Master Thread: Are measurements Everything or Nothing?

[Bozon]

Actually, that is exactly what I would expect. Group confirmation bias. This is why tests need to be double-blind.

That is only If you tell them what are you hearing upfront.

As I know that influence I just let them play with the system for a while and after that we share our observations. It is not an ABX of course but good enough for me.

 

[Beave]

I have 5 DACs accumulated on my rack and only one is being actively used because it's the one with less SQ flaws of all.
The rest sits there collecting dust and everytime I try to use them I spot their faults and return to the same one. All near SOTA implementations, 1 based on 9028, 2 9038 one Akm 4499 and a Chord Hugo2.

The one based on AKM 4490 has the widest sound stage, laser focused, but no rhythm, no texture and gets you ears tired on less than 30 minutes. The rest don't offend as that one. The best of the pack by far is the Gustard X26 pro is clearly more resolved, better peace, better dynamics, texture and good sound stage (but not as spectacular as the AKM on that regard) FR-wise is almost 100% perfect but I can sometimes hear a very small "flavor" of it on some tracks.

All of them, by the theory advocated here, should be transparent and indistinguishable, but that is far from truth (again).

Well, I believe we are all tired of this discussion. So I will rest now.

Oh, I get it now. Amir needs to add tests for DAC rhythm, soundstage width, focus, and texture. Lirpa Labs makes just such test equipment, so let's raise the funds for Amir to buy one.

 

[DonR]

Oh, I get it now. Amir needs to add tests for DAC rhythm, soundstage width, focus, and texture. Lirpa Labs makes just such test equipment, so let's raise the funds for Amir to buy one.

I hate it when a DAC loses rhythm.

 

[GXAlan]

I have a Focusrite Scarlett 2i2 that I use for the measurement microphone Can it be used to capture the DACs output with enough fidelity to make that test?

Yes and pkane‘s free Deltawave software can help elucidate the difference.

You can verify that the Focusrite is good enough by testing the same DAC repeatedly to verify that the recordings are determined to be identical to a high level of confidence.

Then you just compare the best vs worst DAC.

Maybe the best DAC measures identically with repeat measurement but the worst DAC has such run to run variation that it isn’t matched.

If you cannot get your best DAC to be identical from run to run, then you need something like an E1DA Cosmos ADC

 

[Blumlein 88]

I have a Focusrite Scarlett 2i2 that I use for the measurement microphone Can it be used to capture the DACs output with enough fidelity to make that test?

Yes. The limits of fidelity will be either the DAC or ADC or their combined effects. The 2i2 is certainly good enough for this to be interesting for you I do believe. It might take a little bit to learn all Deltawave can do, but not too hard.

You can download it for free here:

DeltaWave Audio Null Comparator | DeltaWave documentation

deltaw.org

A long running thread about using it. You can ask questions of Paul (user pkane who wrote this) in this thread and he or others of us who use it can help you get it going.

Beta-test: DeltaWave Null Comparison software

Folks, Only for the brave: I'd like to solicit your feedback and suggestions on DeltaWave null-testing software. It's an early preview, so please be gentle! Details and software download are available here: https://deltaw.org An example report produced by DeltaWave comparing two USB cables...

www.audiosciencereview.com

 

[Beave]

I hate it when a DAC loses rhythm.

Guilty feet have got no rhythm, just like some DACs, apparently.

 

[DonR]

Guilty feet have got no rhythm, just like some DACs, apparently.

Wham!

 

[IAtaman]

@pkane, as the patron saint of free custom software tools for audiophiles.

 

[Bozon]

Yes and pkane‘s free Deltawave software can help elucidate the difference.

You can verify that the Focusrite is good enough by testing the same DAC repeatedly to verify that the recordings are determined to be identical to a high level of confidence.

Then you just compare the best vs worst DAC.

Maybe the best DAC measures identically with repeat measurement but the worst DAC has such run to run variation that it isn’t matched.

If you cannot get your best DAC to be identical from run to run, then you need something like an E1DA Cosmos ADC

Thanks. I will follow your recommendations.

 

[HarmonicTHD]

Thanks. I will follow your recommendations.

We will see mostly the errors from your measurement setup, your room and your transducers and not any from the DAC. So why not measure the output of the DAC directly? Ah wait Amir already did that. But knock yourself out.

 

[voodooless]

We will see mostly the errors from your measurement setup, your room and your transducers and not any from the DAC. So why not measure the output of the DAC directly? Ah wait Amir already did that. But knock yourself out.

I guess this is where the synergy is hiding?

 

[Bozon]

We will see mostly the errors from your measurement setup, your room and your transducers and not any from the DAC. So why not measure the output of the DAC directly? Ah wait Amir already did that. But knock yourself out.

This measurement will be the output from the DACs reproducing a test track compared to it's original file. The objective is to prove or disprove that while all transparent in theory, the differences on perceived sound quality rank are reflected on the ADC captured track as having the same correlation on digital fidelity and then this supposedly impossible to distinguish by ear is true or false.

 

[Beave]

Wait, what are we (you) proposing to "measure" here? Surely not to characterize DAC performance by using a microphone to measure acoustic output of speakers in a room?!?!

This would introduce all sorts of unwanted variables and can't measure anywhere near the precision that electronic measurements can for a DAC output.

 

[GXAlan]

It’s perfectly fine to use a microphone to measure the DAC output if you can demonstrate that your measurement is reliable enough to have a very deep null when playing the same product twice. This makes it HARDER to show a difference when your microphone has its own noise. If one of the DACs is outputting 4.1 instead of 4V and it causes input clipping downstream, for example, you would have two DACs sounding different in your own room.

But as noted above, he is going to measure output of the interface he had handy. This has a higher chance of showing a difference that isn’t actually audible.

The key is that differences tend to be very small if they actually exist and so recording actual music that you hear the difference in is the best chance of showing a difference.

 

[Beave]

But the differences in DACs are so small that they will be swamped by things like environmental noise, even the position of the person in the room, and of the microphone temperature and the speaker temperature (each time he plays the track, the speaker voice coils will heat up a little, and that will result in differences larger than the differences between any of the DACs).

 

[Bozon]

Wait, what are we (you) proposing to "measure" here? Surely not to characterize DAC performance by using a microphone to measure acoustic output of speakers in a room?!?!

This would introduce all sorts of unwanted variables and can't measure anywhere near the precision that electronic measurements can for a DAC output.

No, I was thinking to connect the 5 DACs to the ADC directly. The measurement microphone is not suitable for capturing music as far as I now.

Now I am thinking that maybe my audio interface doesn't have the required input. I will have to check for that because on my understanding the XLR connectors are meant to be used with instruments not with the typical 4 volts that DACs provides.

 

[rha61]

Those graphs have hugely amplified vertical axis. Some of those variations are less than 0.2 dB! For me, the proper presentation is +- 5 dB, for a total of 10 dB, not 4 dB as JA is doing. It creates concern where none should be there. This is the reason I stopped using my simulated load. It was just showing these small variations and nothing else.

IMHO, it's THD and IMD that could be more telling on a simulated load

 

[GXAlan]

No, I was thinking to connect the 5 DACs to the ADC directly. The measurement microphone is not suitable for capturing music as far as I now.

Now I am thinking that maybe my audio interface doesn't have the required input. I will have to check for that because on my understanding the XLR connectors are meant to be used with instruments not with the typical 4 volts that DACs provides.

The peak is 2V input. That's fine to prove/determine if the DACs sound different. It'll work AS LONG AS you DO NOT turn on 48V phantom power.

Focusrite Scarlett 2i2 Audio Interface Gen 3 Review

This is a review and detailed measurements of the Focusrite Scarlett 2i2 Audio Interface (DAC/ADC). I was talked into buying it recently. The 2i2 costs US $140 including free Prime shipping so very reasonable for amount of functionality. For a budge device, the Scarlett 2i2 looks pretty good...

www.audiosciencereview.com

 

[voodooless]

Now I am thinking that maybe my audio interface doesn't have the required input. I will have to check for that because on my understanding the XLR connectors are meant to be used with instruments not with the typical 4 volts that DACs provides.

No, it will work fine, usually you can set the gain of the input somewhere. Check the manual. Seems like there is a simple button and an automatic mode.

 

[Beave]

No, I was thinking to connect the 5 DACs to the ADC directly. The measurement microphone is not suitable for capturing music as far as I now.

Ah, ok, that makes a lot more sense. The ADC performance will have a tiny impact, but much less than a microphone and speaker would have.