[DUPE]Flattening Audio moves audio slightly

After creating a 4 measure audio file of quarter note snare hits I quantized it 100% to quarter notes. I then double clicked on the audio track and it opened in the sample editor.

Using free warp I moved the last snare hit a bit ahead of the beat.

It appears that the snare hits move slightly to the right (except for the first one) when I flatten the audio. (audio->realtime processing->flatten) I used the poly musical algorithm when flattening.

Please see the two attached pictures. Picture 1 (before flattening) and Picture 2 (after flattening).

BTW I did a search and only found something similar related to Cubase 6.

Thanks for any input regarding the above.


:slight_smile:

Just installing the Cubase 7.5 demo.

I will give this a try and post my results.

Just gave it a try with the Cubase 7.5 demo with the exact same results.

I would call this an important issue that should be fixed.

Free warp is excellent but the realtime algorithms do create artifacts. The flattening should help this artifact problem but is useless to me in its current state where it moves the position of the audio. (see the original pictures I uploaded.) Flattening the audio track also changes the sound of the snare drum. Brings out a weird low mid sound that is not heard before flattening.

:frowning:

How do I bring this to the attention of Steinberg?

By doing this: http://www.steinberg.net/forums/viewtopic.php?f=184&t=61062

Have you a) tried the same test with a “bounce” instead of a “flatten”, and then done the same two things on a brand new project (preferably a Steinberg template, just to be sure there’s nothing hidden inside)?

All about it… http://www.steinberg.net/forums/viewtopic.php?f=195&t=53611



Hey claesbjo - is that long thread for the polymusical too … I thought it was just for Elastique? I think I have read the whole thing once, but it was ages ago … can’t remember!

Thought I should be a little more detailed regarding the creation of my audio file.

I first recorded the snare hits using midi and Superior Drummer 2. Next I quantized the snare hits to quarter notes 100 %. I then created an audio file from the midi. So far so good.

Things go to pieces as described earlier.

Alexis, I will give your suggestions a try. (Have you a) tried the same test with a “bounce” instead of a “flatten”, and then done the same two things on a brand new project (preferably a Steinberg template, just to be sure there’s nothing hidden inside)?)

Thanks.

Seems like something needs fixing about flattening as discovered.

But you seem stuck by it - and had a thought:

Why not move the snarehit in midi instead of riscing all these processing artifacts of audio?
Or was that just a test to reproduce in a simple way?

Yes, I did it just as a reproducible test. I am currently editing some tracks for a client which is all audio tracks.

I have tried a few more things and you can see the results in the attached pictures.

Picture 1 shows the following:

  • first track is unedited quantized snare hits
  • second track shows that I have moved the last snare hit using free warp
  • third track shows what happened when I bounced the free warped track
  • fourth track shows what happened when I exported the free warped track
  • fifth track shows the result of flattening the free warped track

I must say that I am not impressed with the results. I hope that Steinberg can make it a priority to fix this issue. I just find it so much easier to edit something like a vocal track using the free warp method.

At least, on the bright side, bouncing and exporting works properly when the track was edited manually. (cut and moved a section of the audio) as can be seen in Picture 2.


Hello,

This is a duplicated thread:
http://www.steinberg.net/forums/viewtopic.php?f=184&t=62714

Best regards,
GN

Apologies for posting here, I went to comment in the thread linked to by the moderator instead of this duplicated one, but the linked-to thread is locked.

If I understand your tracks in pic 1 correctly, the 3rd one is a bounce of the edited one. It doesn’t look like the same waveform for the edited beat or the unedited beat, surprisingly to me. I’ve never looked at bounced waveforms at that high zoom level before …

Some questions for anyone who might know:

  1. Is this fixed in 7.5.30?
  2. Is this timing error audible only on rapid onset signals like snares, or also on piano, voice, guitar?
  3. Any work-arounds that allow Free-Warp to be while keeping the timing intact?

Thanks -

This hasn’t been fixed that is why it is still on the collected issues forum and not in the solved issues.
This Issue it’s still open.

Best regards,
GN

Thanks for reply, clarification, GN -

How will we know when this has been fixed?

Thanks.

Hello,

It will be included on the list of solved issues, better known as version history that is released with each maintenance update, and the original thread will be moved to the solved issues forum.

Best regards,
GN

I’m confused here, as I didn’t see this issue in the Cubase 8 Collected Issues list?
It does appear in the Cubase 7 list, but as there are no more versions of Cubase 7 forthcoming, it’s destined to always stay there, right?

Any clarification would be much appreciated re: … Is this issue present in Cubase Pro 8 versions?

Hello,

Yes, because it is already on a list somewhere we do not move collected issues from one family to the other.
Probably nobody has reported it in Cubase Pro 8, there are many reasons.

It does appear in the Cubase 7 list, but as there are no more versions of Cubase 7 forthcoming, it’s destined to always stay there, right?

Yes, at least for Cubase 7 users, if it has not been fixed it will remain in our Internal To do list and reviewed on a later stage, and if there is time, it will be fixed.

Any clarification would be much appreciated re: … Is this issue present in Cubase Pro 8 versions?

¯_(ツ)_/¯

Best regards,
GN