Routing Matrices
  • So, the first few patches I've built have been rather large and contained many repeating subsystems. I know I'm not alone. I'm looking at your Cartesian Sequencer @biminiroad, which is really neat. I've found that dragging all those wires takes almost a third of the time, maybe more. Worse is when I change my mind and have to rewire every duplicate node, all the while hoping I haven't skipped an input.
    So for large scale routing applications, I've concocted a few templates that can be dropped into a patch. I've made sets of eight inputs and outputs, which can be snugged up against a column of mux's or quadtomono's or seq16's, then quickly wired horizontally without dragging halfway across a patch. I hope this keeps me from getting confused, and saves time wiring and rewiring.
    the largest one: "routing template 8x16x16x8", replicates a spiderweb I used in a drum machine I made this week: I had 16 sets of 8 outputs routed to 8 seq16's. The first seq16 received the first of 8 outputs in each of 16 sets, representing the sequence of the bass drum. The second seq16 received the second output from each set, representing the snare, and so on.
    Also, I may have accidentally made some art.
    routing template overview.tiff
    routing template 8x8x8x8.audulus
    routing template 8x16x16x8.audulus
    flip 8.audulus
    lowhigh 8.audulus
    lowhigh 16.audulus
    everysecond 8.audulus
    everysecond 16.audulus
  • Interesting. The 8^4 one runs at 12% on my iPad 4, even though it's only routing. So.. 204 connections.. Guess it makes sense that they're not free.
  • Yes, I am surprised. When I get a second I'll see if it makes a difference whether or not they are connected directly to their destination vs through my direct routing nodes. You're right, makes sense that the routing isn't free, and I suppose that each node doubles the original connections. It could still be useful as a placeholder.
  • Ok on my Mac Mini, the 8^4 clocks at about 4%. I don't have an iPad so thanks for that comparison. I made a node that just connects 32 inputs to 32 outputs, and it also clocks 4%, so I don't know. I'm not too worried. This is mainly to save myself time during layout. If I have to pop those connections over to their final destination once the patch is complete, it's probably still worth the lack of confusion.
    ferry 32.audulus
  • Oh, definitely. Anything to save pain during a complex layout.
  • Thanks for these!
  • No problem @biminiroad. Did they save you some time? Do you have any suggestions for other commonly used routings?
  • Any chance of getting this for 3?
  • @altgrendel - which ones do you want, and what do you what to use them for? if you give me some specifics, i'll build it out for you.

    See the attached patch - it's pretty intense on the inside!

    (note: it almost definitely will not load on an underpowered iPad! I made this on the computer and it's very resource-intensive)

    Screen Shot 2017-02-11 at 7.36.00 PM.png
    1920 x 1080 - 2M
    Random Sampling Sequencer BETA 2.audulus
  • I was hoping this was something like a pin matrix, not sure if that's the case. Something like 8x8 was what I was thinking of but if you're building from scratch I think I could DIY from your screenshot. Is there a patch import from Audulus 2.x to current? That might be a good feature request.
  • @AltGrendel - so something like this?

    If so, that's real easy, and I'll do that whenever you confirm that's what you want! :)
  • I just went ahead and started it - you have outputs for both sides, and for each row/column combination there's a sum output that sums all the turned on inputs in that row, and a mult output that multiplies the turned on inputs with one another.

    I'm currently making the matrix for one row/column combination and then going to copypasta it to the rest (so the +/* is missing from the right side).

    Screen Shot 2017-02-12 at 5.51.07 PM.png
    1920 x 1080 - 336K
  • Very grid, much nice!
    Edit: realized things are going to get geometrically complex with both the adding as multiplying functions, looking forward to seeing what Mark comes up with.
  • Tried to post a response, apparently it didn't take. Yes, that's exactly what I was thinking.
  • Well I messed around a bit and made a summing matrix, hopefully that will come in handy for basic modulation routing. Happy patching!
    RHS Routing Matrix.audulus
    Screen Shot 2017-02-14 at 6.38.13 AM.png
    1416 x 1035 - 243K