AudioSnap - Phase relationships when changing tempo

OK - call me thick, but I can't get this to work properly... Despite RTFM'ing the manual and various threads and videos etc.
 
Here's what I'm trying to do -
 
- Two bass tracks recorded at the same time, (1 x DI, 1 x line out from an amp)
- Originally recorded at X bpm
- Want to change them to Y bpm
 
...and keep the phase relationship.
 
The closest I can get is -
 
- Select both tracks
- Open AudioSnap
- Edit Clip Map to tell it the correct bpm and number of beats per bar (3 not 4)
- Right click and MERGE AND LOCK MARKERS
- Keep both tracks selected hit CLIP FOLLOWS PROJECT
- At this point both clips are still in phase, (doing the same thing in 8.5.3 at that point results in out of phase)
- Change the tempo
- Bingo, someone turned the phaser pedal on
- Bounce to track with whatever algorithm and the quality of the phaser pedal gets better, (but doesn't get turned off)
 
Any ideas??
 
Cheers,
 
Q.

Thanks for your reply Rick. I appreciate you taking the time.
 
I followed your course of action and managed to preserve the phase relationship between the tracks using the "split" method. What I ended up with though, was an audible click every time the now time passed through one of the splits. This was despite the splits having been automatically faded in and out when the splitting was performed.
 
I was unable to get around that situation.
 
So, on a whim, I downloaded Reaper.
 
Literally two clicks and this worked perfectly -
 
1. Select both tracks
2. Right clicks and choose ITEM SETTINGS - SET ITEM TIMEBASE TO BEATS (POSITION, LENGTH, RATE)
3. Clicked the TEMPO display and chose the new BPM
4. Pressed PLAY
 
Perfection. Totally in phase. No artifacts - even in real-time.
 
What a joke! Two days mucking around in Sonar and 20 seconds to do it perfectly in Reaper.
 
Q.

Strange, I've successfully repaired bass this way using 8.5.  Maybe AudioSnap in X1b has issues.

I've also used Melodyne and Reaper to change tempo and fix timing and they both work well in most cases.  Glad you were able to figure it out.

OK - I've been mucking around with this more to try and see how I can make Sonar work. Bottom line is that I've finally found a way to preserve the phase relationships when changing tempo, but it is a complete kludge-fest.

The reason why I went back to Sonar is that while Reaper easily handles the phase relationships while changing tempo with great simplicity and ease of use, none of the included time-stretching algorithms are perfect in relation to the slight tonal shift they impart on the resulting audio. You can change and fiddle with a variety of their "Elastique" algorithms, but I found that you are trading off accuracy for one part of the frequency spectrum for smeariness and a lack of definition in another area. None of the included algorithms and associated presets gave unaltered tone - something was always slightly different.

So in the end, here's how I made it work in Sonar -

- Panned the DI track hard left
- Panned the Amp track hard right
- Bounced 'em together as a stereo track
- Enabled AudioCrap
- Generated and then correct the transients while aligning to the left (DI) side of the stereo file
- Updated the clip map to be the correct bpm and beats/bar
- Enabled the option for CLIP FOLLOWS PROJECT or whatever it is
- Changed the tempo
- Bounced the AudioCrap track out as split mono
- Et Voila, two perfectly in phase tracks at the right tempo that still have the correct, original tone
Hope this helps someone someday.
Ciao,

Q.

Qwerty69

So in the end, here's how I made it work in Sonar -

- Panned the DI track hard left
- Panned the Amp track hard right
- Bounced 'em together as a stereo track
- Enabled AudioCrap
- Generated and then correct the transients while aligning to the left (DI) side of the stereo file
- Updated the clip map to be the correct bpm and beats/bar
- Enabled the option for CLIP FOLLOWS PROJECT or whatever it is
- Changed the tempo
- Bounced the AudioCrap track out as split mono
- Et Voila, two perfectly in phase tracks at the right tempo that still have the correct, original tone 

The equivalent of this without combining the two clips to stereo is:





- Disable all markers on the Amped track.
- Massage the markers on the DI track.
- Select the two tracks, right-click one of them, and choose Merge and Lock Markers (which really means "Merge Markers and Lock Clip Positions")
- Change the tempo (Clip tempo doesn;t matter when you're doing an even stretch like this
- Unlock the clips (to allow bouncing).
- Bounce to clips.


This should give exactly the same result. Whether that result is acceptable or not is another question, but it should be identical to the combining/splitting technique. The key is that you're using one set of transient markers for both tracks.


You make an important point about aligning the markers to the earliest of two transients that are nominally on the same beat. This prevents stretching a note in the middle of its attack. This is key whichever technique you use.


And all of this is much more important if you're doing something like quantizing or manually moving transients relative to each other. Doing an even stretch across a whole track by changing tempo shouldn't be that problematic as far as attack phasing goes. But the distortion caused by stretching sustained sounds is always unmistakeable, and often intolerable as far as I'm concerned, even on a single track.


EDIT: All that said, I have to agree the combine-and-split technique is an ingenious workaround. Two thumbs up. 

Hi brundlefly -
 
Thanks for taking the time to reply. Couple of things re: the method you outline.
 
The method you state is actually what I tried originally. I've just been playing with the project again according to your exact recipe and have again emulated the same kind of results. Specifically, I was following your steps up until you say, "Change the tempo (Clip tempo doesn;t matter when you're doing an even stretch like this)".
 
At that point, if I don't fiddle with the clip tempo map settings, then as soon as I click CLIP FOLLOWS PROJECT, (needed to get the tempo adjustment to occur), the clip instantly changes its start time and internal tempo. The clip track I'm working with is in 3/4 - I presume that's the reason. If it were 4/4, I guess it wouldn't matter.
 
So, after I make that change to the clip map, that's when the out of phase thing happens. As you can't globally select CLIP FOLLOWS PROJECT for both tracks, (you have to do it individually), that's where the error sneaks in. Despite the amp track's original markers being disabled and using the MERGE + LOCK command, as soon as you do the CLIP FOLLOWS PROJECT thing, that's where the phase issue occurs.
 
So no, at least on my end here, this method doesn't give the exact same results as the 2 x mono => stereo => split mono method I outlined above. I've replicated this again using 8.5.3 and X1c.
 
Either way - I'm just glad I found a bodgey way of making it work! :)
 
I do agree with you about the resulting tone. The only offline rendering method that works is the solo bass setting. Everything else does some really weird things to the amp track whenever there is a bit of grit... But using the solo bass setting, to my ears at least, the tone of the editted track is pretty much indistinguishable from the original, slower tempo source recordings.
 
Ciao!
 
Q.

OK -

You were correct, I wasn't using the AUTO-STRETCH option. I just gave that a go. Using that selection, you are correct, the time signature makes no difference and I don't need to edit the clip map.

So said, I still have the out of phase issue present using this approach... Given that I don't have a bass phaser pedal, it's actually a pretty good effect, lol... Just not what I need in this track :)

Q.