Improvement to the vocoder effect

I am just guessing from the name that this function has something to do with multiple threads, which maybe Audacity is using for some GUI purpose?

Are you sure the Cancel button continues to work during long calculations?

Yes, I checked that, though “long calculation” has taken on a new meaning :wink:

Just use my declicker :frowning:

Just thought, calculating spectrogram displays seems a great opportunity to make multiple cores useful. And make anticipatory calculations during playback, so tere is no delay in displaying it after you jump a screenful.

A bit off topic, but were can I find the best verson of Chris’s compressor? My current little research project is reading one of the versions.

Here: http://theaudacitytopodcast.com/chriss-dynamic-compressor-plugin-for-audacity/
but note that the license terms are incompatible with GPL.

That points to 1.2.6, I was reading 1.2.7 b1. Should the latter be considered unreliable, experimental, or whatever?

What are the serious license implications for use and for creation of new variants?

1.2.7b was a beta version (denoted by the “b”)
The core code is virtually the same as the 1.2.6 version, but includes experimental features that don’t work quite right.
The 1.2.7b version is licensed under an MIT license, which as far as I’m aware is compatible with GPL v2.

If you want to work on a version, go for 1.2.7b.
If you just want to use a version (unmodified), go for 1.2.6.

Of course if you are wanting to work on a version, it would be worth also looking at the 1.2.6 version to see what changed.

I see a lot of linear-to-db applied to individual samples and wondering whether s-log and scaling instead would improve performance of this one… that may make two performance hacks of things I don’t really understand!

Hi, just bumping this topic for a couple of small tweaks:

  • version 4
  • octaves redefined to support lower sample rates
    vocoder.ny (5.6 KB)