1180 – performance laggy when MIDI In and Out enabled?

Home Forums General News and releases 1180 – performance laggy when MIDI In and Out enabled?

Viewing 15 posts - 1 through 15 (of 15 total)
  • Author
    Posts
  • #778
    msepsis
    Participant
      • Topics: 219
      • Replies: 732
      • Total: 951
      • ★★★

      When I open my panel in 1180 and do not set the midi ports (keep in/controller/out set to none) and rotate knobs I get no lag, so I don’t think it’s caused by Lua.
      When I enable my panel’s midi Out port to the synth I also don’t get any lag.
      When I then enable the panel’s midi IN port from the synth I get terrible lag on many/most/all? modulators on the panel. I’m deving this panel on an old HP D330 on win xp on an intentionally older system but never had these kinds of performance hits with earlier revisions of ctrlr ever. I can literally hear the HD "churning" HEAVY any time a modulator is rotated when both MIDI in and out ports are connected to the synth.

      the synth has a midi in, out and thru port, I’m not coming out of the thru – im coming out of the out port.
      I do not have MIDI thru enabled in any way in ctrlr.

      Normally if I were to see a post like this I would think it’s likely a PEBCAK (newbie) error but I’ve been using this panel in ctrlr for a long time now and haven’t witnessed this kind of terrible lag until the latest ctrlr revs. I read someone mention they’re patch dumps update the modulators "lightning fast" in the newer revs which is disconcerting to me as it’s completely the opposite for me. patch dumps take a good 3-4 seconds to update here. This "lag" is occurring on modulators that have no Lua scripts attached and that are not driving any other Lua scripts at all.

      Monstrum Media | Music, Sound & Software Design, Chicago / San Francisco listen

      #5015
      atom
      Keymaster
        • Topics: 159
        • Replies: 2945
        • Total: 3104
        • ★★★★★

        I’ll check that out asap. I need to install XP to test this, on my Win7 it seems fine but i know XP might cause issues.

        #5016
        atom
        Keymaster
          • Topics: 159
          • Replies: 2945
          • Total: 3104
          • ★★★★★

          Quickly installed XP SP3 in a virtual machine, had no problems with your test panel for opacity, didn’t get any glitches (gave the machine 1 cpu and 512mb of ram).

          #5017
          msepsis
          Participant
            • Topics: 219
            • Replies: 732
            • Total: 951
            • ★★★

            thanks for looking into it. this is an HP D330 w/ a homely pentium 4 proc. Using a moto MTPAV usb MIDI interface. I haven’t had any reason to suspect it but I will try later with a maudio 2×2 usb. Again I get no lag until I enable BOTH midi In and Out devices to the devices connected to/from the actual synth.

            Monstrum Media | Music, Sound & Software Design, Chicago / San Francisco listen

            #5018
            msepsis
            Participant
              • Topics: 219
              • Replies: 732
              • Total: 951
              • ★★★

              Problem persists with any midi interface, and happens on any modulator ([b:2zggmxv8]even on a new fresh panel[/b:2zggmxv8]).. Correction to my earlier findings – this occurs when MIDI out device is set to anything other than "none", not when both in and outs are enabled.

              CPU usage goes from 0 to 10 or 20%, basically ANY midi message that’s sent causes not only (the obvious) MIDI activity LED on my MIDI interface to illuminate but also the system HD light to go.. it’s literally as if my system HD’s activity LED is a second midi activity LED <img decoding=” title=”Neutral” />

              If I set all MIDI In/Out devices to "none" I still get the same behavior with any modulator that is assigned a Lua script (has a script selected under "called when modulator value changes").. any time I rotate a knob that has a script attached, I hear the system disk churn / drive activity light goes off and CPU meter spikes until the moment I stop rotating the modulator.

              HP D330 2.8 GHz Pentium 4, 2 GB RAM, Win XP x32.

              Id be fine with "your machine’s too old/slow" but this issue didn’t appear until somewhere after the 1137 rev and I’ve never ever had this sort of issue with ctrlr on any earlier revs. Something’s wrong.

              Monstrum Media | Music, Sound & Software Design, Chicago / San Francisco listen

              #5019
              atom
              Keymaster
                • Topics: 159
                • Replies: 2945
                • Total: 3104
                • ★★★★★

                Nah, your machine is more then enough.

                I’m outa town again (this time at work again) and i can’t code here (a lot of people are looking at my screen witch suxxx). But once i get a chance i’ll have a look. Can you tell me what XP version you are using, witch service pack etc.

                #5020
                msepsis
                Participant
                  • Topics: 219
                  • Replies: 732
                  • Total: 951
                  • ★★★

                  Thanks atom. I understand the people on your screen thing. keep your job – that’s important <img decoding=” title=”Smile” />
                  it’s XP service pack 3.

                  I have one other windows machine but it’s an exact duplicate of this D330. I could load ctrlr on it and see if I can reproduce it there too just to get some more info. I recently ditched windows 7 on my pentium4 laptop for ubuntu studio as that runs far smoother, boots way faster and is far more reliable when I need to get work done.

                  Also look to my last post for the most conclusive findings.. only midi out needs to be enabled contrary to my original thread subject. If no midi ports are set up I still get this lag/CPU spike/HD reading thing on any modulator rotated that contains even the simplest Lua script.

                  Thanks for being so responsive!

                  -Rob

                  Monstrum Media | Music, Sound & Software Design, Chicago / San Francisco listen

                  #5021
                  atom
                  Keymaster
                    • Topics: 159
                    • Replies: 2945
                    • Total: 3104
                    • ★★★★★

                    I tried this on my VM XP SP3, i tried with a MIDI device and without one (everything on the example for transparency you sent). I can’t reproduce this. There is a CPU spike when i move the knobs but since you manipulate the opacity of components i’d expect that to happen (it spikes to 21% if i move the knob very fast). The CPU does not rise anymore when enabling a MIDI out port (i can see data going out on it, it’s a hardware device connected via usb and it’s blinking it’s led). I need to reproduce this somehow to debug this….

                    #5058
                    msepsis
                    Participant
                      • Topics: 219
                      • Replies: 732
                      • Total: 951
                      • ★★★

                      Can you at least reproduce the HD light activity when you rotate a knob w/ a midi out port enabled? that does not happen in 1102.

                      I wonder if you were to try straight XP (not in a virtual machine from win 7) if you have different results.

                      Monstrum Media | Music, Sound & Software Design, Chicago / San Francisco listen

                      #5059
                      atom
                      Keymaster
                        • Topics: 159
                        • Replies: 2945
                        • Total: 3104
                        • ★★★★★

                        I’ll try to install XP on my company laptop tomorrow and see where that gets me

                        #5109
                        Hecticcc
                        Participant
                          • Topics: 25
                          • Replies: 160
                          • Total: 185
                          • ★★

                          I also experienced laggy behaviour with midi but only when the console is being used for  printing a few things at a time.

                          For instance in a midiReceived script i have some printing to console of channels, numbers and values that reacts on everything coming in for analysing incoming data, and when i  turn off those commands the midi works without a hitch again.

                          If i turn them on again the midi seems to be inert for a second or two and then it spits out everything in quick succesion, as if it’s trying to catch up.

                          This happens on win7 ultimate 32 bit stock (no updates)

                          #5112
                          msepsis
                          Participant
                            • Topics: 219
                            • Replies: 732
                            • Total: 951
                            • ★★★

                            any updates on this issue?

                             

                            Monstrum Media | Music, Sound & Software Design, Chicago / San Francisco listen

                            #5113
                            msepsis
                            Participant
                              • Topics: 219
                              • Replies: 732
                              • Total: 951
                              • ★★★

                              I’m reinstalling winXP on my duplicate, literal clone of this HP D330 tonight.

                              apparently no ability to edit posts with the new site so I’ll just follow up if I have news at the risk of appearing to be having a conversation with myself 🙂

                              Monstrum Media | Music, Sound & Software Design, Chicago / San Francisco listen

                              #5115
                              msepsis
                              Participant
                                • Topics: 219
                                • Replies: 732
                                • Total: 951
                                • ★★★

                                shit i thought I replied last night here. guess not. don’t bother tracking my issue down any further.
                                It was related to my system somehow…. I swapped the actual computer with a freshly installed clone with the same hardware midi interface and it’s actually lightning fast, no lag whatsoever, none of the issues I described earlier in this thread manifested.. probably some windows xp update I ran recently that broke the system or something.

                                going back to my old rule of never putting an xp box on the internet and just running stock sp3 w/ no further updates. Just need to put my dropbox on a NAS.

                                Monstrum Media | Music, Sound & Software Design, Chicago / San Francisco listen

                                #5116
                                msepsis
                                Participant
                                  • Topics: 219
                                  • Replies: 732
                                  • Total: 951
                                  • ★★★

                                  I can completely confirm I’ve licked this issue. I rsynced -ravptz’d my windows xp partition from my backup clone D330 to my “production” D330’s windows partition, rebooted and none of these issues re-occur. My apologies if you spent a lot of time tracking this down.

                                  maybe a good tip to other windows users: if you invest a lot of blood sweat and tears on your audio projects and use windows for any experimental apps or practices I recommend buying two older “shittier” systems, putting XP pro SP3 on em fresh and setting em up with your base apps/drivers then keep one as a backup system and the other as your “production” system…

                                  I tend to prefer to develop on a slower system. When you test on a newer system chances are things will work better than you expect, not worse. That’s always a nice moral boost during debugging… which can sometimes not be so hot on the moral. Besides, older systems can be snagged up for next to nothing.

                                  Monstrum Media | Music, Sound & Software Design, Chicago / San Francisco listen

                                Viewing 15 posts - 1 through 15 (of 15 total)
                                • The forum ‘News and releases’ is closed to new topics and replies.
                                There is currently 0 users and 73 guests online
                                No users are currently active
                                Forum Statistics
                                Threads: 2,495, Posts: 17,374, Members: 77,605
                                Most users ever online was 12 on January 22, 2019 3:47 pm
                                Ctrlr