not sure what you mean. stolen voice? firstly, to deal with latency in audio I suggest recording from the [delread~] object rather than [adc~] directly. secondly, when the sample is triggered, give the [line~] object [1 5( or something (assuming its initialised to zero) to ramp up. Then you already know how long the sample is so at the same time as banging the [tabplay~] and line objects, you can bang [delay n-5] (where n is your sample length in ms) sent to a [0 5( message in turn sent to the [line~] object. you are then ready to retrigger the sample.

Is this at all useful or have I missed the mark?? good luck

M ----- Original Message ----- From: "jfm3" jfm3@ouroboros-complex.org To: pd-list@iem.kug.ac.at Sent: Monday, March 24, 2003 4:06 AM Subject: [PD] percussion sampler

I'm making a percussion sampler performance patch. Latency is a big issue.

In all of the sample playback Pd patches I've seen, There's a 5 ms delay when a voice is stolen while the sample ramps down, avoiding a click that would come from jumping back to the beginning of the sample while not at a zero crossing.

What's the best way to get around that? Is there a way that I can tell when I need to ramp down and when I don't? For example, if I've reached the end of a sample, I don't need to ramp down again since I'm already at zero, I can immediately start playing the sample back from 1 again.

If I were for example to put a [delay 5] after all my control signals, would that add to the 5 ms of regular pd buffering latency, or would it overlap?

Thanks in advance for your illumination.

-- (jfm3 2838 BCBA 93BA 3058 ED95 A42C 37DB 66D1 B43C 9FD0)

---

PD-list mailing list PD-list@iem.kug.ac.at http://iem.kug.ac.at/cgi-bin/mailman/listinfo/pd-list