Virus TI - OS 5.1.1.0 - causing MIDIServer crash

  • excellent news about 5.1.2.0 I will install it on my iMac and MPB and check to see if the issue is resolved.


    agreed, the other issues should be discussed on the appropriate existing theads :


    Disappearing Virus : mac osx - virus midi device disappears
    Summary: this has been going on for a long time in various version at least every 5.x.x version Ive tried!)


    MidiServer CPU usage: OS X Midiserver CPU usage
    Summary: I don't get this issue - but from this thread it looks highly likely its caused by excessive active sensing midi messages, possibly Snow specific?
    @Marc whist what you say is true, the Time is a good indication of actual load, and this was also high... as I pointed in the thread usbd* also looked 'higher than normal', again an indication the active sensing is an issue... of course the question is why so much active sensing... this could be caused by the Virus see issues on usb traffic... the only way to know is to ask the developer to check the code!


    (* yes I'm very well aware usbd is used by other drivers too... so it can only be an indication, but sometimes we have to look for other 'supporting evidence')


    Sadly both of the issues were reported a very long time ago, and neither were resolved, so of course they keep cropping up on other threads... they people believe are related, its a pity the issues are not being followed through to a conclusion which would result in less frustration.


    anyway, thanks again for looking in the MidiServer issue, (Id be interested to know the cause/fix, id your willing to be open about this.. out of idle curiosity), I look forward to testing this later today :)



    Mark

  • Marc whist what you say is true, the Time is a good indication of actual load, and this was also high... as I pointed in the thread usbd* also looked 'higher than normal', again an indication the active sensing is an issue... of course the question is why so much active sensing... this could be caused by the Virus see issues on usb traffic... the only way to know is to ask the developer to check the code!


    we will not look into this further. we know why this is and we also know why it is necessary. unless apple fixes something we will have to go this way. nevertheless, we'll verify if apple has fixed the _something_ in 10.10 or 10.11. if so, we can use their function again and the "load" (which is not load but frequent calls) will be gone. let not make something a problem just because we want to believe that it is one.
    marc
    ps. there will be a 5.1.3.0 today - we broke Virus Control Center in 5.1.2.0 by accident. sorry.

  • 5.1.3.0, cool will wait for that.


    TIME - is a measure of time spent 'on' the cpu by that process (which most would call 'load') , its not frequency of calls.. sure if you make frequent calls (or request frequent callbacks) then this will increase as a result of time spent in functions (particularly noticeable for expensive funcs like kernel calls) and context switching. but regardless, by making such calls you are increasing the amount of cpu time the MIDIServer is using (even if its not directly in your driver code).


    thats no reflection, on it being necessary* ... if its a workaround thats fair enough... but that was never explained on the previous thread, perhaps more transparency about the underlying 'problem' and the workaround would allows users understand/appreciate the necessity. Ive usually found my users are pretty sympathetic when you explain things to them (and conversely are frustrated when you don't)


    * it was simply stated, that the cpu time was high, and that this did not appear to be case on either your machine or mine, so looked 'different'


    'let not make something a problem just because we want to believe that it is one.' is nonsense, answers/solutions yes.. problems no.

  • 5.1.3.0 installed and looking good so far, no midi server crashes, and working well :)
    whats more looking in the system logs, it looks like the driver is spitting out more diagnostic msgs which is a good thing too...


    @Marc, so I had the disappearing midi device, when I turned on the Ableton Push
    (exact sequence below)
    Interestingly, it spat out the following error at the same time, perhaps something for the devs to go on (as previously there was no error/log msg when it disappeared)
    10/6/15 18:54:11,000 kernel[0]: USBF: 358.244 IOUSBCompositeDriver::ConfigureDevice Could not open device (Virus TI)


    my sequence for producing this:
    - power iMac up
    - turn on virus with USB plugged in
    - (I then tested the USB midi was working, and it was fine)
    - turn on ableton push *. (whilst I had console open, to see if I could see any messages)
    bang.. above message, and Virus Synth Midi device disappears.


    * this seems to happen with pretty much any USB device, and only appears to happen if:
    a) its a new device i.e. in the above case, if I restart the virus, I can then switch the Push on and off, and the Virus wont disappear.. only if its the first time since a reboot its seen the device.
    OR
    b) the mac has just woken from sleep


    as previously stated, not big issue, since I try to remember to power the Virus up last after all other usb devices, just a pain if half way thru a session I decide to use a USB pedal (or similar) that I don't often use.. as when I restart the Virus, Ableton (other daws don't) insists on given the Virus a new midi name e.g. Virus Synth #2, which means I need to go thru all relevant tracks and change the input/output routings!

  • unless apple fixes something we will have to go this way. nevertheless, we'll verify if apple has fixed the _something_ in 10.10 or 10.11. if so, we can use their function again and the "load" (which is not load but frequent calls) will be gone. let not make something a problem just because we want to believe that it is one.

    Thanks for mentioning. Can you please tell in which version of OS X that problem didn't happen? Perhaps I could downgrade my system.

  • I think I know the reason, why Access has pulled down the 5.1.3.0
    I noticed, as lot of users, that all the driver part was going better, but what I didn't notice was for example the aftertouch issues in synth.
    I did test and installed back the firmware 5.1.1.0 and then everything works fine. When I update to 5.1.3.0, the in same patches aftertouch will stop working after playing a while (10-20 minutes).
    Since I found it during my jamming session (without computer), I think there can be more issues there, if used with Logic or Live.

  • Maybe it's uncovered some other horrible-ness in the software?
    That could be a good thing in the long-run if there's some corruption or similar lurking in there which can now be investigated.
    Hope we get an update!


    sorry for the delay guys. nothing bad happened, it is just vacation time for some of us and we don't run at full speed right now.
    marc

  • I've managed to successfully install and implement 5.0.8 by bypassing the firmware check (kept the Virus off) upon install & reboot. Switching over to Studio One 3 might have helped too; it manages Virus TI perfectly! Look forward to the next stable TI release.