Please, please fix AudioSnap! ... I am desperate!

[marcL]

This is a problem that I have reported before.

First of all, I like the concept of AudioSnap, it is much better than in other DAWs, because there is the possibility to fix detection issues manually!

The implementation: If you bounce an AudioSnap clip (either by itself, to another track or if you export it), then the notes (resp. the transients) move arbitrarily! If it was only by a minor value I could live with that (and most shifts are minor), but it happens all the time that there are major shifts in some notes that are absolutely unacceptable!

I have been fighting with AudioSnap the whole afternoon and here is my last horrendous example:

I have copied the AudioSnap clip to another track at exactly the same position and then bounced the copy (to make the differences visible). See self above how this note has been shifted by 80 ticks! Like that the music really loses the feel!

Such things make me really sick, it puts me off making recordings and working with music (like today)! Like that AudioSnap/CbB is completely unusable to me! I am desperate and I don't know what I shall do? Switch the DAW? Are there better alternatives? In the past I used Melodyne for timing correction, but the detection is also very weak and often it is nearly impossible to see the note starts between blobs (this is much easier with CbB's wave representation!).

Please fix this issue in the near future!

[Jakub Maggi Malášek]

Yea i point at this problem too... Nothing happend... Now im learning Studio One because i really need to ,,WORK,, not ,,TRYIN,, ...

[marcL]

And again, the bouncing of clips with AudioSnap is fatal !!! ?

I tried to go on with my work on a bass track, but AudioSnap bouncing stopped me! When I had adjusted the transient detection and then used the quantize function all looked pretty well ?. Even the swing was considered (a cool thing that does not exist in most other DAWs, it's a pity that you can only use full percents, no decimals).

But when I bounced the clip I had a lot of note shiftings (just to be clear, not using quantize, but using bounce!), some even worse than in the original clip and in one position there was a bad artifact, the wave form had been changed! Like this AudioSnap is completely unusable (at least for melodic contents) and not only for me !!!

By the way I used the "Radius Solo (Bass)" rendering algorithm. I had tried other algorithms, too, and the shiftings were quite different, but the results were also absolutely unusable. "Elastique Pro" had even much more artifacts, in one position the wave form of a note could nearly not be recognized! ?

Does nobody else have similar problems? Is it just me? Do you use AudioSnap only for percussive contents? What do the developers say?

[Kook000]

Creat new audio track.   Input put aux1.      Set output of audiosnap track to aux 1.   Record onto new track live

[John Bradley]

On 8/2/2020 at 6:11 AM, marled said:

But when I bounced the clip I had a lot of note shiftings (just to be clear, not using quantize, but using bounce!), some even worse than in the original clip and in one position there was a bad artifact, the wave form had been changed! Like this AudioSnap is completely unusable (at least for melodic contents) and not only for me !!!

I ran into the same problem in the past (and for a while, stopped using Elastique Pro for the offline render, as that seemed especially problematic).

I've recently gone back to using Elastique (Efficient / Pro) for realtime and rendering. I use it on all my guitar tracks -- given the latency of my system and/or my sucky playing, I'm invariably 'hot' to varying degrees.

Anyway, the secret I've found to having it work without introducing huge random timing shifts: use as few markers as possible. I'll typically use one marker per bar (or one every couple of bars, if I can get away with it) and disable all the others. I always visually/audially inspect the render (by bouncing to another track and then dragging it right below the original and listening to it 'dry') before tossing the original. Haven't had any problems since I started doing that.

---

Another tip to anyone playing with AudioSnap: if you're working on a signal that's 'clean' (like a vocal, or an acoustic guitar), make sure your markers are moved to slightly before the transients. That way, if there's any difference in the amount of stretching/shrinking going on before and after the marker, the change won't be as noticable. You can sometimes hear weird phasing issues if the marker is in the middle of a chord. Put it right before the attack and the sound of the attack will cover up any such issues. Probably.

This isn't as much of an issue if, say, the audio is dry guitar (feeding into an amp sim in the FX bin), and the amp is heavily distorted.

[Josh Wolfer]

oof. Id recommend emailing support@cakewalk.com. As good as this forum is, this is a serious issue that needs some good attention

[marcL]

On 9/17/2020 at 1:14 AM, John Bradley said:

Anyway, the secret I've found to having it work without introducing huge random timing shifts: use as few markers as possible. I'll typically use one marker per bar (or one every couple of bars, if I can get away with it) and disable all the others. I always visually/audially inspect the render (by bouncing to another track and then dragging it right below the original and listening to it 'dry') before tossing the original. Haven't had any problems since I started doing that.

This is true that if you have less transient markers the risk of problems is reduced! Me too, I reduce the automatically detected transients pretty much, but most of the time I cannot reduce it to only 1 marker per bar in my case. I also noticed that if the corrections (stretching/compression) are really minor, then there is often no problem.

On 9/17/2020 at 1:14 AM, John Bradley said:

Another tip to anyone playing with AudioSnap: if you're working on a signal that's 'clean' (like a vocal, or an acoustic guitar), make sure your markers are moved to slightly before the transients. That way, if there's any difference in the amount of stretching/shrinking going on before and after the marker, the change won't be as noticable.

Again, I agree with this. It is also my experience. But if you do this, then you cannot use the automatic quantizing, i.e. you have to do it manually, since the automatic quantize tries to align it to proper beat positions (whatever you choose 1/16, 1/8, ...).

[marcL]

On 9/17/2020 at 7:00 PM, Josh Wolfer said:

oof. Id recommend emailing support@cakewalk.com. As good as this forum is, this is a serious issue that needs some good attention

I did this more than once!

[Josh Wolfer]

16 hours ago, marled said:

I did this more than once!

Double oof. 

@Noel Borthwick

[Jonathan Sasor]

@marled Can you send us a copy of your project to look at? A lot of AudioSnap stuff is very material-dependent, so if we had a concrete example where you're seeing your project go wrong, that'd be very helpful. Thanks!

[Josh Wolfer]

@marled There you go. Service with a smile  Best of luck!

[Will.]

@marled Your request has been picked up and noted. Send the project. 

Edited September 25, 2020 by Will_Kaydo

[Josh Wolfer]

Yup. Let's get er done.