Scratching a newly recorded sound???

stochastic

So I've gotten the pd xwax library extension installed and working fine, but now I need to understand PD's tabread4~ a bit better - can you guys help?

Essentially I'm looking to complete the following task: record a sample of unknown length (for troubleshooting, so far I've only been using pre-recorded), use my "timecoded" vinyl to control the playback of said sample. I'm okay with there being a start/stop trigger to the recording (if it's needed, I could even live with a pre-detirmined length) rather than having a constant live feed writing to memory (seems wasteful).

The vinylcontroler~ object gives three outlets 1) start/stop decoding bang (signals the object recognition/loss of the timecode signal) 2) pitch of the record (1.0 = 33/rpm) 3) position (increasing from 0 on the outside of the record to 1 on the inside of the record, at a rate which seems to be 0.002 per second at 33rpm)

So I'm wondering what the most elegant way of going about this is, while still achieving a realistic sounding "scratch" manipulation of the soundfile.

Thanks for any suggestions.

sunji

this is a little patch that might do your job. Disect!

http://www.pdpatchrepo.info/hurleur/scratchytable.pd

stochastic

Well that only utilizes pre-recorded soundfile playback, but it does get things to a more acceptable level that I was at before. Thanks. I'll attempt to mod that patch to my liking.

sunji

here is some juryrig on a simple loop patch of mine. Somewhere in there a sound is written to an array called $0-LoopMesa. Probably not helpful in anyway, but I post it anyway.

http://www.pdpatchrepo.info/hurleur/loopstatADV.pd

nestor

position (increasing from 0 on the outside of the record to 1 on the inside of the record, at a rate which seems to be 0.002 per second at 33rpm)

I would base your [tabread4~ patch on that.

This will give you a loop that will index according to the position of the record if I understand correctly.
You'll have to do some math based on the actual length of the record in ms to figure out your multiplication factor. For example, if the record is 22min long, and your sample is 30sec long, you should multiply by 44.

--N

P.S. Where did you find the xwax external????

nestor

P.P.S. it is called [vinylcontrol~]

Interesting way of going about positioning;

relative position; the start is the first cue-point (first tracked timecode where the needle was dropped) and can be reset to the current timecode-position by sending a bang to the third inlet. The end position is the last sane timecode on the record. The position is calculated using (current_position-start_position)/(last_position-start_posit ion), thus giving a range between 0 and 1 between the end position and the relative start position and negative values for positions before the start position.

-- From http://karhumusic.sesser.at/vinylcontrol/vinylcontrol.html

So, you may have to do a little more math than I thought...