Multiple MIDI messages with iO Dock and FCB1010

edited December 2012 in Equipment

Bad news/good news, of course it may depend on if you're a glass half empty or half full sort of person :)

A little background: I was trying to program the FCB1010 to send multiple MIDI commands (5 PC/2 CC/1 Note on-off) to be able to automate multiple bindings into 1 foot press.

Good news: the FCB1010 can be programmed and will send all 7 (5 PC/2 CC/1 Note on-off).

Bad news: they don't all make it to the App. Some are intermittent, while others are 100% failure.

Good news: I have not had a failure when only 1 MIDI command is sent from the FCB1010.

Now an explanation of what I have found by trial and error and experimentation.

Hardware:

iPad 3 v6.0.1, Alesis iO Dock v1.0.7, FCB1010 v2.51e, M-Audio MidiSport 1x1

Software:

iPad - Loopy HD v1.3.1, MIDI Monitor v2.0

PC - Windows XP SP3, FCB1010 PC Editor v2.99-6, Bome's SendSX v1.22, MIDI-OX v7.0.2.372

connected:

FCB1010 connected to iO Dock, PC connected (USB) to iO Dock (basically MIDI Thru to the PC)

FCB1010 connected to MidiSport connected (USB) to PC connected (USB) to iO Dock

  • The following results:
    • All MIDI commands from FCB1010 show on PC (MIDI-OX or SendSX)
    • some command from FCB1010 are intermittently missing in MIDI Monitor display and expected response in Loopy is not correct.

PC connected (USB) to iO Dock

PC connected (USB) to MidiSport connected to iO Dock

The following results:

  • manually sending a MIDI stream from SendSX with 1ms delay between commands, the following results:

    • All MIDI commands are displayed in MIDI Monitor
    • expected response in Loopy is not correct
  • manually sending a MIDI stream from SendSX with 50ms delay between commands, the following results:

    • All MIDI commands are displayed in MIDI Monitor
    • expected response in Loopy is correct

Conclusion:

  • the iO Dock and/or the iPad can not accept MIDI commands as fast as the FCB1010 sends them (based on MIDI Monitor and Loopy)
  • Loopy does not respond reliably to MIDI commands sent less than about 50ms apart (this is purely an observation. I have read posts that iOS does not put MIDI at the top of its List)

I do not have another way to get MIDI commands into my iPad from the FCB1010, and cannot send them any faster with SendFX (1ms delay between) to further narrow down the problem. Alesis has taken quite a beating for it's iO Dock, hanging and missing notes. They have not exactly been open and forthcoming about the issues in the past (shame on them!), pointing at Apple while Apple points at them...

I don't know what Michael might be able to do at the Application level. It might be worth while contacting the Yahoo FCB1010 Forum. Maybe there's a possibility of a MOD for the FCB1010 to configure sending speed (just thinking out loud)

Comments

  • Imagine how good the world's software would be by now if everyone sent in bug reports and requests with this level of detail. Nice work @ganthofer.

  • Tell me about it =)

    Thanks heaps, @ganthofer. Loopy does have a bounce filter in its MIDI implementation, which might possibly be causing problems. I've since decided to remove the filter totally (probably an overly complex solution to a nonexistent problem, to be honest), so it'll be interesting to see what happens in the next version.

    Given that there are messages missing in MIDI Monitor, I'm inclined to think that I'm not going to be able to do anything about that - chance are it's either a hardware or an OS problem, as you say. But, I'll be able to fix the 50ms problem, which is what the bounce filter will be causing, although that should've only applied to multiples of the same message.

  • @Michael - I believe the reason I settled on 50ms, is because I tried sending a number of different message combinations ( some were multiples of the same message - toggle mute - to immediately stop playback when count-in/out is on ), and that was when all of the missing/dropped messages stopped happening. So I'll be interested in testing it again with a future release.

  • That'll almost certainly be it then - my bounce filter threshold was actually exactly 50ms.

Sign In or Register to comment.