A/B-testing D/A-converters - how?

[zongkong]

I'm about to write a thesis on the (eventual) differences in certain qualities between two D/A-converters
(an M-Audio USB Audiophile and an Apogee Rosetta 200 - to be honest I take it for granted that there will be a noticeable difference in a lot of areas. Then again, I've never done a proper A/B-test with converters, so I really can't say I know for sure.).

Now, in order to do this, I'll do some listening tests of the double blind type
(i.e. for every quality tested, the participant has to listen to 3 soundclips - one A, one B, and one double of either A or B. The participant will not know there is a double, or even that they are listening to different converters).

But - I've come upon a problem, since I don't know of any A/B-ing software which can let me use more than one audio driver.

Do you know of such a program? Or do you perhaps know anyone who would be up to creating one for tests like these?

I already know about pcabx.com, however none of those programs let's the user do this.


Many thanks!



This is how I plan to hook things up btw:

For the Rosetta 200:
A/B-program -> Fireface ADAT out -> Rosetta 200 ADAT in -> Rosetta 200 Bal. out -> Monitor mixer -> Speakers

For the USB Audiophile:
A/B-progam -> Fireface SP/DIF out -> USB AP SP/DIF in -> USB AP RCA out ->
Monitor mixer -> Speakers

[Ruben A]

whatever you do.. be sure the material you test with (audio) is more than cd-material (44.1khz). Use fx. a turntable with a proper piece of music.. record it (min. 48khz -> higher = better) (a/d - maybe record for both cards) and then play it thru a decent program.

Or better; make a monitor signal thru a program and let the turntable (dont forget step-up, RIAA) play directly!!! Which means, the sound will go A/D -> around a program and then out D/A which gives you an idea of the converter for both recording and playing(then you might use two computers; one for each soundcard and a y-split connecter for the RCA-outputs from the turntables step-up)

A lot of people makes the mistake with testing using a CD-players digital output. It simply doesnt take the converter to the limit resulting in a very small difference between converters in general.
So when going analogue (min. 48khz) it gives you a better result. That´s why some people says that vinyl still sounds much better than CD´s..
Here you can play around with the sample-freq... see how high you can go with the cards wihtout loss.

good luck (psst you will be able to hear a difference!)

[zongkong]

Thanks Ruben!

I should've included in the original post that I'll be testing music material, as well as just voices, in different bit depths and sample frequencies (recorded with either a U87 or a AKG 426 into Millenia HV3D preamps and a PT 192 A/D).
I haven't done the design for this yet though.

But back to the main problem - how to actually set up this test. How do I let the participant just be able to press maybe 5 buttons -
the first 3 to select what converter they want the soundclip in question to be played through,
the next one to rewind the clip,
and the next one to go forward to the next clip?

[christtjj]

you think you can actually hear the difference between 44.1 and 48?

it's a copy protection scheme, not a quality enhancement.

[zongkong]

Originally posted by christtjj
you think you can actually hear the difference between 44.1 and 48?

it's a copy protection scheme, not a quality enhancement.

Actually, I don't know for sure. But under good circumstances, it wouldn't really surprise me if I did.

What I meant with different sample rates though is not only 44.1 and 48, but also 88.2 and 96. But then again, I haven't really designed this part of the listening test yet.

[Ruben A]

the higher - the better.. no doubt.

ofcourse, you would also be able to hear the difference with 44.1 khz. Just wondering how much "quality" the soundcards actually offer? And what with the rest of the testing equipment? Do you actually know how this sounds? What´s the reference?

@ christtjj


the reason why 48khz-> is to place focus on raw recordings.. fx. vinyl, which is very good for testing.. yeah, take it higher if possible?

[zongkong]

Originally posted by Ruben A
the higher - the better.. no doubt.

ofcourse, you would also be able to hear the difference with 44.1 khz. Just wondering how much "quality" the soundcards actually offer? And what with the rest of the testing equipment? Do you actually know how this sounds? What´s the reference?

Hmm I'm not sure I understand what you mean, but the soundcard, the device generating the digital signal, will be an RME Fireface 800, from which it will go to the digital inputs of the USB AP and the Rosetta 200.

The speakers that will be used are either going to be a pair of Genelec 1022A
http://www.genelec.com/products/history/1022a.php

or a pair of Adam S3-A
http://www.adam-audio.de/studio/midfield/s3a_data.htm

in a radio broadcast control room (acoustically treated and designed).

As for "reference", I'm not sure what you mean?

[Ruben A]

reference is everything; the sound from loudspeakers, mixers, cables, room, music.. . But most of it stay the same even if you replace the soundcards. I just mean, you should get an idea of what it sounds like before concentrating on the souncards D/A.. test the system with a lot of music so you know what it can handle and what it can´t!
Then it would be easier for you to find the difference between the cards.
It´s possible the best card can add some more width/height to the sound-image - but the equipment aint good enough, it won´t show you. So if all setup seems new to you - you never heard it before - it would be hard to get an idea of where the differences might be between the cards..

you get it?

[zongkong]

Ah =)
now I'm with you!
Haha, I'm gonna sound like a know it all now, but here we go:
the room (with it's 1022 Gens.. I love them, if not only because the look like Darth Vader's helmet) is without question the best listening room I've ever been in. It's simply put a place that allows for even the smallest of details to come through. Then again, that would be my subjective opinion about it, but like I wrote earlier it is a properly acoustically designed and built room.

As for the soundcard part though, like I wrote before I'll only use the Fireface as a soundcard, for both converters.

Thanks for your input Ruben, keep it coming!



Back to the main topic - how to get a hold of a program that can let me do what I want it to (look at the third post if you haven't already), or how do I find someone who would find it interesting to create such a program?

[Scoz]

I don't know if they'll be of use but you may find someone who either knows about a program or is interested enough write a program for what you are doing at http://www.hydrogenaudio.org