DAC types and their sonic signature

[zalive]

I think this above speaks volumes.

Gents, no point in arguing with dogma. Leave him to it.

Dogma?
This was not a bright comment...

 

[March Audio]

Dogma?
This was not a bright comment...

Its entirely accurate. Your commentary so far is entirely typical of the "subjective audiophile" who refuses to recognise the proven science which contradicts their entrenched beliefs.

 

[pkane]

However I think one needs some experience in blind testing before he can actually get to the point of having really equal mind state regardless of sighted or blind. It's not something ensured when it's relatively new. It's a different situation and with the different situation you can easily expect a different mind state than usual.

That is something that's absolutely true. Training is necessary. A different mind state occurs when switching from sighted to blind precisely because there are so many false differences that are normally experienced when sighted. These 'obvious' differences while sighted disappear when blind, resulting in confusion. That confusion passes with some experience. And then you can learn to hear the real differences.

What you say is completely true. But you can't escape the human nature, which is prone to making conclusions and creating 'facts'. If no difference is confirmed through the blind test, human nature will take this as 'there is no audible difference'.

If I can't detect differences in a blind test over multiple attempts, I am comfortable making the conclusion that I can't hear the differences. I wouldn't reach that conclusion for anyone else but me. If I hear differences when sighted, I can make no such conclusions.

Something else escapes here. Blind tests do not only include hearing a difference. To audiophile, it's also about a preference. So, if audiophile claims he heard a difference in a blind test and he preferred device A in front of device B, would you find value in his judgement of preference, since it can be quite different to your preference? Or would you just find worth in hearing a difference but completely dismiss a personal choice 'which sounds better'?

First thing's first: in order for a preference to exist and be valid, there must be an audible difference. If there's no audible difference, the preference is based on something other than sound.

Once a difference is established (in a proper blind test), a preference can certainly be expressed. This is useful information, especially when done in large groups. Some common preferences can then be discovered. A single individual expressing their preference in a blind test is mildly interesting, if I have some way to compare my own tastes to his. Otherwise, it's just someone else's preference. I don't particularly care if you like Coke or Pepsi, and I don't particularly care if you like digital or vinyl -- means nothing to me.

 

[Xulonn]

I agree, but many audiophiles care for smaller differences too, not just the huge ones.

Indeed, often for differences so small that they cannot actually be heard, only imagined...

 

[Blumlein 88]

The yt link I put, I wanted to hear whether other people like the sound which I like to.

I like the sound from the yt a lot. I presume there is really good mic job beneath and obviously a good sound recording too. But all this won't ensure good sound if the system beneath have a good sound.

And that's the reason why I put it. I really wanted to here whether others like as well what I like. Obviously my impressions were sighted but I wanted blind listening impressions from the others to say whether they like it too.

But to me it looks like the others are for some reason afraid to write down their blind impressions of this sound and stand behind them. I don't understand quite why because as you said, there are no valid conclusions from this and there can be no personal judgement too. It was meant to be the only thing which was said, solely about a sound preference.

Sorry, I didn't get back on your linked video as I asked for the link a second time. Unfortunately I clicked on the link and saw the video meaning I didn't hear it blind.

I think the sound you have there is good, enjoyable, a bit smoothed over in a pleasant way, but commensurate with tubes being involved. Would I have thought that blind? I think so, but I bungled hearing it blind first.

 

[Blumlein 88]

Furthermore, even blind listening impressions are not reliable. They might be more reliable than sighted impressions because of avoiding bias induced by knowing what's in system, but they're not completely reliable because the nature of listening is subjective. There's no even guarantee that blind listening should be more reliable than sighted listening because of a different mind state when blind listening and sighted listening. Especially when blind test is done in a situation when a person might fear he/she will be judged by the results and a personal reputation will be in the line.

What about blind results when listeners didn't know they were being part of a blind test?

 

[Blumlein 88]

I didn't want to argue about the preference.
I was curious about it.
And it's not even important if to me if it sounds different in the nature. I was currious about impression of how that specific audio sounds to the others.

And yes, in case when it's a decent mic'd audio, my opinion is it will say something about the sound. It won't say it all, it can't give a completely accurate impression as when listening it live, but it can still present some of its quality. Make an audio of a lousy sounding system, you won't be able to make it sound nice. Or at least not unless you DSP it or similar.

This is harder than it sounds. Making system recordings. I've a few I could share (or I think I still have them). You'll have a very different idea of the overall sound depending upon where I put the microphone. I even have an idea of where I can place them and give a reasonably close to how it sounded live result. It could tell you something, but it is rather limited.

You have the same problem with sighted listening. How useful is it? Not 0%. Not 100%. On basic tonal balance, on whether something is clipping or underpowered, on some aspects like that maybe 30%. Though some particular conditions can lower than to maybe only 10%. Like, "I hate box speakers, the boxes always sound colored." What are the chances that person will hear a box speaker and not give a lowered account of its real virtues or lack thereof? Or, "I hate horns speakers, they always have that honky colored sound"?

 

[zalive]

If I can't detect differences in a blind test over multiple attempts, I am comfortable making the conclusion that I can't hear the differences. I wouldn't reach that conclusion for anyone else but me. If I hear differences when sighted, I can make no such conclusions.

Do you ever reach inconclusive % after series of blind test, and what do you do if this happens?
Make twice as much repetitions?
What if no polarization towards either 50% or a high % should emerge?
Say, you do 30 repetitions and have a 70%...what would you conclude?

 

[Blumlein 88]

Indeed, often for differences so small that they cannot actually be heard, only imagined...

Yes, this is a good point. If someone adds a subwoofer and say the bass is more abundant I doubt we need a blind test. If they added a new power cable to their amp and tell us it was so much better in the bass it was like adding a subwoofer, we'll need a measurement or blind test to give this any credence.

 

[Blumlein 88]

Do you ever reach inconclusive % after series of blind test, and what do you do if this happens?
Make twice as much repetitions?
What if no polarization towards either 50% or a high % should emerge?
Say, you do 30 repetitions and have a 70%...what would you conclude?

That there is only a 5% probability that the result was from random guessing. If you repeat it and get 70% again, chances are less than 1% it is a lucky result instead of a perceived difference.

 

[zalive]

Sorry, I didn't get back on your linked video as I asked for the link a second time. Unfortunately I clicked on the link and saw the video meaning I didn't hear it blind.

I think the sound you have there is good, enjoyable, a bit smoothed over in a pleasant way, but commensurate with tubes being involved. Would I have thought that blind? I think so, but I bungled hearing it blind first.

Thanks, a honest reply. I should have warned to ensure a black screen after clicking or at least stress out more it's a video. crossed my mind...
Anyway, your sighted impressions are quite similar to my sighted impressions. The difference is I didn't find the sound smoothed but rather realistic. But possibly you're right. But if so it's a very fine level of smoothing in my view...

I came to that vid because I was interested in that particular DAC, it was an option for me.

 

[zalive]

What about blind results when listeners didn't know they were being part of a blind test?

You should describe the scenario, I'm not sure what's your idea.

 

[zalive]

That there is only a 5% probability that the result was from random guessing. If you repeat it and get 70% again, chances are less than 1% it is a lucky result instead of a perceived difference.

I agree completely (without checking the probability, I trust you calculated it right).
But I get the notion some folks here acknowledge only results pretty close to 100%, if not 100%.

 

[Blumlein 88]

You should describe the scenario, I'm not sure what's your idea.

Well one I can think of is unintentionally self inflicted. A few audiophile friends gathered trying out a new DAC one of us had. Took over and hooked into a good system with both new and old DAC going thru a preamp so we could switch. Listened to the owner's DAC. Swapped to the new one. During the swap over during discussion and checking levels were about the same we then listened to the new DAC. Discussion about how it sounded different, whether or not it was better, the consensus was a pretty good improvement overall. Then it was noticed during the level checking it was left on the original DAC for both presentations accidentally. We all reached a firm consensus on the new DAC's improved qualities and even described them more or less the same. Except we hadn't heard the new DAC. A little egg on our faces.

A similar accident happened trying out a new expensive digital cable. The DAC had three coax inputs so we connected the new one and had the original cable already connected. You could swap inputs remotely which we did. Listened, swapped, listened some more. Came to a consensus conclusion about how one sounded vs the other. This remote let you cycle threw the inputs with one button. For over an hour we compared. Then it was noticed we had gotten swapped around sometime during the audition. The improved new cables qualities were actually paired with the old one. Oops. You might think they still sounded different. But upon making sure and trying again everyone still thought the new fancy cable was clearly better. Which of course makes no sense either way. Everyone was employing regular old sighted listening as we didn't know we had accidentally done a blind test on ourselves.

I understand the regular audiophile ways of doing things. I was one of those guys for a long time, but should have known better. It is fun in a different sort of way, you like being able to periodically improve your sound. Even if it is all in your mind it has a genuine satisfaction to it. The personalities of different designers and their varying philosophies of good sound are more deep and interesting than simply looking at an amp test results and thinking well this one is clean. I get it, heck I lived it. In time thru accidents like above, eventually bother to test myself, and just looking at the facts electronically, I became convinced it is all a result of biased effects of sighted perception. A odd experience as you begin to investigate is being so sure how something sounds different, and knowing from your experience it isn't small at all. You confidently think identifying one vs the other is child's play. Match levels and blind the listening and a very disconcerting evaporation of those perceptual differences occurs.

 

[Blumlein 88]

I agree completely (without checking the probability, I trust you calculated it right).
But I get the notion some folks here acknowledge only results pretty close to 100%, if not 100%.

I did the quick short cut calc on what is significant. It might be off one choice worth, but it is basically correct.

Something you get 20 of 20 is a difference worth worrying about. 70% with enough trials indicates something is real. You have to at some point get very near the edge of our perceptual abilities where you can identify things more than by random chance yet it is a minor difference. That is obvious. Now there are some things that test out as being a perceived difference, but a small one. One where with several hundred trials listeners get 55% correct and something is genuinely different. But if you are getting 55% rather than a guessing 50% obviously the difference has to be close to trivial. And if it is so close to flipping a coin you couldn't ever develop strong feelings about such a difference when you almost as often as not miss the difference. Then again, we know people are highly prone to hear differences when none exist and get strong feelings about it when they are reinforced with other knowledge from sighted listening.

 

[March Audio]

Yes, this is a good point. If someone adds a subwoofer and say the bass is more abundant I doubt we need a blind test. If they added a new power cable to their amp and tell us it was so much better in the bass it was like adding a subwoofer, we'll need a measurement or blind test to give this any credence.

Or just slap them around a bit and say dont be so silly

 

[solderdude]

I'm lazy, really, to care for setting up a blind test. But it's not all. This process is boring. I'd rather allow myself to make a mistake than do this as a usual routine when choosing equipment. There is in fact even more reasons....

This is an honest answer.
When making the decision between buying a $100 and $2000 DAC you would rather allow yourself a $1900 mistake.
That's your decision.
When the reason would be that the 2k$ DAC looks better in the system (and overall), pride of ownership, quality feel and has more/better functionality/options then that would make sense. Even when it performs better there is a good reason for choosing the expensive one over the cheaper one.

I have similar experiences as @Blumlein 88 where one can clearly hear differences where there are none in reality simply because one knows (or should I say believes) what is playing. Those made me think about how the brain operates.
(This was well before the age of internet and wide spread opinions that are easily accessible from the web)

I don't care if you want keep living the dream but that is no reason to tell others that like a more realistic approach they are biased and stupid to trust on measurements that don't say/mean much if anything to you.

 

[Blumlein 88]

Or just slap them around a bit and say dont be so silly

I prefer to tell them, in highly technical language, how the extra bass from the power cord is very likely to overload both woofer and tweeter. You might blow the woofer (you hear the extra bass right?), but people are less sensitive to the high treble, and it only makes sense the power cord has increased the high treble. So it could burn out a tweeter. There is a good reason your amps didn't come with a 'dangerous' power cord like this in the box.

 

[abm0]

The yt link I put, I wanted to hear whether other people like the sound which I like to.

But if the benefit of Filterless NOS is all due to the ultrasonic content, anyone listening to that video through a DAC with a filter will have nothing special to hear. Hell, maybe even the codec used when adapting that recording for YouTube has already thrown out all ultrasonic content. In this case your test proves nothing (even if the effect you're talking about is real, you've nullified it by uploading to YouTube, and if not the listeners will nullify it themselves by listening with a filtered DAC).

 

[zalive]

But if the benefit of Filterless NOS is all due to the ultrasonic content, anyone listening to that video through a DAC with a filter will have nothing special to hear. Hell, maybe even the codec used when adapting that recording for YouTube has already thrown out all ultrasonic content. In this case your test proves nothing (even if the effect you're talking about is real, you've nullified it by uploading to YouTube, and if not the listeners will nullify it themselves by listening with a filtered DAC).

The benefit in filterless NOS R2R is in avoiding producing ringing. Filter produces ringing.
For some technical reasons filtering are mandatory for sigma-delta OS DACs (significantly degraded performance without a filter), while it is not so with the NOS R2R - even if there is audible degradation from the ultrasonic content, you actually get more degradation from ringing if you use filter.

Now objectivists here will claim ringing is not audible. Obviously there's no DBT proof that it's audible. There's no proof it's not audible either. What's objective, though, is that ringing is objectively measurable within audible range. That's the only fact here.

Few local audio designers that I communicate with consider filtering being generally harmful for the sound. Now, while phase linear filter should be better than a classic brick wall filter which inverts phase, it still does cause ringing in the same manner.

The best part with R2R is that it's not that expensive to make a DIY NOS R2R and then experiment. Finished boards are available with more than one option. One might make a blind test on his own, then decide whether he preferred such DAC or some sigma-delta with those perfect measurements. Or would he find those to sound the same.

Everything else is pretty much theory.