Max Patch Slowing down when using multiple Peakamp~, line and multislider objects...


    Apr 05 2019 | 3:49 pm
    Hi everyone,
    I'm pretty much done with a pretty big patch composed by a heavy GUI interface (Many sliders, knobs, etc). I'm experiencing a big lag in the interface and I understood that is coming from this part of the patch (when I remove it everything is pretty smooth again). There are 10 sound sources (I just made 8 for the demo) that are analysed via Peakamp on Computer 1 and then sent via OSC to Computer 2 which adjust the data received and generates the visuals in Jitter. Each sound source is then directed inside the patch and used as modulation sources.
    How can I optimise this patch to make it run smoothly? I noticed when experiencing this lag, that apparently I'm using very little CPU (~10%) and GPU on my new MacBookPro 2019 with this config: i7 2.6GHz and Vega20 16GB of Ram. Is there a way to address more efficiently Max to the computer resources? Also I'm trying to trigger events with peaks, but I have often unwanted repetitions, any suggestions to make it more accurate? Thank you very much in advance for your precious help!

    • Apr 05 2019 | 6:22 pm
      PS. All the knobs and fader are from live, I already set the "update limit" to 20ms and ticked "Defer automation output"
    • Apr 05 2019 | 8:32 pm
      I'm not experiencing any lag with your patch (macbook 15" i7 16Gb osx 10.12), but my guess is that the lag is coming from the peakamp's, reporting new values at a 15 ms time-interval and then all the slides that have to calculate there new values. You might want to try and build the entire envelopefollower in the signal-domain. To avoid the repetitions you can then use [thresh~], which has a lo and hi threshold.
    • Apr 05 2019 | 10:56 pm
      Hi Timo, this is what I thought... too much infos for the data domain! Thank you very much for your suggestion!!
    • Apr 06 2019 | 12:47 pm
      you could also try to not run the peakamaps in sync, i.e. use time values which are individually, randomly between 15. and 16. ms.
    • Apr 12 2019 | 10:20 am
      So, I did all the programming in MSP, problem fixed! Thank you very much Timo for the nice threshold~ trick, it works perfectly... Best!