparent." Thus people like the OP take it as fact.
EDIT: There is also the issue of how many trials would be needed to be convincing. @ThorstenL advice was to do only a few trials each day, as fatigue can set in quickly. Who is going to pay for everyone involved in the project to spend days preparing, practicing, and conducting daily listening tests for however many days? Who is going to pay for people's time off work?
Also, what equipment is going to be used to verify the test dacs measurements? What I have access to now is okay for most stuff but definitely not SOA. More cost and complexity involved for that.
Who is going to flip the switches, who is going to be a qualified independent 3rd party observer, etc.?
BTW, I know it is a hypothesis, but its stated like a fact in the reviews that a particular dac is "audibly transparent." Its not, "we think this dac is probably audibly transparent." Thus people like the OP take it as fact.
EDIT: There is also the issue of how many trials would be needed to be convincing. @ThorstenL advice was to do only a few trials each day, as fatigue can set in quickly. Who is going to pay for everyone involved in the project to spend days preparing, practicing, and conducting daily listening tests for however many days? Who is going to pay for people's time off work?
Last edited:
No one. Maybe you take this all a tad too seriously. Just a thought.
Regarding taking things too seriously, I think its reasonable to point out something about human thinking. Its easy to say, why don't you do a controlled listening test? Its easy to say because it seems easy enough to do. Its just one example of how our human biological minds often tend not consider ambiguous details. Its why a lot of people who have to estimate how much time it will take to do a job, often underestimate (often by a lot). Long ago, when I used to do occasional estimates of how much it would cost to install something like a church PA system, I would figure out how long it would take if we just did the work, then I would double it as a contingency allowance for things that come up as the work is being done. Usually it was still an underestimate.
Last edited:
And then you brought your perfect measuring and sounding DAC to a listening test. Tension, excitement.
There you are, it is clear to you that your DAC is a winner. The competition is meagre. Chinese stuff with all their flaws and some former british brands ready made DACs. Even one tube DAC. Blind, double blind, in reverse and upside down: your DAC is clearly the best. Only the deaf would not hear that. Your DAC is audiophile truth. You wrote papers and can show those too. Easy peasy.
The verdict of the jury is that the SMSL D6s is the winner.
There you are, it is clear to you that your DAC is a winner. The competition is meagre. Chinese stuff with all their flaws and some former british brands ready made DACs. Even one tube DAC. Blind, double blind, in reverse and upside down: your DAC is clearly the best. Only the deaf would not hear that. Your DAC is audiophile truth. You wrote papers and can show those too. Easy peasy.
The verdict of the jury is that the SMSL D6s is the winner.
No. The listening test has to come here. I have the speakers and the room to judge soundstage. That's probably how to tell the dacs apart blind, because its probably not going to be by "distortion."
Ah - your DAC is tailor made for your room!/system?
Well, if its listening optimised, thats whats happens 😉
//
Well, if its listening optimised, thats whats happens 😉
//
Until such time as someone, anyone comes with a valid test showing a pair of decent dacs that manage say -110,dB thd + n and similar for imd can be identified by listeners I'll roll with the, not worth worrying about theme I'm comfortable with.
Benchmark DAC-3 and Topping D90 were compared here. They both measured pretty well over at ASR. Easy to tell apart by soundstage. The D90 had another quirk which some people didn't seem to have trouble hearing, for which the only measurement I have seen that might explain it is one they don't do over at ASR.
A lot of people are deaf to such things though. One easy exercise if you still have enough hearing left is to listen to the sound of a room. Clap your hands and listen to the how the room echos and decays. Try to listen that way while someone is quietly talking and you are trying understand them. All of a sudden it gets harder to understand what they are saying as you hear the room echoing and decaying with each word. Fortunately, the brain normally filters all that low level room sound out as noise, so you only hear the words being spoken.
Most people have no idea how much low level sound their brains are discarding as unwanted noise. IMHO, that includes some sounds produced by a music reproduction system.
A lot of people are deaf to such things though. One easy exercise if you still have enough hearing left is to listen to the sound of a room. Clap your hands and listen to the how the room echos and decays. Try to listen that way while someone is quietly talking and you are trying understand them. All of a sudden it gets harder to understand what they are saying as you hear the room echoing and decaying with each word. Fortunately, the brain normally filters all that low level room sound out as noise, so you only hear the words being spoken.
Most people have no idea how much low level sound their brains are discarding as unwanted noise. IMHO, that includes some sounds produced by a music reproduction system.
Last edited:
As Mark4 explained, these need to be properly funded otherwise all will be just anecdotal evidence …
For your information, I'm an exception to the rule, but in general, autistic people are supposed to hear more than neurotypical people because their brains discard less information.
Interesting. Yet you don't own an Amanero for listening to some of what you design? Is that related to how you hear or something else?
I assume you mean the FPGA configuration file I posted a couple of days ago (which appears to have a bug in the FIFO control part). I wrote that because you and bohrok2610 both seemed to be interested in it and because it was a nice puzzle - actually still is because of the bug. I never had any intention to listen to the result myself. I do listen to a DAC that uses essentially the same sigma-delta algorithm, though.
- Status
- Not open for further replies.