Time Correction Overload

Today during my audio production class I was demonstrating to my students how time correction impacts digital audio when pitch shifting more than a few semitones. To illustrate this I pitched down a chunk of music an octave with the time correction on in Pro Tools. The degradation was clear, but it occurred to me that it would be even more obvious if I shifted it back up to the original pitch with time correction enabled again.

This created an interesting way to effectively down-sample the audio. Intrigued, I applied the same technique over several times to hear what would happen on multiple passes. This is something I’m likely to explore more, but I tried it again on the snippet of music from Unprocessed Rhodes Pedal Noise going down two octaves and back up again. It sounds like the audio has been boiled in a pot of bathroom chemicals. Delicious!

Time Correction Overload

6 thoughts on “Time Correction Overload

  1. I’ve found for maximum artefacts you should pitch up (with the time correction on) and then pitch it back down. I think you lose more that way.

  2. Sounds good, Matt. I’ll give that a try. I think the interesting thing about going down is that the time correction algorithm has to remove data, therefore effectively down-sampling. Going up first might be interesting because the algorithm has to inject data into the clip to achieve the time correction. There’s lots more to explore here.

  3. Hi,
    I actually wrote the elastic audio feature in ProTools and I’m curious about which algorithm you were using. I’m thinking it must have been the Polyfonic one.

    Anyway, yes there are lots of interesting artifacts that can be had from over stretching. Try ripping vocals apart with the monophonic alg. Especially just in the decay of a word – it gives a nice weird glittery effect.

    Have fun!
    S

  4. @Stephen That’s really interesting. I’m not sure which algorithm it was. I had “Time Correction” checked while pitch shifting under the Audio Suite menu in Pro Tools LE 7.4.2 if that’s any indication. Cheers!

Leave a Reply