board w/ stereo codec (pcm5102a or wm8731) and 8 x 12 bit ADC (MCP3208), suitable for use with raspberry pi models pi 2, 3, zero, A+ and B+ (*). 16bit / 48kHz, 10VPP, w/ buffered CV inputs etc

view from behind (w/ wm8731 and w/o rpi attached):

and here's some kind of demo/test (this is the wm8731 version, w/ model A+); the object used is simply disis_munger~ ; two samples coming from eurotrash, into the left and right channel inputs; parameters more or less randomly modulated (grain size, # grains, pitch, delay length, stereo spread) (much the same could thus be achieved with the (old) pcm5102a version using files on the uSD card) :
https://soundcloud.com/menschenimhotel/ ... edium_test
BOM and panel svg is here: https://github.com/mxmxmx/terminal_tedium
build guide here:
https://github.com/mxmxmx/terminal_tedium/wiki
here's the old/rpio demo:
http://soundcloud.com/menschenimhotel/rpio-1/s-UHNCvnot sure how useful this is in demonstrating the sound quality of the thing ... but here's a few minutes of not-particularly-inspired audio
*
NB: models 3A+ and 3B+ are compatible in principle, but won't work out of the box. see here.
original post:
... courtesy of PCM5102A, which is an i2s stereo audio DAC. the point was/is to get decent sound quality at modular synth levels, rather than having to use an usb audio interface or the native pwm output. at any rate, this isn't supposed to show off some pd patch.
.. sorry i didn't come up with something more decent (or at least systematic); truth is, i didn't get round to spending a great deal of time on/with the software side of things, so it's just some fairly straightforward pitch shifting and time-stretching, where pitch and stretch factor are controlled by control voltages (some random-ish sample+hold stuff; not that it's noticeable).
i figured i use some voice sample (suzanne ciani), which is what you can hear the first 10 seconds or so. that's just the wav file (aplay), the remainders is puredata resp. csound. edit: updated the file to make it somewhat less repetitive. it's still just various samples run through a phase vocoder patch; from around 1:00 you'll notice it's actually two rpios, one sounding a bit duller as it went through a
PT2399 based delay, with the feedback cranked up.
the outputs go into a L-1 stereo mixer (one via the delay) then straight into my audio interface. everything is stereo, 16bit 44.1khz (pd) or 48khz (aplay). well, some of the wav files were actually converted from mp3.



