C15 (SOLVED): Two same audio samples won't cancel each other when insert effects are used

Cubase
1

Hi. I had to scratch my head trying to understand why I can’t null the same audio sample on two different tracks (inverted phase on one track). On one track there were some effects that were turned off. On another track, no effects at all. And the tracks couldn’t null. When I removed those plugins, the tracks nulled.

Is it ok to have this behavior? So Cubase cannot compensate for that latency and make a perfect null test without completely removig or shutting down the effects (alt+click on the effect slots)? That’s is not right. It should compensate for that, otherwise we have those time asynchronization all over the project.

However, later I found another weird bug and it was even worse. I used the same plugins this time - DaTube (native Steinberg plugin). So that they had the same latency and they didn’t null! Without effect, nulls perfectly.

So I tried the same test in Bitwig. I used Arturia Dist Tube Culture. And it did null perfectly. But it didn’t in Cubase. So seems like Cubase has a fundamental problem here.

Here is a very short 2 tracks NullTest project with DaTube. One track is phase inverted.

Null test.zip (62.9 KB)

You can even try Magneto, Distroyer, Tube Compressor. Or Waves compressors. They won’t null

2

That’s intentional so you don’t get glitches when un-bypassing the effect.

3

A “bypassed” plugin is still active in the signal chain, so it’s using the same amount of memory and adding the same amount of latency, but the effect isn’t in the audio path. When you turn it off, you’re unloading it and freeing up system resources–and there’s no more latency.

4

Main post has been extended.

5

Main post has been extended.

6

But there should be delay compensation when using plugins that introduce latency, no?

7

I took a look at your project. I don’t think this has anything to do with latency. It’s just that the DaTube plugin is non-determinisitic, so you shouldn’t expect a duplicate track to sound the same (or null when inverted). If I mute both DaTubes, or mute one and delete the other, it nulls for me.

8

That’s not true. I have tried Arturia Coldfire, Dist TUBE-CULTURE, Steinberg Magneto, Waves CLA pliugins. They don’t null ONLY in Cubase. But in FL-Studio and Bitwig they null perfectly. I can’t believe I just switched to Cubase and already found such a fundamental bug that nobody has reported in years? That’s just too much for me…

If you have already made a few tests, just compare the same plugins in other DAWs as well. You will get a better picture.

9

I can’t reproduce what you’re reporting. Sorry.

Try inverting the phase by another means. e.g. from the audio menu.

10

What exactly you cannot reproduce? My project is already done with the inverted phase. Just press play. What do you hear? I did the invertion even with the 3rd party plugins - same result. Try Steinberg Distroyer plugin. Same thing. Do you have any Waves Compressors?

11

As I already explained, you shouldn’t expect that to null. If you need further explanation, I’ll leave that to somebody else.

12

What saturation plugins do you have? Maybe I have them too, so i can try out in two different DAWs. And btw, DaTube is not to that extent non-linear plugin to have such a huge difference in null test.

13

I tested the same thing today. I found that when a send is on the track but turned off — while the send track itself is enabled — the sound doesn’t match.

14

i repeated test. files are equal now. so it was glitch

15

Intentional what? Various tracks time asynchronization? It’s all should be latency compensated, no matter what.

16

Your original post has been heavily edited, so that first post from me that you’re referring to is in reply to a post from you that no longer exists :grinning_face: .

For the explanation of what you’re seeing, see my second post.

17

Your second post was not related to the problem though. It didn’t explain why the tracks didn’t null. And mlib answered you, btw.

18

It says exactly why they don’t null and why you should not expect them to. If I don’t respond to you further it is only because I can’t think of another way to explain this to you. Sorry.

19

Turning off the effects on one channel shouldn’t cause the time de-synchronization btween the rest channels. You are trying to state the opposite..

If you are talking about your post about DaTube, then I have already answered too. All Arturia distortion plugins null in FL-Studio and Bitwig, but not in Cubase 15.

20

The issue has been found. I did the phase invertion in Cubase in the Pre-Section. In that case the tracks never nulled. If I used some third-party plugins for phase inversion at the end of the Channel effects chain, then it was a perfect null.