This is a proof-of-concept port to JS of the Phase Vocoding Analysis/Resynthesis routines from Tom Erbe's program "SoundHack".

nodejs npm

npm install

./bin/cli.js -h

./bin/cli.js -i path/to/soundinput.aiff -o path/to/soundoutput.aiff -b 4096 -v 1 -t 4

• the Aiff writer does not guard against writing files larger than the format allows (~4GB) which will cause integer wrap-around in the headers, potentially corrupting the file.
• only AIFF files are readable. Files in AIFF-C format are not readable (including AIFF-C/swot little endian files). Throws an error if you try it.
• only Hamming windows are used for windowing before doing the FFT and for the Overlap Add resynthesis step.
• only timescaling is available. SoundHack also can do pitch scaling, that is not implemented.
• not usable in-browser as it uses core NodeJS classes for file access on disk. This could probably be factored out so samples could be fed from a fetched network file or stream.

• the cli.js file refers to code in the ./dist directory. If you want to edit files in the ./src directory and then use the ./bin/cli.js you must run:

npm run-script build

first so that your changes are available to the cli.

npm test

npm run-script lint