VI playback sounds different in LPX vs CB

Similarly the analog filters in the D/A may change things by a minute tiny amount from playback to playback, but again, the effect of moving your head a few inches in any acoustic space will be orders of magnitude larger than the differences between the filters, because A/D and D/A filtering is designed to alias outside the audible spectrum, so the chances of it being audible are nil. So is the audio coming out of the speakers precisely the same? Possibly not, though it's not within human capabilities to detect that difference reliably, we're talking changes of thousandths of a dB at most...
 

Henu

Senior Member
I didn't find it mentioned before, but while (besides of pan law) I think this is 100% utter nonsense, there is one thing which could theoretically be happening and that's upsampling differencies between DAWs. Yet, I don't know per se how this is handled in LPX.
 

babylonwaves

Darth Fader
Unless you're dealing with either a deliberately nonlinear plugin as many are - or badly coded third party plugin or external gear, offline and online bounces are identical. It's another myth that they are different. The math is simply done faster than realtime, with the expense of not being able to monitor the bitstream sent realtime to a D/A.
all native hosts skip samples when playing back in realtime, depending on the load. and that, obviously, changes and eventually degrades the sound. so, in order to get a good result, you need to bounce offline. it's not a myth at all. also, there are differences in between core audio and asio drivers.
 
Last edited:

Sheridan

New Member
Hi everyone!

Long time lurker here, chiming in because I think I might have something to contribute to this thread.

Just like for the OP the perceived difference between Logic and Cubase had been bugging me for some time, so last year I bit the bullet and performed the null test. Unfortunately I don’t have the sound files any longer so I can’t submit any "evidence", but I will walk you through what I did and you can decide.

I loaded up a strings patch from Albion One in Kontakt in both Cubase and Logic on the same Mac computer. I reset the round robins to ensure the same wave file would play when the note was triggered. Pan law was set to the same in both DAWs and there were no automation or additional effects applied (Admiral Bumblebee has shown that these are indeed different between DAWs).

I then captured the resulting audio outside of the DAW using Rogue Amoeba’s Audio Hijack software. This produced two wave files, one from Logic and Cubase respectively. I imported the files in Cubase and phase inverted one of them.

At first they didn’t null, there was still some sound coming out from the stereo master. I thought that maybe there was a fundamental difference after all. But when I looked more closely at the two original waveforms I noticed that the Logic one was ever so slightly lower in amplitude.

So I increased the level of the Logic file with 3 dB and presto, now the files nulled! Well, at least the stereo master level peaked below -110 dB, which I think was due to the difficulty in lining up the waveforms optically, even at max zoom I was probably a couple of samples off.

So IMHO Living Fossil is correct that the difference is due to the output levels, and this impacts the perception. I had always thought that the mid range in Kontakt in Cubase was slightly clearer, now I am confident that Cubase and Logic sounds the exactly the same and have opted for Logic for workflow and stability reasons.

Of course, if anyone else could take the time to perform a similar test that would provide independent verification and we could bury the topic once and for all.
 
OP
M

Mishabou

Active Member
Hi everyone!

Long time lurker here, chiming in because I think I might have something to contribute to this thread.

Just like for the OP the perceived difference between Logic and Cubase had been bugging me for some time, so last year I bit the bullet and performed the null test. Unfortunately I don’t have the sound files any longer so I can’t submit any "evidence", but I will walk you through what I did and you can decide.

I loaded up a strings patch from Albion One in Kontakt in both Cubase and Logic on the same Mac computer. I reset the round robins to ensure the same wave file would play when the note was triggered. Pan law was set to the same in both DAWs and there were no automation or additional effects applied (Admiral Bumblebee has shown that these are indeed different between DAWs).

I then captured the resulting audio outside of the DAW using Rogue Amoeba’s Audio Hijack software. This produced two wave files, one from Logic and Cubase respectively. I imported the files in Cubase and phase inverted one of them.

At first they didn’t null, there was still some sound coming out from the stereo master. I thought that maybe there was a fundamental difference after all. But when I looked more closely at the two original waveforms I noticed that the Logic one was ever so slightly lower in amplitude.

So I increased the level of the Logic file with 3 dB and presto, now the files nulled! Well, at least the stereo master level peaked below -110 dB, which I think was due to the difficulty in lining up the waveforms optically, even at max zoom I was probably a couple of samples off.

So IMHO Living Fossil is correct that the difference is due to the output levels, and this impacts the perception. I had always thought that the mid range in Kontakt in Cubase was slightly clearer, now I am confident that Cubase and Logic sounds the exactly the same and have opted for Logic for workflow and stability reasons.

Of course, if anyone else could take the time to perform a similar test that would provide independent verification and we could bury the topic once and for all.
If you reset the round robins, pan law was set to the same, no automation or additional effects applied and all levels at unity...then why would one DAW be 3 dB louder than the other ?
 

Living Fossil

Senior Member
If you reset the round robins, pan law was set to the same, no automation or additional effects applied and all levels at unity...then why would one DAW be 3 dB louder than the other ?
Could be related to the master output level or similar.

But:
Why don't you just post two snippets?

This thread is going nowhere without evidence.
If you're too lazy to do two little bounces, there is no sense to debate any further.
But who knows, maybe it's not even laziness but the anxiety of being expelled from Placebo-land.
(which, of course is the most beautiful place in nonexistence)
 
all native hosts skip samples when playing back in realtime, depending on the load. and that, obviously, changes and eventually degrades the sound. so, in order to get a good result, you need to bounce offline. it's not a myth at all. also, there are differences in between core audio and asio drivers.
No, they don't. It's trivial to play back a PCM audio stream without "skipping" samples.
 

Dewdman42

Senior Member
That's the first time I've ever heard that claim too. I tend to doubt its veracity, but I remain open minded for the moment until we hear more.

I do not personally think for one second the real time bounce will render anything different then non-real-time bounce. I am pretty confident they would proceed equal results, tested easily with a null test. The only difference is that one can do it faster then the other.
 

babylonwaves

Darth Fader
No, they don't. It's trivial to play back a PCM audio stream without "skipping" samples.
another poster already said that when you do a null test with two audio files from LX and CB, once level matched both cancel out perfectly. my comment was in regards of what the OP said and also the majority of poster here: VIs. and that's a different story because e.g. Kontakt performs differently on both platforms. there are Kontakt patches which overload the last core in Logic pretty fast and those might be still fine playing the same notes in cubase. that's why you need to bounce before you compare.
 
another poster already said that when you do a null test with two audio files from LX and CB, once level matched both cancel out perfectly. my comment was in regards of what the OP said and also the majority of poster here: VIs. and that's a different story because e.g. Kontakt performs differently on both platforms. there are Kontakt patches which overload the last core in Logic pretty fast and those might be still fine playing the same notes in cubase. that's why you need to bounce before you compare.
If Kontakt is merely triggering samples, it's linear PCM audio streaming. It will be identical. It's not the host DAW that's the difference. If Kontakt is actually overloading a core and therefore causing audio glitches, that's not really anything to do with the DAW unless the DAW is written in an inefficient manner (which they aren't). If the CPU is overloading, we're not talking about one DAW "sounding" better than the other, we're talking about an actual fault/bug. That's very different.
 

foxrec

New Member
Hi everyone!

Long time lurker here, chiming in because I think I might have something to contribute to this thread.

Just like for the OP the perceived difference between Logic and Cubase had been bugging me for some time, so last year I bit the bullet and performed the null test. Unfortunately I don’t have the sound files any longer so I can’t submit any "evidence", but I will walk you through what I did and you can decide.

I loaded up a strings patch from Albion One in Kontakt in both Cubase and Logic on the same Mac computer. I reset the round robins to ensure the same wave file would play when the note was triggered. Pan law was set to the same in both DAWs and there were no automation or additional effects applied (Admiral Bumblebee has shown that these are indeed different between DAWs).

I then captured the resulting audio outside of the DAW using Rogue Amoeba’s Audio Hijack software. This produced two wave files, one from Logic and Cubase respectively. I imported the files in Cubase and phase inverted one of them.

At first they didn’t null, there was still some sound coming out from the stereo master. I thought that maybe there was a fundamental difference after all. But when I looked more closely at the two original waveforms I noticed that the Logic one was ever so slightly lower in amplitude.

So I increased the level of the Logic file with 3 dB and presto, now the files nulled! Well, at least the stereo master level peaked below -110 dB, which I think was due to the difficulty in lining up the waveforms optically, even at max zoom I was probably a couple of samples off.

So IMHO Living Fossil is correct that the difference is due to the output levels, and this impacts the perception. I had always thought that the mid range in Kontakt in Cubase was slightly clearer, now I am confident that Cubase and Logic sounds the exactly the same and have opted for Logic for workflow and stability reasons.

Of course, if anyone else could take the time to perform a similar test that would provide independent verification and we could bury the topic once and for all.
..yes i did the null test too and it cancels but there is still a difference in my ear, its not about the loudness +- db..its about the sound, its more clean open and 3d in cubase...i just cant understand why am i hearing this...
 

foxrec

New Member
Hi everyone!

Long time lurker here, chiming in because I think I might have something to contribute to this thread.

Just like for the OP the perceived difference between Logic and Cubase had been bugging me for some time, so last year I bit the bullet and performed the null test. Unfortunately I don’t have the sound files any longer so I can’t submit any "evidence", but I will walk you through what I did and you can decide.

I loaded up a strings patch from Albion One in Kontakt in both Cubase and Logic on the same Mac computer. I reset the round robins to ensure the same wave file would play when the note was triggered. Pan law was set to the same in both DAWs and there were no automation or additional effects applied (Admiral Bumblebee has shown that these are indeed different between DAWs).

I then captured the resulting audio outside of the DAW using Rogue Amoeba’s Audio Hijack software. This produced two wave files, one from Logic and Cubase respectively. I imported the files in Cubase and phase inverted one of them.

At first they didn’t null, there was still some sound coming out from the stereo master. I thought that maybe there was a fundamental difference after all. But when I looked more closely at the two original waveforms I noticed that the Logic one was ever so slightly lower in amplitude.

So I increased the level of the Logic file with 3 dB and presto, now the files nulled! Well, at least the stereo master level peaked below -110 dB, which I think was due to the difficulty in lining up the waveforms optically, even at max zoom I was probably a couple of samples off.

So IMHO Living Fossil is correct that the difference is due to the output levels, and this impacts the perception. I had always thought that the mid range in Kontakt in Cubase was slightly clearer, now I am confident that Cubase and Logic sounds the exactly the same and have opted for Logic for workflow and stability reasons.

Of course, if anyone else could take the time to perform a similar test that would provide independent verification and we could bury the topic once and for all.
...and i did another test, send 2 files to my friend who works in logic too without telling him who is who...and he said too that the file 2(cubase) sounded more open and 3d....regards
 

Zero&One

Senior Member
...and i did another test, send 2 files to my friend who works in logic too without telling him who is who...and he said too that the file 2(cubase) sounded more open and 3d....regards
Post them up with a poll.
A
B
Same

I suspect it’s just a placebo, but would be interesting
 

Bear Market

Active Member
..yes i did the null test too and it cancels but there is still a difference in my ear, its not about the loudness +- db..its about the sound, its more clean open and 3d in cubase...i just cant understand why am i hearing this...
Confirmation bias?
 

Sheridan

New Member
I would like to stress that in my tests the audio was captured outside the DAW applications and when adjusted for loudness the files nulled. Logic = Cubase = Pro Tools = Kontakt Standalone.
 

foxrec

New Member
if you have cubase 10 too pls make a test, make one 8 bar midi file and use the same in logic and cubase export at the same volume without any plugin...i would like to hear your opinion, regards and stay safe
...with the same vst au instrument...