Tidal with Supercollider 3.9
by roger pibernat
Hi,
sorry if this has been asked before.
When I run Tidal with Supercollider 3.9 I get the following error
*** ERROR: SynthDef dirt_delay2 not found
FAILURE IN SERVER /s_new SynthDef not found
Then when I run a pattern I keep getting this error endlessly.
FAILURE IN SERVER /n_run Node 1003 not found
I also noticed this on SC's log when recompiling the class library
exception in GraphDef_Recv: UGen 'SwitchDelay' not installed.
which lead me to suspect it's related to the dirt_delay2 issue.
Is there any way I can fix this?
Best
Roger
(+34) 639 81 26 71
@rogerpibernat <https://www.instagram.com/rogerpibernat/>
http://www.rogerpibernat.com
2 years, 6 months
SuperCollider: distort a sum of signals, and voices that receive instructions more than once
by Jeffrey Brown
These are two separate issues, but related.
A sound I fell in love with as a young guitarist was sending the signal
from multiple strings into a monophonic distortion box. Unless the pitches
going into it are perfect multiples of each other, the result is a bunch of
interference tones that weren't present in any of the original signals.
In Tidal, I can create a multi-frequency signal, and apply the "shape"
(distortion) parameter to it. But doing that distorts each signal
individually, resulting in none of those delicious interference parameters.
When I look in SuperDirt -- which I understand almost none of -- I see
something called globalEffects. The compressor sits there. But I don't want
something literally global, because that would sound bad. I would like to
be able to select, for each signal, how much of it goes into the distortion
unit, and how much bypasses it.
Actually the ideal would be to be able to create and destroy any number of
named effect units from Tidal, and mix sounds into each effect unit for as
long as the unit existed. This would let you do things like monophonic
portamento -- gliding a single persistent voice between successive pitches,
as in [1], or like slide guitarists do, or the synthesizer in Snoop Dogg's
Gin and Juice[2]. A sequence of distinct successive notes can't sound like
that; it has to be a single continuous waveform, responding to instructions
received over the course of its life. (The same is true of classic
didgeridoo and dubstep rhythms, or wah-wah guitar work ala Hendrix's Voodoo
Child (e.g. the sound that starts at 0:43 here[3]).)
I talked to Julian Rohruber about creating and destroying persistent named
voices. He liked the idea but I think he said it would be a big refactor.
He never said that about a distortion box, though!
[1] https://www.youtube.com/watch?v=S6M843hOkQs
[2] https://www.youtube.com/watch?v=fWCZse1iwE0
[3] https://www.youtube.com/watch?v=IZBlqcbpmxY
--
Jeff Brown | Jeffrey Benjamin Brown
Website <https://msu.edu/~brown202/> | Facebook
<https://www.facebook.com/mejeff.younotjeff> | LinkedIn
<https://www.linkedin.com/in/jeffreybenjaminbrown>(spammy, so I often miss
messages here) | Github <https://github.com/jeffreybenjaminbrown>
2 years, 6 months
stacking multiple params with a rhythm (related to superdirt MIDI)
by Mike Hodnick
This question doesn't seem appropriate to go in the SuperDirt MIDI github issue, and it's a bit too verbose for Slack, so I'm asking here.
With the new SuperDirt MIDI, I'm able to control a single CC param on my synth just fine:
d1 $ s "midi*8"
# n "0"
# midicmd "[noteOn, control]"
# midichan "15"
# ctlNum 1
# control (scale 0 127 $ slow 30 sine)
Above, it plays a pattern of 8 notes, and oscillates CC param 1 from 0 to 127 over 30 cycles. Every time a MIDI note plays, CC param 1 changes its value. Great!
But then I want to control two params:
d1 $ s "midi*8"
# n "0"
# midicmd "[noteOn, control]"
# midichan "15"
# stack [
ctlNum 1 # control (scale 0 127 $ slow 30 sine),
ctlNum 2 # control (scale 0 127 $ slow 20 sine)]
The problem is that the two CC params now only change once per cycle, instead of once with each note that is actually played. What I'm hearing is something more like # control (discretise 1 $ scale 0 127 $ slow 20 sine).
The only way I've been able to get around this is to duplicate the sound pattern into the ctlNum pattern:
# stack [
ctlNum "1*8" # control (scale 0 127 $ slow 30 sine),
ctlNum "2*8" # control (scale 0 127 $ slow 20 sine)]
So, how can I put two ctlNum patterns into a stack and have them fire off with the same rhythmic pattern as the sound pattern, but without duplicating the pattern string?
-Mike
--
Mike Hodnick
mike(a)kindohm.com
2 years, 6 months
