DAC blind tests? EVER?

[Killingbeans]

Yeah, that Airist really is poorly made.
In case the poor design isn't the explanation, do you, or anyone else here, have a suggestion to what the explanation could be, except perhaps not level-matching them completely?

Haven't seen any info on the reconstruction filter used in the MOS16. If it's a super slow variety with very early treble roll off, that might give an audible difference.

 

[Blumlein 88]

He did level-match them - or at least tried his best.
If he weren't to use a microphone and the pots he used, how would a consumer then level-match two DACs?

To expand upon what BDWoody has already posted.

Use one source and set a comfortable listening level with music. A level you can use for the listening comparison of each DAC.

Run a test tone thru, 400 hz, 440hz, 1 khz probably around -20 db is a good level. Use a voltmeter to measure the voltage at the speaker terminals. Switch to the other DAC or DACs and adjust level until it matches your first source within +/- 1% of the voltage you read. You now have adequate level matching to proceed.

The voltmeter need not be expensive or super accurate. Any inaccuracy will be the same for either DAC being tested. For matching purposes it will be fine.

 

[solderdude]

Someone here did what he considered a very easy test of telling DACs apart

Its not someone... It is Marv. He has an agenda and believed to have the goldenest of ears. One should not take him serious unless one is an SBAF believer, only then his word is gospel.

 

[board]

Its not someone... It is Marv. He has an agenda and believed to have the goldenest of ears. One should not take him serious unless one is an SBAF believer, only then his word is gospel.

I don't think I've been to that forum before, so I hadn't heard of him before, but if he's unreliable then ...

 

[board]

To expand upon what BDWoody has already posted.

Use one source and set a comfortable listening level with music. A level you can use for the listening comparison of each DAC.

Run a test tone thru, 400 hz, 440hz, 1 khz probably around -20 db is a good level. Use a voltmeter to measure the voltage at the speaker terminals. Switch to the other DAC or DACs and adjust level until it matches your first source within +/- 1% of the voltage you read. You now have adequate level matching to proceed.

The voltmeter need not be expensive or super accurate. Any inaccuracy will be the same for either DAC being tested. For matching purposes it will be fine.

Okay, thanks. When I compared amplifiers I ran a 100 Hz tone (as that's where most music peak) then used my MiniUSB Umik-1 microphone and a phone app to measure the volume levels and managed to match them to 0.3 dB difference between three amps. When someone doesn't have a voltmeter I would still say that this would be pretty accurate matching, although it might be less reliable than a voltmeter. I ended up turning the brightest amplifier one notch down, and it still seemed brighter than the rest.
Actually, a friend of mine, who's an acoustics engineer, said that if he were to do a test with volume levels matched he would measure it with a microphone rather than a voltmeter. So, I'm thinking that if possible, measure with both a voltmeter and a microphone to make sure that nothing is off.

 

[solderdude]

Unreliable is a big word.
Big imagination and having a feeling of grandeur is perhaps more accurate. His followers think of him as an undisputed guru.
On certain things I even agree with some of the things he writes.
I put him in the 'grandpa McClown' category. Some truths mixed with fanatasy (while stating it as a fact).

 

[Blumlein 88]

Okay, thanks. When I compared amplifiers I ran a 100 Hz tone (as that's where most music peak) then used my MiniUSB Umik-1 microphone and a phone app to measure the volume levels and managed to match them to 0.3 dB difference between three amps. When someone doesn't have a voltmeter I would still say that this would be pretty accurate matching, although it might be less reliable than a voltmeter. I ended up turning the brightest amplifier one notch down, and it still seemed brighter than the rest.
Actually, a friend of mine, who's an acoustics engineer, said that if he were to do a test with volume levels matched he would measure it with a microphone rather than a voltmeter. So, I'm thinking that if possible, measure with both a voltmeter and a microphone to make sure that nothing is off.

Firstly, .3 db is enough to make the louder one sound better even though it won't sound louder to you. The caveat being there is no other difference. If one is really brighter as in different frequency response that would be measurable without a listening test. You want to get within .1 db to be sure. .2 db might work, but more than that can confuse a result.

I don't know how a microphone would be better in a scenario with amplifiers or anything else as long as the same speakers are used. Phone apps or even good SPL meters are more difficult. If you have to use a microphone, using your Umik, use the readout on something like REW by looking at the FFT. Just using the SPL function other noise can make it hard to get precise readings. Especially using lower frequencies like 100 hz low frequency noise even from a truck a block away could change the reading by a small amount using raw SPL numbers. You also could have standing waves at 100 hz in your room meaning you need to be extremely picky about maintaining the exact microphone position for setting levels. Using filtered pink noise with a microphone or SPL meter is probably a better choice.

I've done testing just to see about matching with a sound level meter. On tones even at 400 hz you could stand in a different position of the room (leaving the meter in the same position) and the disturbance of where you were varied the level .3 to .4 db. Where you were altered the standing waves a bit.

 

[Tangband]

The Yamaha WXC-50 has a pretty good DAC implementation so should be transparent, as other good DACs, if the Denafrips had sounded better if would possibly have been beacause of some coloration. Good you couldn't hear a difference, so you didn't get a dud.

Its the preamp ( when enabled ) in Yamaha wxc50 that is the bottleneck in this streamer. The dac with toslink input has been reviewed by LTS in Sweden , and they couldnt detect it in the before/after testing ( blind ) . Its the cheapest dac they have tested that couldnt be detected . This is an achievement because they could detect a much more expensive chord dac…..

This is the reason I bought wxc50 and one can do slightly better with a really good DDC as a digital preamp - toslink into the Yamaha wxc50 and only use it as a DAC .

As a streamer ( in player-mode ) without volumeregulation the Yamaha wxc50 is really good sounding .
.

 

[Blumlein 88]

Its the preamp ( when enabled ) in Yamaha wxc50 that is the bottleneck in this streamer. The dac with toslink input has been reviewed by LTS in Sweden , and they couldnt detect it in the before/after testing ( blind ) . Its the cheapest dac they have tested that couldnt be detected . This is an achievement because they could detect a much more expensive chord dac…..

This is the reason I bought wxc50 and one can do slightly better with a really good DDC as a digital preamp - toslink into the Yamaha wxc50 and only use it as a DAC .

As a streamer ( in player-mode ) without volumeregulation the Yamaha wxc50 is really good sounding .
Not as good as a Linn Akurate DS player, but almost.

Do you have links to those DAC tests we could translate and view? Or is it subscriber only content?

 

[Tangband]

Do you have links to those DAC tests we could translate and view? Or is it subscriber only content?

The reviews are only in printed form . ( Above you have the conclusion in thumbnail form and swedish )

 

[Blumlein 88]

The reviews are only in printed form .

Could you describe their basic methodology if it isn't too much trouble? Not asking for anything super detailed. Also do they use any special test signals other than music? If I'm asking too much don't worry just say so.

 

[Tangband]

Could you describe their basic methodology if it isn't too much trouble? Not asking for anything super detailed. Also do they use any special test signals other than music? If I'm asking too much don't worry just say so.

They use special test signals for the most, and also music. They have a blindtest panel with different people that do series of listening tests before and after the object tested. All levels are calibrated within 0.1 dB .
The testresult must be statistic relevant to draw any conclusions.
One interesting fact is that DAC:s that measures very well sometimes can be detected, with small sounddeviations.

 

[Blumlein 88]

They use special test signals for the most, and also music. They have a blindtest panel with different people that do series of listening tests before and after the object tested. All levels are calibrated within 0.1 dB .
The testresult must be statistic relevant to draw any conclusions.
One interesting fact is that DAC:s that measures very well sometimes can be detected, with small sounddeviations.

Is one of those test signals the impulse like sound of a metronome they used in amplifier testing?

 

[Tangband]

Is one of those test signals the impulse like sound of a metronome they used in amplifier testing?

One of the test signals ( besides music ) is the sound from two drum-sticks when they test dacs , it seems.

 

[Thomas_A]

What was the reference in that LTS test? I assume they tested the DAC.

 

[RobL]

With THIS song, I can easily distinguish my RME dac from any other dac I have in my house, which doesn’t say much but includes 3 different Airport express’s, a couple of Amazon echos w/ line outs and my Airpod pros. Tried it with 2 different sets of speakers (Genelec 8361 and Jbl 705p). The RME places the alternating guitar picks solidly in the center of the image, everything else I’ve tried it through plays them more ephemerally…no particular solid image. Everything else in the song sounds the same between all of them. I haven’t looked for any other examples, just stumbled on this one.

 

[Axo1989]

With THIS song, I can easily distinguish my RME dac from any other dac I have in my house, which doesn’t say much but includes 3 different Airport express’s, a couple of Amazon echos w/ line outs and my Airpod pros. Tried it with 2 different sets of speakers (Genelec 8361 and Jbl 705p). The RME places the alternating guitar picks solidly in the center of the image, everything else I’ve tried it through plays them more ephemerally…no particular solid image. Everything else in the song sounds the same between all of them. I haven’t looked for any other examples, just stumbled on this one.

Ok, that was a cruel trick: looks like Avicii sounds like country ...

Yes, sounds like two tracks of the guitar spread/split left and right to me. On both YT and Apple Music versions (third track on album True) via AirPods Max and Mac > Mojo > Sony Z1R. Not ephemeral or poorly defined, just not centred. It would be an unusual mix to centre that guitar, although anything's possible. I don't have an RME to try though, but I'd check your settings/setup of that DAC.

 

[RobL]

Ok, that was a cruel trick: looks like Avicii sounds like country ...

Yes, sounds like two tracks of the guitar spread/split left and right to me. On both YT and Apple Music versions (third track on album True) via AirPods Max and Mac > Mojo > Sony Z1R. Not ephemeral or poorly defined, just not centred. It would be an unusual mix to centre that guitar, although anything's possible. I don't have an RME to try though, but I'd check your settings/setup of that DAC.

Thanks for the reply. I have EQ off in the RME as I use GLM, I’ll have a look at the rest of the settings but am not sure what else might impact imaging? Kinda bothers me now as I’m not sure if it’s giving me a correct image in other material.
Does anyone have an RME Adi2 that can play the track through it and report if they get the same image from it?

 

[SIY]

With THIS song, I can easily distinguish my RME dac from any other dac I have in my house, which doesn’t say much but includes 3 different Airport express’s, a couple of Amazon echos w/ line outs and my Airpod pros. Tried it with 2 different sets of speakers (Genelec 8361 and Jbl 705p). The RME places the alternating guitar picks solidly in the center of the image, everything else I’ve tried it through plays them more ephemerally…no particular solid image. Everything else in the song sounds the same between all of them. I haven’t looked for any other examples, just stumbled on this one.

Have you tried this with controls?

 

[Jimbob54]

Thanks for the reply. I have EQ off in the RME as I use GLM, I’ll have a look at the rest of the settings but am not sure what else might impact imaging? Kinda bothers me now as I’m not sure if it’s giving me a correct image in other material.
Does anyone have an RME Adi2 that can play the track through it and report if they get the same image from it?

The different guitars seem to drift a bit in the image plus there are obviously multiple guitar tracks confusing things a little. In no set up were they centred though. You might have accidentally enabled mono on that particular I/O profile on the RME , or significantly adjusted the L/R balance without knowing. Or possibly got one of the crossfeed settings enabled.