Branches hack suggestions


#1

Hi
time comes and goes, back on MI forum gain after 5 years ( shruthi time ) or so …
anyhow i jumped into avr coding lately…Atmel studio 7 up and running for editing
MI repository cloned and and running in vagrant …just kind of workflow since i like the ms ide and intellisense … anyhow i plan at first hack a bit branches …here my suggestion …put at the moment i’m up for other input as well

i call it for now BranchesShruhti
For short:
In this mode branches should play given patterns ( those from shruthi like o-o–oo-oo-o…)
with included displacement map …see more below …
-In 1 still clock or other gate pattern in ( shared on both channels )
-In 2 becomes a Reset

In Detail:
-pressing both button simulant longer ( switch into parasite mode )

  • maybe i let both let blink or turn on for a sec to red to signal it ( details )

Panel functions in this mode:
-In1 acts as clock in or other sequenced gate in ( as normal )
In1 acts also as In for channel 2
In2 becomes a rest functionality in this mode ( could be as well a other sequencer pattern just In1 gets reseted more often )

2 sub modes gets implemented
-pattern select
-probability between out 1 and 2

Pattern select:
Pressing top button enter sub mode 1 pattern selection

  • top poti select the patterns ( idk yet like 64 patterns to select )
  • bottom poti selects the displacement of the pattern
    (short remember the shruthi with this glass…options displacement maps like 1,2,3,4,5,6,7,8 paly pattern normal order 8,7,6,5,4,3,2,1 in reverse order or doing like 1,3,2,5,4,7,8 whatever 64 of those mapping will be there i guess
    =>64 patterns x 64 displacements => 64*64 different variations = 4096 im fine with that :slight_smile:
    (note:first patterns could be simpler like o-o-o-o or o—o---o----o … to get also a clock divider out …anyway will check on tweaking the firmware later )

Portability:
Press bottom switch enter sub mode 2
Top poit for channel 1 as normal and change probability of channel out1 1 and 2
Bottom poti for channel 2 as normal and change probability of channel out2 1 and 2
but internal patterns will be taken rather than the in2 signal ( because those acts as reset )

i have to spend latch or toggle mode to select the sub modes hm… maybe some has other ideas or suggestions as well … keep update on my progress in interest …
maybe only by pressing the bottom button cylce in modes by showing with the led of 1 sec in witch mode … i’m not a big fan of those and i would miss some more feedback with leds to show some modes …anyhow … those is now top on my head … will see where it goes …
but basically the pattern select would be done once and later only play around with probabilities of each channel which is the nature of this module

if some want to join let me know, some python scripts hacking to generate the resource look ups would be nice :slight_smile: i just have to find out to share the repository but should be easy by adapt the makefiles i guess
br widy


#2

Why not just implements a basic Euclidian pattern sequencer, rather than replicating the Shruthi-1 arp patterns?

a|x


#3

@toneburst u are right …but i own 2 modules they do in sum 5 eucl. patten generators …so i go back to more straight patterns…
euclidian will need amount lenght and shift …ok lenght could be fixed …and would work…
but im not that big in eucl. generators …there exits more algorithm wich i dont have an idea …:slight_smile:


#4

Had some fun with branches anyhow that was the outcome for short
ERBSE ( Euclid Random Branched SEquences ) german word for pea … in other words from a smale pea some big plants/sequences will grow/generated …

  • Still a random module if no mode was selected
  • Add a toggle v2 which toggles when previous and current out was/is low
  • Each channel gets a 16 step sequencer with fixed length ( shruthi patterns and simple one to act as clock divider, each knob controls pattern number for each channel )
  • Euclid sequencer with max 16 steps hit and length control shared for both channels ( could be clocked individual ) (32 steps will costs to much resources )
  • Channel could be linked to each other and out overrides hardware in
  • Reset detection to force internal sequencer to first beat on long hardware in idle ( design decision rather than lose on in as reset )
  • Reset all manual ( mode selections not the sequencer tracking)
  • Led feedback for all modes and selections (at least i try with a simple led)

This was more or less i collected from requests here and what i found on muff
At the end its still a random module and it doesn’t matter in which mode was selected if not happy with the result push switches or move knobs until a valid result gets generated … thats more or less the root of this module and without a display the options may confuse.

Finally it will try generate nice bread and butter grooves for some addition sequences in the back of a mix

More here

Still have to run some tests too, if anyone would try it just feel free and maybe report bugs to me :slight_smile: I also try to get a video done on weekend.
br widy