Variable decay time and size when using multiconvolve~

    Nov 17 2013 | 10:51 pm
    Just wondering if this is possible or not, as in the M4L convolution reverb. Looking inside it seems to rely on resampling the buffer with a mysterious object bufresample~ but I can't seem to find any info on it or how it works?

    • Dec 09 2013 | 8:38 pm
    • Dec 10 2013 | 12:55 pm
      Look into the full suite of externals called the HISSTools where multiconvolve~ comes from:
    • Dec 10 2013 | 7:05 pm
      Thanks, I have looked but am none the wiser. No mention of bufresample~ in the documentation. Any hints?
    • Dec 10 2013 | 9:48 pm
      Hmm.. I assumed it was in the library - it looks like it's unique for the live device, you're right it seems mysterious :-) Check in with Alex Harker, the developer of the externals - he's also quite active on these forums
    • Dec 11 2013 | 1:26 am
      I'm way, way interested in this too! For our convolutions we've been creating multiple impulses corresponding to the respective sample rates, which is a total pain. I'd love to get the bufresample~.
      As a side note, anyone have any news on 64 bit versions of the HISS tools?
    • Dec 11 2013 | 4:46 am
      Good timing re 64 bit - Pierre Alexandre and Alex Harker just released version 1.1 which is compatible with both 32 and 64 bit: :-)
    • Dec 11 2013 | 6:09 am
      Woohoo! Huge thanks for the tip!
    • Dec 12 2013 | 9:33 am
      Thanks for the replies, will email ALex when I have a minute. Jeff, I'll keep you informed if I have any breakthroughs.
    • Jan 29 2014 | 3:20 pm
      bufresample~ was written for the device, but it is perfectly possible to use elsewhere. At this stage it is now possible to make that object publicly available, and write a help file for it.
      If I remember correctly off the top of my head it will take a message (probably process) followed by an output buffer, then an input buffer, then I think a ratio for resampling, and finally a sample rate (in case the original is in a different sr). I can't easily check that right now as I am between computers, but it is something along those lines. If that doesn't seem right, or it isn't possible to figure out from the device then let me know and I'll look in more detail,
      There is also a message to set up the resampling filter (which impacts quality and speed), but I forget what that is. The defaults sound OK I think. Internally the resampling uses a windowed sinc approach....
    • Feb 25 2016 | 7:14 pm
      It's been quite a long time anyone replied now but i'm interested in this too ! I'm not very experimented on Max. I used multiconvolve~ to apply a 5 IR convolution on the 9 outputs of an HOA (High Order Ambisonics) spatialization system (externals by the CICM MSH Nord Paris 8). It works good but I want now make the decay time, size and eventually eq and damp variable in real time, as in the M4L Convolution Reverb Pro.
      I tried some things, including play the IR edit it (time stretch, eq...) and record it in a new buffer but obviously it takes the time to record in the buffer, so it's not real time. I also search the web and Cycling' 74 forums for hours but I still don't know how to do. It's for an exam so I really need to find a solution.
      So I would be grateful if you can help me doing it :)
      Thanks, Clément