DAW audio engine test request

[stefancrs]

I haven't done any tests or calculations, but I'd expect a 64-bit summed mix and a 32-bit ditto cancel out to about -150dB or so (just as 32bit mixes does against each other). I could be quite off with this ofcourse, but I expect the difference to be inaudible.

There are plenty of people who claim to hear the difference, though personally I wouldn't comment too enthusiastically! Technically I would assume that it does make a difference though, or Cakewalk/Mackie wouldn't have wasted their development time.

It might very well make a difference. But it might make a difference that happens below -150dB. I've together with others here at kvr compared a few hosts at 32bit mixing (cubase, energyxt, podium, tracktion, vegas and. uhmm. some more) and they all cancelled out to about -130 to -160dB when mixing 16 channels of channel-unique full-range white noise. Peak amplitude, not RMS.

BTW, regardless of if there's an audible difference to 64bit mixing compared to 32bit or not does not have anything to do with whether it's worth implementing or not. If 64bit mixing means more people buy Sonar, it's most likely enough of a reason to implement it. And nothing wrong with that imho, it's not up to cakewalk to educate their customers.

[headquest]

Interesting - thanks Stefan.

It might very well make a difference. But it might make a difference that happens below -150dB. I've together with others here at kvr compared a few hosts at 32bit mixing (cubase, energyxt, podium, tracktion, vegas and. uhmm. some more) and they all cancelled out to about -130 to -160dB when mixing 16 channels of channel-unique full-range white noise. Peak amplitude, not RMS.

Fair enough - that doesn't suprise me. I use a combination of programs, and personally I find it no easier to get a good mix in Sonar with 64-bit summing than I do in Ableton or Audition.

BTW, regardless of if there's an audible difference to 64bit mixing compared to 32bit or not does not have anything to do with whether it's worth implementing or not. If 64bit mixing means more people buy Sonar, it's most likely enough of a reason to implement it. And nothing wrong with that imho, it's not up to cakewalk to educate their customers.

. I doubt anyone bought Sonar (or Tracktion 2) specifically for that feature

[bullshark]

All the test I've read about, including the one conducted here by stefancrs, conclude to the exactness of all daw software rendered output. I will not dispute that, the proof's in the pouding as they say.

What I'm wondering about now thought is the "live" output of different daw software, or even difference between one software live output and it's own rendered output, especially using VSTI.

Why? Well, I've built this drumtrack in eXT using DR008 as a sampler. When done, I rendered the tracks from eXT and brought them into Reaper. Well, they do sound different, and not even so subtil a difference, easy enough for anybody to hear. So I then brought the rendered track back in eXT, they sound somewhat like they do in Reaper, but not like the live VSTI track.

I can't explain the why, or devise a test to prove that without question, maybe someone here with more brains than I could prove/disprove this somehow; although I can't see how to disprove this as my ears are not golden so if I can easily hear it, it should be because there is a difference.

[Pipelineaudio]

All the test I've read about, including the one conducted here by stefancrs, conclude to the exactness of all daw software rendered output. I will not dispute that, the proof's in the pouding as they say.

What I'm wondering about now thought is the "live" output of different daw software, or even difference between one software live output and it's own rendered output, especially using VSTI.

Why? Well, I've built this drumtrack in eXT using DR008 as a sampler. When done, I rendered the tracks from eXT and brought them into Reaper. Well, they do sound different, and not even so subtil a difference, easy enough for anybody to hear. So I then brought the rendered track back in eXT, they sound somewhat like they do in Reaper, but not like the live VSTI track.

I can't explain the why, or devise a test to prove that without question, maybe someone here with more brains than I could prove/disprove this somehow; although I can't see how to disprove this as my ears are not golden so if I can easily hear it, it should be because there is a difference.

We routinely used to record a pass of our daw to Masterlink and compared it to the offline render. The results often do not null. When you attempt to null them, certaion things will JUMP out from time to time. I think it was explained to me that the offline render if far more accurate, and the realtime playback is "as fast as the audio hardware and drives can keep up"

[bullshark]

Thanks Pipe; and that was using strict audio files right? So it is possible that the realtime, live output from different daw would sound different, probably more so using tone generator in VSTI format, while their rendered output null to almost perfection.

This could explain why so many are adamant that software x sounds different to software y, if they were listening and comparing only the live output of both and never bothered to compare the rendered output (which, frankly, is the only one that matter at the end of the day).

[Vitaly]

The difference between offline and real-time rendering could happen (and actually should) if you use delay-based effects as chorus, flanger, reverb(modulated). Because, every time you render offline their LFOs start from some initial (the same) phase. And during real-time playback their relative phase can be anywhere. Of course, these two mixes won't cancel each other and will sound 'different'

[Pipelineaudio]

ahhh another good point!