A REAL DAW summing test

[Pipelineaudio]

my theory is that the same people that worry about this crap are the same people that never make any music

Since its an issue that comes up every 15 minutes or so, I thought it would be nice to put it to rest once and for all

[Cordelia]

my theory is that the same people that worry about this crap are the same people that never make any music

Since its an issue that comes up every 15 minutes or so, I thought it would be nice to put it to rest once and for all

I get the impression he's talking about the "my DAW beats your DAW" crowd. Those of us very weary of this discussion thank you, Pipelineaudio.

[headquest]

Surely it would only have been put to rest if all the hosts had nulled?

Robert Henke/Ableton claim that independant tests they commissioned reveal differences between the major hosts, and that on the basis of their findings they have made improvements for Live 7. See the OP in this thread for a summary I gave before of Robert's comments:

http://www.kvraudio.com/forum/viewtopic ... 48&start=0

[Pipelineaudio]

All the ones we tested nulled except for cubendo, under certain circumstances, but that was already a known issue

[headquest]

All the ones we tested nulled except for cubendo, under certain circumstances, but that was already a known issue

Sorry, must have misread - I though you had mentioned issues with Logic and something else too. I'll go and re-read it.

Just out of interest, as Live (one of my main hosts) and Audition (the other) aren't listed... what do you make of Robert Henke's points?

(As with this question in general, I don't know too much about it, and I would certainly like my own mind to be at rest on the whole matter. So thanks again for your helpful research, and for sharing it with us here at KVR )

[stefancrs]

Surely it would only have been put to rest if all the hosts had nulled?

Robert Henke/Ableton claim that independant tests they commissioned reveal differences between the major hosts, and that on the basis of their findings they have made improvements for Live 7. See the OP in this thread for a summary I gave before of Robert's comments:

http://www.kvraudio.com/forum/viewtopic ... 48&start=0

Keep in mind that if there's the slightest difference in dB on the main outs, which you compensate for in the nulling test, you will get floating point errors (inaudible ones mind you, around like -150dB or something!). So they won't null. Audio-wise they will null, but not mathematically.

That being said, if you use 24bit integers you "only" have about 144dB dynamic range, which means that any errors below that threshold will go away. Furthermore, how many DAC's have a 144dB SNR? Be happy if it's slightly above 100dB.

[Pipelineaudio]

[
Just out of interest, as Live (one of my main hosts) and Audition (the other) aren't listed... what do you make of Robert Henke's points?

You are free to download the testfiles and add live and audition. I will update the zipfile with the result if you do. I didnt mean to exclude anyone, just posted what people sent me immediately.

As to Robert Henke's points, Justin Frankel uploaded a picture last night of some experiments he was doing, its possible these are along the same lines

http://www.cockos.com/forum/showpost.ph ... stcount=37

Im not at all competent in math, nor am I able to grasp concepts much more complex than the basics of a camshaft timed, internal combustion engine, so I don't really understand this whole issue except for the surface, but it seems to say that it isnt much of a problem at low track counts, but could become something in a real mix.

[headquest]

Thanks for the feedback (both of you). I'm off to work shortly, but if I get time later I will download the test files.

For all your modesty pipeline, you are far more knowledgeable than me with all this - I'm a complete layman (cough: musician) so thanks for bearing with me!

[Epiphany]

My girlfriend is the first audience for my tracks as they roll off the line; she is not musical or technical in the slightest. Recently, she commented on a track as "it sounds cleaner than the other stuff". I'd just switched to SX3 from FLStudio. Oddly, despite all the science, I thought exactly the same thing, which is why I switched.

[stefancrs]

You probably just work a bit differently in it.

[pough]

my theory is that the same people that worry about this crap are the same people that never make any music

Nah. I never worry about this stuff and I never actually make any music. Also, people have to make lots of music and switch hosts to be able to think that they're hearing something that isn't there.