Serious Question: How can DAC's have a SOUND SIGNATURE if they measure as transparent? Are that many confused?

[raif71]

I have tested 12 non-eccentrically engineered DACs level matched and blind from £1000 to &14,000 and I could not hear any difference of note, in as much as if there did seem to be a difference I could not repeatedly recognise it or tell one from the other, so I am content they all sound the same to me, or at least so close to the same to be inconsequential.
I can't speak for anybody else, obviously.
Having written that, I have also tested my hearing in my room at home and find if the music is playing 60dB quieter than my average level I can only just hear it.
On that basis I do not find it credible that I would be able to hear distortion at a level of -60dB whilst music was actually playing since I can only just hear that level of sound when it is not playing. Since my listening level allows 20dB for peaks I also am content that if the SNR or SINAD is better than -80dB then it would be ridiculous to imagine I would be able to hear noise less than that.
So based on logical deduction (IMHO) anything with distortion better than -60dB and noise less than -80db would obviously sound the same to me here in my room.
There is almost no electronics available on the market which exceeds these limits so, for me, I am content that any difference I would hear in sighted tests non-level matched I will have imagined.
That is my logic for me tested on me by me

You're the expert and I take your word for it

 

[Frank Dernie]

You're the expert and I take your word for it

I am the expert on testing me in my room only.
It isn't difficult so anybody can do it. I recommend it it.
It is made easier for me by having a volume control calibrated in dB. Just try it, turn down the music until you can only just hear it then try to imagine being able to hear either distortion or noise at that level whilst the music is playing at normal levels. It is most enlightening.
If stuff needed to be better than this there would not be a person on earth who would ever consider listening to an LP as acceptable, never mind the numerous enthusiasts considering it better, since they have never been as good as this!

 

[nhatlam96]

Here is a list from HeadFi of blind tests over time. I could not access the Stereomojo link, but it may be archived elsewhere via the Wayback Machine.

https://www.head-fi.org/threads/testing-audiophile-claims-and-myths.486598/

the amp and dac tests were tested with speakers and are quite old. Im searching for new tests with headphones + dacs + headphone amps. I find it odd that I can't find a single article, which proofs that measurements are everything. They only say that double blind test and volume matching is important, but nobody ever did an experiment or test and shared their result in a full article.

 

[Julf]

Actually a lot of people think like us but the experts here still maintain that when doing blind tests coupled with volume level matching, you can't really tell the difference!!

Prove us wrong...

 

[Julf]

the amp and dac tests were tested with speakers and are quite old. Im searching for new tests with headphones + dacs + headphone amps. I find it odd that I can't find a single article, which proofs that measurements are everything. They only say that double blind test and volume matching is important, but nobody ever did an experiment or test and shared their result in a full article.

Do you have access to the AES library?

 

[nhatlam96]

Do you have access to the AES library?

If you mean this, then no.

 

[Julf]

It is also important to distinguish between blind tests based on "which one do you prefer, A or B" and ABX that tests if you can hear a difference.

 

[Julf]

If you mean this, then no.

Indeed. Unfortunately most of it is only accessible to members.

 

[mansr]

I, personally, have zero interest in the DAC measurements on this site since the measurements can't form a basis for my choice.

If I were considering a new DAC, I'd check the measurements here to make sure it wasn't doing something stupid. Beyond that basic sanity check, I agree that they don't really say much.

 

[nhatlam96]

Amir tested, reviewed and listened to countless dacs and amps and he repeatedly said that there is no audible difference when properly engineered. I just would like to see him maybe make an article about it, where he would do a proper blind test between multiple amps and dacs with full documentation and some pictures. I think this would fix this topic once and for all.

 

[Chippyboy]

Hi,

Really surprised this thread is still going on.

But for me, the answer is that there can be a difference between sound output of different DAC's (as a complete device : from digital input to analog RCA/XLR ouput).
If there was no difference then there would be no point in measuring.

Here are three simple "facts" that tells me that YES, indeed there can be differences in sound produced by a DAC (as a complete device : from digital input to analog RC/XLR output) :

- A LOT of mesurements show that the ESS SAbre DAC have a bump on a given frequency range when the other brands do not show that ;
- some manufacturers implement a "filter" on their DAC and allows the user to change between filters. Amir has measured differences between those filters. So yes, according to how the output is then "filtered", the DAC (as a complete device : from digital input to analog RC/XLR output) can sound differently ;
- almost ALL of the DAC's (as a complete device : from digital input to analog RC/XLR output) tested by Amir have different measurements that goes from pretty bad to superbly good... sometimes using the same DAC chip/SoC (so the sae DAC, but as a chip, not the complete device) .

And may I add a really stupid question ?
If a DAC (as a complete device : from digital input to analog RC/XLR output) can't have a sound of it's own, then there would be no point in testing the devices I think...

Regards.

P.S. : not trying to troll. Just looking for explanations...

The facts you quote do not demonstrate that audible differences exist, merely that technical differences exist. Your post proves that it is *possible* that audible differences may exist, but you provide no evidence that this is actually true. Merely that it is a possibility, which may actually prove to be untrue.

 

[BDWoody]

Amir tested, reviewed and listened to countless dacs and amps and he repeatedly said that there is no audible difference when properly engineered. I just would like to see him maybe make an article about it, where he would do a proper blind test between multiple amps and dacs with full documentation and some pictures. I think this would fix this topic once and for all.

It wouldn't prove anything to people who don't 'believe' in controlled testing.

It would be easier to just get a single example to disprove the assumption...

Despite offers of cash prizes and world renown, not a single person has shown that they can pick one competent DAC over another listening to music, in a normal room, using comparable filters, under controlled circumstances.

How do they get around it? There must be something wrong with controlled testing, because 'I know that I hear!'

It goes round and round the toilet bowl, but never flushes down...

 

[nhatlam96]

It wouldn't prove anything to people who don't 'believe' in controlled testing.

It would be easier to just get a single example to disprove the assumption...

Despite offers of cash prizes and world renown, not a single person has shown that they can pick one competent DAC over another listening to music, in a normal room, using comparable filters, under controlled circumstances.

How do they get around it? There must be something wrong with controlled testing, because 'I know that I hear!'

It goes round and round the toilet bowl, but never flushes down...

If the test is done properly and fully documented, then there should be no flaws to critize on. So non believers of controlled tests have no other choice than to believe in amirs test

 

[Chippyboy]

What's missing from this debate is what does "good" sound like. There would surely be few people (if any) who would claim that no tube amplifiers in the world, at all, sound good. That no hifi system with a tube amp produces a pleasing sound. That all decent transistor amps sound better than all decent tube amps.

And yet it is true that the vast majority of transistor amps - properly designed ones at least - measure significantly better than their tube-based counterparts. So how can it be that these - relatively speaking - terribly performing tube amps can sound great. Fantastic, even. If good sound was defined by low distortion and other such technical parameters, then transistor amps would sound better than tube amps, and yet clearly many times they do not.

What does this tell us? Well, a number of things. First that the human ear/brain is not a particularly good measurement instrument on any objective level. We are insensitive to all kinds of distortions which are easily measureable. And also, that there is no hard and fast correlation between levels of distortion and perceived sound quality. Or at least if there is one, no-one has demonstrated what that correlation is.

The conclusion therefore is that endlessly chasing incremental technical improvement in some parameter or other - SINAD, for example, down to 0.00001% from 0.0001% etc - is a completely pointless exercise. First, because we have no remote possibility of being able to hear such a change and second, because even if we could, there is no evidence it would actually sound "better".

We are after all, in search of "good" sound aren't we? So we can listen to records, enjoy the music and smile. Not sit there thinking "this sounds bloody awful but I will put up with it because I know the DAC distortion is only 0.00001%".

 

[Frank Dernie]

If the test is done properly and fully documented, then there should be no flaws to critize on. So non believers of controlled tests have no other choice than to believe in amirs test

There is no point.
Who would organise it and who would participate?
Would it change the mind of the disbelievers?
It is easy to do a small test for oneself and, despite substantial prizes having been offered for anybody who can tell the difference in blind tests, both amps and DACs, over many, many years not one single person has ever done so.
The nearest we get is for people to claim the test is invalid, which is lame, at best.

In the end the disbelievers won't believe and, incredibly to me, people don't do simple tests for themselves.
I know the majority of people are not technically minded but checking it out for oneself ain't hard.

 

[BDWoody]

If the test is done properly and fully documented, then there should be no flaws to critize on. So non believers of controlled tests have no other choice than to believe in amirs test

You underestimate the ability of people to hold onto ignorance.

If they don't believe in the premise of blind testing, they won't care about the results.

We've got a guy shilling his snake oil audio cables on another thread. When asked for blind test results, he said being tested prevents you from being in the alpha brain wave state required in order to hear the difference. Ummm...ok.

How do you prove anything to him, or anyone who buys that crap?

 

[Frank Dernie]

What's missing from this debate is what does "good" sound like. There would surely be few people (if any) who would claim that no tube amplifiers in the world, at all, sound good. That no hifi system with a tube amp produces a pleasing sound. That all decent transistor amps sound better than all decent tube amps.

And yet it is true that the vast majority of transistor amps - properly designed ones at least - measure significantly better than their tube-based counterparts. So how can it be that these - relatively speaking - terribly performing tube amps can sound great. Fantastic, even.

What does this tell us? Well, a number of things. First that the human ear/brain is not a particularly good measurement instrument on any objective level. We are insensitive to all kinds of distortions which are easily measureable. And also, that there is no hard and fast correlation between levels of distortion and perceived sound quality. Or at least if there is one, no-one has demonstrated what that correlation is.

The conclusion therefore is that endlessly chasing incremental technical improvement in some parameter or other - SINAD, for example, down to 0.00001% from 0.0001% etc - is a completely pointless exercise. First, because we have no remote possibility of being able to hear such a change and second, because even if we could, there is no evidence it would actually sound "better".

#2,003
Think it through logically for yourself.
For me there are 2 "levels", if you like, one the point at which shortcomings become audible and secondly the point at which audible shortcomings may (or may not) become pleasantly euphonic.
2% distortion and SINAD of -70dB is fairly good for an LP yet many enthusiasts will claim them better than digital (where better indicates their preference, obviously, not fact).
Most amps have better than 0.1% (-60dB) distortion and I personally can only just hear a faint sound if I turn the volume down by 60dB, so the likelihood of me being able to hear anything at that level while the music was playing, distortion or otherwise, is faint to sod all.

It is so easy to get an idea of what you can hear for yourself I am surprised people still want to be told by somebody else.

 

[Plcamp]

From another post regarding the audibility of different amplifiers:

I liked this. I have always thought a properly in-spec operating amp, driving a load within its current capability, will sound the same as any other amp likewise driving within its capability.

As far as DACS go, I will say that my new e30 DAC sounds cleaner and more detailed than the DAC built into my Sony Bravia tv, but (aside from idle background noise level) isn’t noticeably different to other DACs I own like the one in my Yamaha RX-A3030 or my minidsp2x4hd.

 

[Julf]

What's missing from this debate is what does "good" sound like.

There are two major schools of thought. One is that a system should sound "pleasing", "musical" or whatever, as in "coloring the music in the way I like". To me that is like saying "I want extra salt and ketchup on all my food, even in a Michelin star restaurant" or "I like everything shaded a bit pink, so I wear tinted glasses when I go to art exhibitions".

The other school believes "good" is defined as "accurate", as in reproducing the original signal with as little audible distortion, coloration or added noise as possible.

 

[Julf]

So non believers of controlled tests have no other choice than to believe in amirs test

Time and time again they have proven to us that they are perfectly capable of falling back to "double-blind tests are flawed" and other excuses to stick to their beliefs.