Hello all,
Looked for some documentation on the broad advantages and disadvantages of using tabread4 as opposed to readsf summarized. Would love to get a quick rundown of them.

William

readsf~: good for reading files from disk at the current samplerate of pd, use it when you just want to play back an audio file without time stretching
tabread4~: use it when you're doing granular stuff (like time stretching). A few things to be aware of when using tabread4/soundfiler:

1. it stores the sound file completely in ram (whereas readsf~ incrementally reads into a buffer in ram, not needing the whole sound file at once)
2. because pd is single precision, there is a limit to how precise table lookup is. specifically anything greater than 24 bits (the significand + 1 of a 32 bit floating point number) will lose resolution. there is a help file detailing this. The second inlet of tabread4~ was added to circumvent this problem by using double precision addition in the object, but it's a hassle

basically: use readsf~ to play back sound files, use tabread4~ if you need more of the granular/time stretching stuff

Thank you so much! That has been a great help!

As a side note, there's also [readanysf~]:
http://puredata.info/downloads/readanysf

• support for multiple channels
• can upmix/downmix channel configurations
• pcm seeking
• looping (internally and is sample accurate)
• varispeed play (resampling)
• plays MANY file formats
• plays both kinds of music, ... country AND western.