Virus TI - OS 5.1.1.0 - causing MIDIServer crash

    Just ran BeyondCompare on the 5.1.0.0 driver vs. 5.1.1.0.


    ...but the first obvious difference is 5.1.0.0 driver file is 71,492 bytes, and 5.1.1.0 is 175,136 bytes.


    There is commonality (from the binary comparison), but the 5.1.1.0 version has a big block on the end, of what looks debug string data. It's probably a debug build.


    In which case, it could be that the difference in size and addition of debug has moved a problem around either in memory, or the extra code has changed timing.


    I'm sure Access are aware of that though. The null pointer issue might have been there for a long time, but the new release is statistically more likely to hit it.

    'i think we would have had a lot of people writing into support if there would be a general issue.'


    I'm also experiencing this issue, I wonder if there are others who have also just put up with it?

    Thanks to thetechnobear and all of you for being so deep involved in this frustrating behavier of our loved TI.
    I spent so much time in trying to solve my the USB-Midi Problems with my Polar and my very new MacBook Pro, that i decided to gave up, make music over DIN-Midi and wait ;) The German Support "Jörg" was very patient and professionell but always signify me, i am the only one having this Problems with my machines.
    So the crashing Midi Server issue was nether solved. Byside i have with my old TI2 desktop and an Mac Mini 2012 there where some Strange USBMidi behavier too.
    Now: if something in the Midi Setup changes, the Midiserver Crashes, my Polar Keyboard disapear, but the Synth can be played by other gear eg. Push. this can be 100% reproduced.
    I hope you guys from Access solve this, and my Polar do not loose its place beside my Elektron Keys, Sub 37... and what ever is comming up in my little Studio. Thanks to all, spending so much Time without being paid..
    Greetings
    darius

    More info: that's with Virus plugged directly into the Mac.


    Plugged in via hub doesn't work.


    It's also still affecting the Steinberg dongle (dongle doesn't work with Virus also connected).


    Ableton is OK though.


    So - if I buy a Kanex Thunderbolt to USB3, I'm guessing that I *should* be OK.


    I may have a way forward.

    I had Virus connected directly and the Syncrosoft dongle connected via my Belkin 4 port hub (which is meant to be MTT).


    The dongle doesn't seem to like being on the hub at all. It's like that hub is adding enough latency, that the host thinks something has gone wrong and rejects it.


    Of course, it could just be that this hub is too old hat, even if it is MMT with USB2 connection to the host.


    Which model hub do you have Marc?

    I think I've mentioned it previously, but my ultimate aim for this system is to buy a Thunderbolt PCIe chassis (end of this week hopefully), to host my 3 RME HDSPe cards with 3 Multifaces. The Fireface UC then gets used with my MacBook for gigs.


    That removes some of the USB bandwidth hogging and gives me back my 24 analogue ins and outs.


    At that point, there may be enough spare USB bandwidth to cope with Virus and dongles etc.


    If not, I'll grab a Thunderbolt to USB dongle, which should give Virus its own dedicated connection. ...or I might do that anyhow next month.


    That just leaves dongles and USB hosted MIDI connections to various modules and keyboards connected to the MacPro's USB.


    I'm fairly confident that I'll have something working now that I can see Virus working as a plug-in, with Ableton.

    Quote from Marc

    we've upload version 5.1.2.0 which hopefully fixes the MIDI server related issue reported by some of you. The new build can be found within the public beta section of our download page. I'd be interested to learn how this version works for you.
    best, marc


    Hopefully you have in plan to drastically reduce it's MIDIServer CPU consumption!
    How come in Windows it works fine, but in OS X it doesn't!

    just tested die 5.1.2, works really better, MIDI Server don't crash.
    the Virus (Polar) Keyboard disappear, if something in die MIDI Setup changes,
    like doing this before, but u can restart the Virus and nothing crashes.
    I am almost satisfied ;)


    Marc: DO know this behavior, that the virus midi keyboard disappears in then MIDI CONFIG (OSX) ,
    when an other USB device like an HD is being switch on, although on an complete other USB Port.


    So this update was the right way,
    THANKS Darius

    Quote from DariusW

    Marc: DO know this behavior, that the virus midi keyboard disappears in then MIDI CONFIG (OSX) , when an other USB device like an HD is being switch on, although on an complete other USB Port.


    i haven't seen this so far. it sounds strange thought. the way USB works is that a NEW device cannot connect if power/bandwidth/etc. doesn't suite. it is not like an existing device gets removed. and AMS (i think this is what you mean with midi config) should not care about a hard drive at all. do you see log messages when this happens?
    marc

    Quote from Taavi

    Hopefully you have in plan to drastically reduce it's MIDIServer CPU consumption!


    you have posted this on many occasions and it appears to me that you believe that what either "top" or activity monitor tells you something about the real CPU consumption. that's wrong. it gives you only a figure illustrating how frequently something is required (or "called"), independently on how much CPU the "something" will require in the end. i know, sounds strange but what i'm trying to bring across: the MIDI driver does not tax the CPU a lot, it is just contacted very frequently. read up on that stuff please. it is not that simple.
    marc
    ps while you're at it, read the forum guidelines as well. don't post unrelated stuff.