there are five voices for glitching. each voice has individual volume, panning, gating, positions and probabilities. each voice contains a softcut loop with random tape modulations. each voice also emits a supercollider engine that adds a wobbly resonant low pass filter to each glitch to get that 90’s feel.
Requirements
audio input
norns
Documentation
quickstart: put music into line-in. set norns global tempo in clock -> tempo to tempo of music. open glitchlets and press K1+K2.
all five glitchlets can be consciously controlled via global params or quick menu. quick menu:
first set clock->tempo then reload glitchlets
K1+K2 does quick start
hold K1 to turn off glitches
K2 manually glitches
K3 or K1+K3 switch glitchlet
E1 switches parameters
E2/E3 modulate parameters
note: make sure to restart norns the first time you install because it has a new supercollider engine that needs to be compiled.
Had a very quick play and it’s great! Being very non-‘beat’/sync-centric, one thing I’d love added is some kind of realtime encoder-based control of the BPM
guess I could somehow control the norns clock bpm with midi, tho I like the idea of being able to use it with as few external controllers as possible, generally
@Cementimental as for the bpm - i can do that. but, just fyi the bpm basically controls the resolution for selecting sample start/ends. the resoution is quantized, but i’m not sure it matters much since the line-in audio might not be synced to the start of the whole loop… the bpm also controls the total amount sampled (basically its the “seconds-per-beat x number-of-beats-per-loop”). if you’re interested in just changing the total amount sampled, you can also do that from changing the number of beats per loop and leaving the resolution the same. i would suggest the latter instead if that’s your use-case!
I spent a pleasant hour or so with glitchlets last night - this particular occasion made a Volca Modular go to rather interesting places, especially with select MS-70CDR effects patches following.
I found mapping a midi controller (in this instance a Launchpad Mini) to sample length and glitch probability gave useful results, but I also completely failed to record anything as I was so involved with the script.
all that is un-holy!
THIS is fracken AMAZING!
(you were right…i absolutely dig this script!)
it’s sounds crazy awesome!
weird thing though…
i have the top and bottom graphics as well as the play bar…but…there are no sound columns like in your video examples.
weird…
(running on a stock norns.)
here’s tonight’s session…
lot’s of Glitchlets with @tyleretters Arcologies woven in there as well as a couple of Tocantes going through @Justmat Bounds and Foulplay controlling a DRM mkIII:
@SPIKE I really like that noiz. And the visuals are also neat.
First impressions of glitchlets. I definitely didn’t have this script completely dialed in when I made this track. The glitches are loud compared to the drum part that is being glitched. But the effect is very musical. I’ll take more time to dig in and really understand. I thought adjusting the softcut output would change the volume of the glitches, but I didn’t hear a change.
Beets into Glitchlets+layered acoustic guitar+vocals through mlr
@reallyok this is great! the glitchlets sound like a wonderful companion for your guitar playing and the plumbutter!
@McNUG wow! i LOVE that! the glitched drums sound incredible paired with that guitar. i would love to hear more music like that!
sorry - yeah the softcut volume won’t work because i have a different parameter tied to it. you can adjust the volume of each glitchlet (glitchletX -> volume). you can also control the master volume with "glitch volume" (controls the stuttering effects volume) and “warb volume” (controls the engine volume for those big bassy warbly sounds / other glitch effects). i do find that the volumes need to be tied below 0.5 or they are a little too powerful.
@SPIKE thought you might like it! it sounds awesome through your setup
as for the sound column - that should be based on the waveform. in my video i had dry drums as input so it has obvious peaks/valleys from bass/snare hits. if you input music it might not have peaks/valleys. the reason for the display of the waveform is to help inform of where the beat actually is (since the start of the loop is not synced to the start of the input). then you can carefully position each glitchlet to correspond to a specific portion of the waveform. there might be better approach to this (like automatic beat finder?). but all of that is if you want to take the carefully curated approach to glitchlets. i find glitchlets sound pretty good just making them all randomized (K1+K2).
this update requires a norns restart because engine changed
finally figured out how to use supercollider’s beattrack so that should help synchronize incoming audio so all the quantization works better. to make best use of quantization make sure to set the global tempo before starting glitchlets (so that parameter offsets are quantized). syncing takes about 10 seconds to take hold.
here’s a demo with amen break:
changelog
feature: added a beat finder into engine which syncs to audio after ~10 seconds, so quantization is better
feature: better glitching sounds
ux: better defaults when pressing K1+K2
ux: engine glitches with softcut (better sounding)
hey ya!
just updated…so i think i discovered something…or i am not grasping what is happening.
so i launched Glitchlets on a norns and a shield norns.
BING!
there’s the analysis waveform just like in your vids!
then…i realized that the Cirklon wasn’t on…as soon as it booted up…VOOOP…the waveform graphics disappeared on the norns but remained on the shield since it’s not connected to the MIDI clock at the moment.
