so. since i can't afford a fancy loopstation pedal, i'm making one in pd and taking advantage of my low-latency soundcard.

i've made a set of midi trigger pedal buttons - 8 channels, two buttons each channel (one for "record", one to toggle "play" or "silent"). you stomp on the "record" button on the first track and start playing music, then when you're done you hit the button again and the track starts playing back on loop. every subsequent track records in sync with the first loop. you can alter the volume of each of the 8 loop tracks, or stop and start them at will.

i use the [tabwrite~] object to record each loop, then play them back using [tabread]s controlled by [line~]s.

despite driving myself bonkers trying to program the operation of the button interface, so far everything's shaping up well. but the biggest annoyance is the 'pop' that you get when a loop of a constant but unsteady sound (ie, a single sustained violin note, recorded badly) jumps back to the zero position to start playing again - due to a difference in amplitude between the end of the loop and the start.

could anyone suggest a way that i could automatically alter the end of each recorded array so that over the final few milliseconds the amplitude is smoothed (or 'faded') to approach the amplitude at the exact start of the loop? i'm guessing this might eliminate the jump in amplitude that causes the 'pop' effect - let me know if you think i'm wrong

err, hope this is understandable.

cheers,

jon.

you could simply use an extra line~, triggerd by the same bang as the tabread driving line, to quickly fade in (and quickly fade out again at the end, using del and samplelength-5ms or something). Or look into B09.sampler.loop.smooth from examples. There are also higher level abstractions for this in rj lib, or try the xsample objects.