Sliding Stretch for tempo change versus GarageBand (and how selection choices matter for SS)

I use Audacity’s Sliding Stretch a lot, usually for tempo but sometimes for pitch too. (The latter I find tricky and imprecise, but immensely helpful when needed.) I often use Sliding Stretch when linking together selections from various takes, since I can adjust the length of notes I didn’t quite play in time with my click track (actually, my iPhone metronome, which I then record to make a visual click track for editing).

I tried to Google this, but I didn’t find any discussion of it. The quality of a stretched or shrunken note or short passage changes drastically depending upon how much space I leave around it when I select it and then apply Sliding Stretch. (When possible, I isolate the notes before working on them and re-integrating them.) A tiny change makes the difference between an ugly misshapen note (waveform) and a good one. Surprisingly, I sometimes even manage to make a better, more uniform note than what I started with, by trying 10-15 times until I get a good one!

When it comes to speeding up an entire track, Sliding Stretch can distort notes quite a bit, something I didn’t catch on to for a while. I tried my usual tricks on a track I recently recorded on an instrument I can’t play up to speed. I put a lot of work into stitching together pieces of the tunes – the first time I ever tried recording tunes I honestly couldn’t play all the way through without seriously messing up. Crazy, but it worked. Then I wanted to speed it all up, but Sliding Stretch distorted some of the notes so much that it destroyed the dynamics (and looked really ugly – I can often detect sound quality issues better after I see a potential problem visually, such as with an airy or fuzzy or spiky looking note). My usual trick didn’t work on my entire track because including differing amounts of space in the selection of the track affected different notes differently; my less-perfect notes would go lopsided in different directions, different notes would get airy etc so there was no one best sped-up version. I also tried various percentage changes (since I don’t care exactly what final tempo I end up with), which made a difference but again, some of the notes would be better and others would be worse.

In desperation, I looked up how to speed up a recorded track in GarageBand. Sorry to have to report this, but GarageBand did a much better job, with hardly any distortion of waveforms. Anyway, in case anybody else is pulling out their hair trying to make a quality tempo change to a whole track, I thought I’d report my experience here. I’m sure it depends on instrumentation and other factors to some extent, but I won’t be speeding up whole tracks with Audacity anymore, unless somebody tells me something I don’t know about how to do it better.

I still will use Audacity to record (when not overdubbing) and for almost all my editing. I even sometimes record in Garage Band and then export the tracks to Audacity for noise removal and editing. In general, Audacity works much better than GarageBand for my purposes. I just ran up against one limitation that I thought maybe was going to ruin a project.

  • Kate D
    I’m using Audacity 2.4.2 on a Mac Air running Catalina 10.15.7.