A/B Testing two audio interfaces

[Skezza]

I have two almost identical audio interfaces but one I've modified slightly. How would I go about A/B testing them to see if made a difference?

Recording a square wave through both and comparing? Sine wave? Two WAV audio recordings?

I'm new to this. Please be gentle lol.

 

[kemmler3D]

I think the best (i.e. most thorough way to identify differences) would be to use Deltawave (app developed by @pkane ) to compare the output of each device. Do a "loopback" recording where you record the output of the DAC on its own ADC input, or if you have a better ADC on hand, use that.

You can use any audio to compare, but if it were me I'd use all of the above, plus sine wave sweeps.

This requires some nuance to do correctly, but probably worth the effort.

When / if you find differences revealed by Deltawave, you can go back and listen A/B to see if what you hear is consistent with those measured differences.

 

[Blumlein 88]

Yeah Deltawave and maybe Multitone would be useful. Both from pkane. Search for pkane and links are in his signature. Free and excellent. Also nice that they work with two different ASIO devices on the same computer at the same time.

 

[DVDdoug]

If there's an audible difference it would likely be hiss from the mic preamps. But make sure to match the gain.

 

[Sokel]

Download Multitone Analyzer,prepare a short loopback cable (which is the interface? ) and we will help you with settings and measurements.
Since they are the same the results will show you if the mod worked or not.

 

[Skezza]

Thank you. I'll get on that tomorrow.

To provide context of what I've done, I have removed the TL072 and installed what I think/hope is an OPA1642 as the preamp. I did this by removing the SOIC from the PCB. I've done plenty of SMD rework so fairly straightforward.

One thing... I don't know that the chip is legitimate. I believe they are, but I'm not sure. Is there any way I can prove it?

 

[Skezza]

Does anyone know of any good methods of identifying an OPA1642 and proving it's real / fake?

 

[Sokel]

Does anyone know of any good methods of identifying an OPA1642 and proving it's real / fake?

It's performance maybe?
Comparing it with another decent opamp should reveal if something is wrong.

 

[Skezza]

Thanks for everyone's advice so far. I've not had an opportunity to fully prove this mod has made a significant difference although I believe I can hear more clarity. I do plan to post some more details at some point, but currently haven't had chance.

I was told I could do a simple noise floor test by plugging in my guitar, turning the knob up and just recording what was being picked up and subsequently analyzing that data, by increasing the volume.

So that's what I did:

The top is the TL072

The bottom is the OPA1642.

Would I be wrong in thinking that this demonstrates that there's less noise in the background of the OPA1642?

Again, please be gentle, I'm still a n00b as you can tell.

 

[Sokel]

Thanks for everyone's advice so far. I've not had an opportunity to fully prove this mod has made a significant difference although I believe I can hear more clarity. I do plan to post some more details at some point, but currently haven't had chance.

I was told I could do a simple noise floor test by plugging in my guitar, turning the knob up and just recording what was being picked up and subsequently analyzing that data, by increasing the volume.

So that's what I did:

The top is the TL072
View attachment 351422
The bottom is the OPA1642.

Would I be wrong in thinking that this demonstrates that there's less noise in the background of the OPA1642?

Again, please be gentle, I'm still a n00b as you can tell.

It's far better to do the test with the loopback as suggested in the previous posts.
Op-amp differences can be so small that are negligible but it's interesting to see it in a way that will be immediately visible.

It's even easier than the test you did,and a great (free) tool to diagnose stuff all over your chain.

 

[Skezza]

Having a punt exactly as you guys have described.

It might be that I should do a new thread, documenting what I'm doing, but basically what I've done is swap the TL072 for an OPA1642 (assuming not fake). I've done nothing else. What I was hoping is that by upgrading the preamp, it would improve the sound quality (if only a small amount).

 

[Skezza]

So to test this idea, I went to https://www.kozco.com/tech/soundtests.html and downloaded organfinale.wav and played it back through the jack input from the headphone jack of both the original interface and the modified one. I recorded it using a typical DAW and then output to the same length, same bitrate etc WAV file.

I'll admit that I don't fully understand what it's telling me currently, the only place where there's an obvious difference is under Loudness

The blue line is the modified version. The white line is an original without the modification.

Is there a part of this tool I should be looking at?

 

[Skezza]

I'm going to try again, I totally didn't spot the record function... Learning all the time though!

 

[Skezza]

This time I've tried a sine wave.

Starting to think my ears aren't playing tricks after all There may be more I can do with this little interface, but it appears to have done something, I think?

 

[Skezza]

I'm going to keep doing some more investigations. Still need to understand these rather awesome software packages and how I best go about testing.