Greetings! I am preparing the long awaited 1.4 release of my critically acclaimed Rack plugin Computerscare Modules and am soliciting anyone willing and able to install the pre-release to give it a test before I submit to the official Library. I’d especially appreciate if anybody has an existing patch using any computerscare stuff, because I don’t want to inadvertently break any existing functionality! Here are the release notes:
New Modules:
-Horse a Doodle Doo: pseuso-random trigger & CV sequencer
I built it for Linux, and verified that animated GIFs work in the blank panel (I have priorities). I’ll try and incorporate a few other things into whatever I work on today and see where it gets me.
Oh excellent, thank you very much! I realized after my post that I had actually never tested on Linux and thought that the GIF stuff would be the most likely to have a platform-specific problem. Did you create a distribution release (accomplished via make dist on the command line)? If so, would you be willing to share the package with me so I can upload it to the github pre-release page?
Pattern: Changes pattern: input for voltage control.
Length: Pattern length: length of pattern, input for Voltage control.
Density: Pattern density: from 0 to 100%, 0 being silence and 100 being a new trigger out for every trigger in the clock input. input for voltage control.
Poly Knob: Control the number of output voice
Questions:
How are patterns generated? Are they random per Rack session or do they change when you randomize the plugin?
The polyphony modes are a bit weird. Initially it only used 2 voices no matter what the polyphony was set to. Randomizing the module got it into using 16 voices. This seems like a bug, like there was a missing initialization or something.
“Each Channel outputs independent pulse & CV sequence” - Are they in lockstep or offset or ?
“All channels triggered by channel 1 sequence” - or ‘every note a chord’ mode
Trigger Cascade “each channel is triggered by the previous channel’s trigger sequence” - This seems to indicate that the step of channel N is advanced by the trigger out from channel N-1.
EOC Cascade “Each channel is triggered by the previous channel’s EOC” - This seems to work , but what it seems to mean is that if e.g. channel 1 has a 16th note trigger, channel 2 will get one trigger at the end of channel 1’s sequence.
The patterns are not random. The settings of the pattern, length, and density knobs completely determine the gate and CV sequences. Randomizing the module will change the pattern because it will change the knob values. The algorithm to generate the CV is some big complicated sum of sine waves that seemed to work well enough to give a lot of variety. You can check out the actual algorithm in the source code
Is it possible you were providing a 2-poly-channel signal into one of the module inputs? If the poly knob is set to “auto”, (all the way to the left), the polyphony channels will be the greater of all connected polyphony inputs. I haven’t noticed this behavior, and yes it would be a bug. Also, the polyphony knob should not be randomized when you randomize the module. If you have a patch file that exhibits this behavior I’d love to take a look!
I don’t understand the question here
Yes it seems that you are understanding the functionality correctly. I think I actually left the trigger outputs for these modes somewhat… unfinished. I couldn’t think up how the trigger outputs should behave for channels 2-16 in these modes since their CV is all derived from channel 1. Perhaps I should explicitly set the trigger polyphony output to 1 channel only in these cases, I’ll have to think about it.
Thank you very much for trying it out and for your feedback!
Do the sequences on each channel all start together, and last the same length? i.e. in lockstep with each other.
As for the two voice thing, this shows up unpredictably leading me to think there’s a bug in there somewhere. This patch demonstrates it. Every time I load it, regardless of the mode, Horse a Doodle Doo only sends triggers on the first 2 voice channels.
The sequences all start together, but any given sequence might have a rest (no trigger) on its 1st step, so all sequences will not necessarily output a trigger on their first beat. I find it best to patch a reset cable to Horse a Doodle Doo to make sure everything is consistent.
If the length spread knob is set to 0% then each channel will have the same number of steps. If greater than 0%, as the channel number increases, the length of that channel’s sequence will increase. If length spread is set to less than 0%, then as the channel number increases the length of that channel’s sequence will decrease to a minimum of 2 steps. You can provide a polyphonic signal to the length CV input to explicitly set the length of each channel’s sequence.
Thanks for the patch file, I will take a look this evening.
Thanks for sharing the patch, there is a bug involving the “steps spread” knob functionality. What I wrote here is not functioning as I intended:
However, even once I fix this bug I believe this patch would not output triggers for the higher channels because the “density spread” knob is turned far to the left, which means that most all the higher poly channels density will be evaluated as 0%, meaning no triggers. For example, with density set to 50%, and the spread set to maybe -10%, the density of each channel will decrease as the poly channel goes up, eventually reaching 0% (no triggers) and staying there for all the higher channels.
I think I might exempt this and the length spread knobs from randomization because it’s pretty easy to get empty or otherwise dull sequences.
