Is there really no audible difference between different DACs?

[pma]

i asked why make measurements on dacs if they are audibly transparent

Because you/we have to know IF they are and there is no other way than measurements. It is a necessary condition. You know, there are necessary and sufficient conditions. It may happen that the set of measurement is incomplete and something is overlooked. Like for example here
https://www.audiosciencereview.com/...easurements.10129/#lg=attachment40622&slide=0

You would not see it in a THD 0dBFS spectrum, , you do not see it in THD calculation, because disturbing lines are not harmonic. You would not suppose that it is happening. It may be that different parts/units may need different measurements. The set is never complete.

 

[VintageFlanker]

@CelticForest and @WreckTangle are one and the same person , it's a guy who signs up to the forum ( and many others ) under multiple different names .

Both now banned .

can you please stop being so pathetic and go elsewhere mate! Have you nothing else in your life!

Hum...

Oh my God... you have got to be kidding me.. how can an OPA2134 ever 'conduct' the special sound from vinyl and tubes.
Opamps block the 'magic' ... you ought to know that by now .
NO opamp ever sounds good. It MUST be discrete !
And are those dual gang potmeters ? ... not stepped goldplated silverlined rhodium contacts with special 0.001% tolerance, induction and capacitance free HD-moviefilm resistors ... blasphemy !
That opamp will impart it's horrid sound signatureon the supplied signal and that is the reason why everything sounds the same when AB testing this way.

@solderdude and @maty are one and the same person.

 

[Berwhale]

Well it was entertaining while it lasted...

 

[ernestcarl]

DT880 (250 Ohm which has a slightly higher treble peak than the 600 Ohm)

one ply of toilet paper lowers the treble a bit:

That is almost exactly what I have done -- only I used a very thin piece of some old, weathered white t-shirt -- black would've worked the same, but I'm afraid it might impart too much of a dark signature.

 

[Deleted member 3116]

So you are saying the differences were above -70dB ?
And around 1kHz while listening to complex music and those audible variances weren't masked by the music at all?
And an external power supply changes that ?
What DAC has such poor PSRR and is that demonstrable (recording of output or measurements could be called demonstrable) ?
Did you keep this DAC and if so can you show the influence to be there.

It's this measurement actually https://www.audiosciencereview.com/...nts-of-topping-d50-dac.2403/page-2#post-68200

As I said I dont believe it has anything to do with this noise, this is merely the only difference seen with static measurements (of a different DAC and power supplies .
How do we know the DAC doesnt measure differently when playing music compared a simple sine wave was what I asked?

 

[RayDunzl]

How do we know the DAC doesnt measure differently when playing music compared a simple sine wave was what I asked?

What's simple about a sine wave?

 

[Berwhale]

What's simple about a sine wave?

It's deterministic rather than stochastic?

 

[Deleted member 3116]

I have encountered supposedly inaudible or unmeasurable differences before, another example was blind testing between ODAC and D50 ( a different D50 to my current one infact, display failed after a couple weeks on first and I got sent a replacement). they were discernable in blind tests, D50 was favoured as sounding best, but I can never guaranteed there wasnt some measureable difference, since im relying on 2nd hand measurements.
I plan to buy soekris DAC at some point to further test this, soekris is one of the most affordable R2R Dac that measures well enough to be audibly transparent.
If a DAC has measurably audible noise or distortion then its junk, I would like to trust measurements represent everything we can hear and/or capture all possible sources of distortion for the sake of convenience, cost and peace of mind but there is mounting personal evidence that this may not be case,.
Blindly believing measurements tell us everything is just as harmful as believing they dont, but I have seen many here have done tests and found no audible differences themselves so their objective stance is fully justified.
If all the ones raving about actually hearing inaudible differences between equipment would do some controlled tests and provide proof that might actually get us somewhere... I dont doubt a large portion would fail at producing any proof.

 

[maty]

I have encountered supposedly inaudible or unmeasurable differences before, another example was blind testing between ODAC and D50 ( a different D50 to my current one infact, display failed after a couple weeks on first and I got sent a replacement). they were discernable in blind tests, D50 was favoured as sounding best, but I can never guaranteed there wasnt some measureable difference, since im relying on 2nd hand measurements...

With some cheap tweaks the old ODAC sounds much better. Its biggest weakness is that it has insufficient noise filtering and...

 

[Deleted member 3116]

With some cheap tweaks the old ODAC sounds much better. Its biggest weakness is that it has insufficient noise filtering and...

old as in first version? mine was rev b with burr brown chip , tested on here.

 

[maty]

Yes, the first version. I followed its development from the beginning. It took me a few months to buy it just in case, as is my custom with new products, whatever they are.

https://nauscopio.wordpress.com/201...ro-audiomalaga-de-audio-planet/#comment-13442

 

[LTig]

Well it was entertaining while it lasted...

No, his writing style sucked.

 

[solderdude]

How do we know the DAC doesnt measure differently when playing music compared a simple sine wave was what I asked?

The closest thing we have right now to test DACs to see how they react to a bunch of different tones is the multitone test.
There are tones over the entire range which are not harmonically related (so they can mask each other).

Another option is to use @pkane nulling software where you can use real music. Just record what has been played and then null it.
Problem is nulling 2 different files exactly is a bit difficult so you will always 'measure' some differences. Also phase differences will show up as amplitude differences where phase differences could be benign and not audible yet a null would show differences in FR which thus aren't really there.

Music is just a bunch of sinewaves anyway that differ in amplitude and frequency and thus answers to the same laws as a sine wave.

Add to that a DAC simply gets a digital word and then has to output the voltage belonging to that value. very quickly followed by the next value.
There is no analysis going on. The DAC just does what it is asked.

Of course the digital filters used before the actual conversion process do take following and preceding (in most proper filters) into account.
Not to 'look for how something should sound' but to interpolate the to be expected voltage values between sample values it receives more accurately then when it would only 'draw' a straight line between sample values.
This is a mathematical thing and should comply to the sampling theorem.
Some manufacturers deviate from it on purpose though.

Those filters thus also do not analyze or change sound... they simply predict more accurately the inbetween values based on current, following and previous sample values so they can feed all of those calculated (not guessed but mathematically calculated) to the DAC chip which then outputs these analog values (or in case of SD an averaged approximation) at a very high speed so all that is needed afterwards is a very simple low pass filter at a very high frequency.

 

[pma]

IME multitone often does not show issues and is too easy to pass. Like here:

https://www.audiosciencereview.com/...dac-vs-cd-player-shootout-measurements.10129/

 

[Berwhale]

No, his writing style sucked.

It wasn't the writing style that entertained me.

 

[solderdude]

Here is a multitone of a DAC that shows it has quite a lot of issues

and from another high end DAC

and here one from a PC soundcard:

When a DAC passes this test well it means distortion is very low and little issues can be expected when it concerns distortion.
Here is what a good one looks like:

So the multitone is not easy to pass and indicative of certain issues, but of course not the sole important thing just like SINAD is not the sole number to judge quality on.

 

[pma]

When a DAC passes this test well it means distortion is very low and little issues can be expected when it concerns distortion.
Here is what a good one looks like:

So the multitone is not easy to pass and indicative of certain issues, but of course not the sole important thing just like SINAD is not the sole number to judge quality on.

No it does not. Multitone did not show level dependent non-harmonic spuriae that occured at 1kHz/-60dBFS sine test and are very audible. They are audible with very low volume passages of classical music, e.g.
Low frequency distortion you would see with low frequency sine. The case is that one has to tune wide scale of levels, frequencies and combinations of frequencies to get view wide enough. A standard wideband multitone with fixed crest factor does not tell the story.

1kHz/-60dBFS. This behaviour was not disclosed when tested with a multitone. Multitone result was close to perfect. These non-harmonic, signal related spuriae are perfectly audible.

Another DAC without issues:

This is a correct behaviour and no spuriae are audible.

Multitone test of both DACs was almost identical.

No spuriae around 7.5kHz. Issue not disclosed with the multitone test.

 

[RayDunzl]

It's deterministic rather than stochastic?

I have to look:

Wiki:
"In deterministic models, the output of the model is fully determined by the parameter values and the initial conditions initial conditions.

Stochastic models possess some inherent randomness. The same set of parameter values and initial conditions will lead to an ensemble of different outputs."

Hmm...

 

[raif71]

@CelticForest and @WreckTangle are one and the same person , it's a guy who signs up to the forum ( and many others ) under multiple different names .

Both now banned .

can you please stop being so pathetic and go elsewhere mate! Have you nothing else in your life!

Aw man, cmon. Just when I think I'm not alone...this happens. Is this ban permanent? The other guy, @protoss at one point banned (I think) , is he still around?

 

[VintageFlanker]

Aw man, cmon. Just when I think I'm not alone...this happens. Is this ban permanent? The other guy, @protoss at one point banned (I think) , is he still around?

Hope he's not. Heam, do you miss that guy? :

Sorry i offended your cult leader.

Your leader needs to be checked. And I will be questioning him.

You just watch over in the corner cheerleader.

Such a weak-ling you are. Arent you little duck-ling.

So what if I leave. The kool-aid party continues, drinking the host blood and everyone circle jerks each other to infinity. Never knew such weaklings are found at a science base forum.