Logic dropping MIDI OUT during a track

Question

Level 1

10 points

Logic dropping MIDI OUT during a track

Running Logic 8 on a Macbook Pro for live performance. 98% of the time it's working very nicely, but have had a couple of instances where the MIDI info being sent to external processors and lighting controllers is being 'dropped'

There are a couple of additional Audio tracks (software instruments bounced to WAVs) which continue to run and no error message comes up... it just seems that the MIDI instructions stop being transmitted. MIDI IN still stays in operation as Superior Drummer keeps being triggered by the kit without missing a beat

All returns to normal on opening the next song (calling up the next song via OnStage does involve Logic being closed and re-opened with the next song then ready to go) and there's no need to reset anything else

This has happened on two different MIDI interfaces - first time on a USB Edirol UM-1 and second time on a Saffire Pro 40 via FW400, so that would indicate it's not an external hardward issue and with the Logic 'reboot', it would appear to be a software problem

Is this a known issue..?

iMac 24" 3.06GHz 4Gb & MacBook Pro 15" 2.4GHz 4Gb, Mac OS X (10.5.6)

Posted on Jan 18, 2009 1:17 PM

Reply

Answer 1

iSchwartz

Level 5

4,321 points

Jan 18, 2009 2:43 PM in response to Cover'd

There are known issues with Logic playing back MIDI data correctly when playback is initiated from somewhere in the middle of a song, but I'm imagining that when you playback a song it starts from the top. If that's not the case, that might be a culprit. Please confirm.

When you say MIDI info is being dropped, does it always drop out from a particular point in the song(s), or do the dropouts randomly occur?

Once MIDI drops, does it ever come back on those tracks?

Are any of the MIDI regions used for lights and effects looped regions?

Does the MIDI drop out simultaneously for lights and external effects, or have you run into an instance where only one of those drops out?

Is it possible that during one of the songs, MIDI input intended to target Superior Drummer is actually being routed to one of the tracks outputting to your lights & external MIDI gear? That could happen depending on how you have your MIDI routing programmed in the clicks/ports layer of the environment, in combination with the track that's selected when a song loads.

Message was edited by: iSchwartz

Reply

Answer 2

Pancenter

Level 6

14,788 points

Jan 18, 2009 10:12 PM in response to Cover'd

Adding to iSchwartz's comments.

How much automation (if any) is being used?

pancenter-

Reply

Answer 3

Cover'd Author

Level 1

10 points

Jan 19, 2009 1:06 AM in response to Cover'd

Many thanks for your initial responses 🙂

The arrangements are set up in a very consistent way in as much as there are up to 10 External MIDI tracks depending on what a given song requires:

MIDI Ch 1) KORG Triton (program changes and notes)
MIDI Ch 10) Drummer's click & count in click through stage monitors via Alesis DM5 (notes)
MIDI Ch 11 & 12) Guitar processor program changes (program changes)
MIDI Ch 13 & 14) Vocal processor program changes (program changes and notes)
MIDI Ch 11, 12 & 16) Lighting controller scene changes (controllers and notes)

The use of these channels and the reason for some shared channels is a throw-back to our previous set up of external sound modules using Midi Channels 2-9 for instruments (now replaced by internal Software Instruments which have been subsequently bounced to two stereo WAVs). The MIDI channels have been retained in case a sudden revert to the old set up is ever required (this could be changed if necessary however)

Where MIDI Channels are shared, each external processor is only responding to instructions relevant to it's operation (for example, the guitar processors don't react to notes, whereby lighting controllers do)

All MIDI regions run the full length of the song with note, controller and program change events popping up as and when required during the song

To answer each point in turn:

+There are known issues with Logic playing back MIDI data correctly when playback is initiated from somewhere in the middle of a song, but I'm imagining that when you playback a song it starts from the top. If that's not the case, that might be a culprit. Please confirm.+

Indeed, every song starts from 1.1.1.1. and runs to the end. No looped sections

+When you say MIDI info is being dropped, does it always drop out from a particular point in the song(s), or do the dropouts randomly occur?+

No, nothing consistent. One night fine, the next night not and vice versa. It's happened once on two different songs (so far!)

+Once MIDI drops, does it ever come back on those tracks?+

No - but all seems to return to normal when Logic closes one project and opens the next after receiving instruction to do so from

+Are any of the MIDI regions used for lights and effects looped regions?+

No - just a constant stream of events or occasional instructions as and when necessary, but all in a linear format with no loops

+Does the MIDI drop out simultaneously for lights and external effects, or have you run into an instance where only one of those drops out?+

The seemingly all quit together - the Drummer's click (Ch10) disappears (yikes!) and the lighting (11, 12 & 16) stays on the last instruction received etc

+Is it possible that during one of the songs, MIDI input intended to target Superior Drummer is actually being routed to one of the tracks outputting to your lights & external MIDI gear? That could happen depending on how you have your MIDI routing programmed in the clicks/ports layer of the environment, in combination with the track that's selected when a song loads.+

No - it's a very clean set up. There have been no additional routing or port adjustments made. Superior Drummer gets it's MIDI IN from the Alesis DM5 via MIDI Ch10 (MIDI thru disabled in DM5 to stop the loop-back) and the projects are always saved and open when called with that track selected

+How much automation (if any) is being used?+

None at all

Thanks for your questions - I hope it's of some use in eliminating a few possibilities! 🙂

Reply

Answer 4

iSchwartz

Level 5

4,321 points

Jan 19, 2009 2:08 AM in response to Cover'd

OK!

I'm inferring from the info you've supplied a few scenarios. Here's #1 --- you've got a 1-output MIDI interface (ouch!) and you're daisy-chaining (ouch!) all of that MIDI gear. If that's the case, it's entirely possible that the MIDI output is getting into a buffer overload state because it can't handle the bandwidth of all the data you're sending. And/or the buffer overload doesn't occur in the interface but rather, a bottleneck occurs in one of those piece of gear from which the MIDI is being thru'd.

Scenario #2 --- you've got a big ol' MIDI thru box and all of your gear is receiving all data from the single out of your MIDI interface all the time. If that's the case then it's the interface's MIDI out getting into a buffer overload state because it's being asked to send too much data simultaneously.

Is your setup like either one of those that I described?

Even if I'm off the mark here, IMO the two interfaces you mentioned are not adequate for your rig. I would like to strongly, emphatically, unequivocally, unambiguously, and explicitly BEG you to get a *multi-port MIDI interface* and fan out the MIDI data to your various devices that way.

😉

It makes sense to me that if your MIDI goes down mid-song that the system would seem to 'recover' when a new song loads. That would be because the MIDI interface (or the thruput on the device which I'm suggesting gets a MIDI bottleneck) has a chance to 'clear out'. There's the possibility that Logic even resets the MIDI drivers when a new song is loaded (though don't quote me on that).

Anyway, please post back with details of your MIDI setup if the ones I described don't reflect the rig you're running.

Cheers!, Tables, and a Small Settee,

-=iSchwartz=-

Message was edited by: iSchwartz

Reply

Answer 5

Cover'd Author

Level 1

10 points

Jan 19, 2009 2:22 AM in response to Cover'd

You're surmising well 🙂

It's set up option 2 - the MIDI Out is being sent to a self-powered 1 in - 8 out MIDI splitter - very little daisy-chaining save for one or two units

I would agree with your thoughts except that we've run this set up for years on a poxy Compaq lappy / Win 98 / Cakewalk / Van Bascos MIDI Player / MIDI output through Joystick port into the same splitter etc and NEVER ONCE had the same thing happen, not even a hint of it. In fact, due to the external sound modules being used back then, the MIDI steam was actually way, way bigger in that set up!

The only thing which has changed is the addition of a MacBook Pro, Logic and a drastic reduction the MIDI data being transmitted. Surely 'upgrading' the kit in this way hasn't revealed that the MIDI handling capability has actually gone backwards..?

Reply

Answer 6

iSchwartz

Level 5

4,321 points

Jan 19, 2009 2:22 AM in response to Cover'd

Glad to see my frontal surmising lobe is working at 5 AM LOL!

I would agree with your thoughts except that we've run this set up for years on a poxy Compaq lappy / Win 98 / Cakewalk / Van Bascos MIDI Player / MIDI splitter and NEVER ONCE had the same thing happen, not even a hint of it.

You're breaking my heart!!! 😉

OK then... I'm sorry to say, I'm stumped. But I'll keep your problem in mind and if I think of any other possibilities I'll post back and let you know.

Best,

-=iS=-

Reply

Answer 7

Cover'd Author

Level 1

10 points

Jan 19, 2009 2:40 AM in response to Cover'd

Thanks for your assistance thus far 🙂

It does seem crazy that such an upgraded system couldn't handle this - so that's why I think it's more of a software issue

Maybe Logic isn't 'man enough' to handle this MIDI data..? Nope, it can't be that as thinking back, the points at which these songs went funny there would have been little going on - no sudden splurdge of Program Changes etc

Before bouncing the Software Instruments down to two WAVs during pre-production, the MBP & Logic were processing far more MIDI information and the only errors would be the standard 'System Overload' type until the song had been run through once - nothing like the same random drop out we've experienced twice (different song, different night, different MIDI interfaces) to date

Of course bouncing all the Software Instruments to two WAVs has reduced the load on the system to practically nothing now too...

Humm..?

It almost seems as though communication with the MIDI OUT port gets lost / broken... but the MIDI IN stays connected and fine... that in itself points to Logic and not hardware... humm again!?!?!

Reply

Answer 8

Bee Jay

Level 6

10,900 points

Jan 19, 2009 4:49 AM in response to iSchwartz

Glad to see my frontal surmising lobe is working at 5 AM LOL!

Hehe. Hey iS!

As to the OP's issue. Hmm - I've never had Logic drop MIDI data before (although I did get SIDS last night - god, I thought that one was nipped in the bud a long time ago...!)

You've been troubleshooting well. What I'd try to do is narrow down when the MIDI stops - is Logic not sending it to OSX, or is OSX not sending it to the MIDI driver, or is the MIDI driver not sending it along the MIDI interface.

Ruling out Logic is the easiest - try setting up a similar test case in a different app (say, the Ableton Live demo) and seeing whether it still happens - if it does, it's most likely MIDI interface/driver related, and a different interface might cure it.

You could set up a virtual MIDI monitor through the IAC bus or MIDIpipe and seeing whether Logic still sends MIDI along these when it "drops". This would indicate that it's not Logic that's dropping, but the data is being "lost" somewhere after Logic and before the interface. This might be tricky though, because it may only drop on that particular port with the interface drivers on, but it's something else to look at.

Lastly, check your versions (OS version, Logic version, MIDI driver versions) and make them up to date if not. Trash the Logic MIDI caches and AMS/MIDI prefs, trash your MIDI drivers and reinstall them. Does this cure it? Could you try a different MIDI interface (borrow one if necessary) to rule out anything funky with the drivers on your system? Also, try it in a different (new) user account, to rule out anything funky with your user on your system.

You could also upload a test song and we could try it out to see if we could replicate the loss of MIDI on our systems.

Some ideas to pursue, anyway...

Reply

Answer 9

Eriksimon

Level 6

15,428 points

Jan 19, 2009 5:46 AM in response to Cover'd

How old are your MIDI-interface and 1 to 8 splitter? Maybe some IC's (those handling OUT) are getting old/defective/tired/burned out...? The only way to test this unfortunately & obviously is to get your hands on new or known to work (similar) hardware (borrowing it from a fellow MIDI-er or your 'regular' music store perhaps?) and hook that up to your rig.
Hardware can malfunction overnight... like any technology...

Oh, and another unasked question: do you have any other apps open? Is your machine connected to the internet?
Have you used Activity Monitor to see if anything unusual is happening with any apps at the time of the drops? Did any of your processors spike?

regards, Erik.

Reply

Answer 10

iSchwartz

Level 5

4,321 points

Jan 19, 2009 7:17 AM in response to Bee Jay

Hey Bee Jay, nice to see you. 🙂

Did you get SIDS on an external MIDI track?

I get the equivalent of SIDS all the time with instrument tracks all the time. The behavior hasn't changed since L7. 😟

I concur with your idea that the OP upload a test song. I'd like to see what's going on myself. This is a very interesting problem.

Cheers!

-=iS=-

Reply

Answer 11

Bee Jay

Level 6

10,900 points

Jan 19, 2009 8:01 AM in response to iSchwartz

Did you get SIDS on an external MIDI track?
I get the equivalent of SIDS all the time with instrument tracks all the time.

Software Instrument track - basically, once a plugin has killed that channel, it stays dead, even loading other instruments into it. Haven't had it for a long time, but it's (sort of) reassuring to know that it's still lurking there ready to strike at any time. Forewarned is forearmed yada yada...

Perhaps I should start http://www.thedefinitivelogicbugandworkaroundsite.com once and for all... 😉
(with you as top contributor, of course! 😉 )

Steady now - fermusic's on the case - he'll ~~confuse~~sort out the OP's problem in no time. :o

In this case, my gut says it's probably not Logic, but something system related. With a clean OSX and a reputable MIDI interface/driver combo, I'd be willing to lay (some) money that the OP's problem would go away. But that's just Troubleshooter's Knee at work - it will be interesting as the OP seems to have great troubleshooting and dialog/feedback skills to get to the bottom of it.

Reply

Answer 12

iSchwartz

Level 5

4,321 points

Jan 19, 2009 7:59 AM in response to Bee Jay

Bee Jay wrote:
Software Instrument track - basically, once a plugin has killed that channel, it stays dead, even loading other instruments into it.

Oh, that ol' chestnut. Right. Yes, I've had that happen too. Only happens once in a while.

Perhaps I should start http://www.thedefinitivelogicbugandworkaroundsite.com once and for all...

I'll split the cost of the URL with ya LOL!

Steady now - fermusic's on the case - he'll ~~confuse~~sort out the OP's problem in no time.

"This is your captain speaking. Brace for impact!"

Reply

Answer 13

Eriksimon

Level 6

15,428 points

Jan 19, 2009 8:37 AM in response to iSchwartz

iSchwartz wrote:

"This is your captain speaking. Brace for impact!"

My plane always flies no problems. Without using checklists! 😉

Reply

Answer 14

Jan 19, 2009 8:49 AM in response to iSchwartz

iSchwartz wrote:

Steady now - fermusic's on the case - he'll ~~confuse~~sort out the OP's problem in no time.

"This is your captain speaking. Brace for impact!"

Are you sure man?
I always give personal experiences

Everything is always tested!

Don't spread false info about me and replay with Technical report
you will be surprised about my technical skill level

thanks

Reply

Answer 15

Cover'd Author

Level 1

10 points

Jan 19, 2009 10:12 AM in response to Cover'd

Wow! Great info and input here guys! Thanks a lot for taking the time to respond 🙂

I'll freely admit I'm a total newb to Mac & Logic (three months now) and a lot of the technical questions (IAC bus & MIDIpipe?!?!) are way above me... at the moment! I'm sure I can learn how set up some tests with a bit of direction

With the virgin state of the system in mind then, I can confidently say that the OS is running pretty clean (10.5.6), the Drivers for the Saffire Pro 40 are spanking new as it's a very new product. The Edirol UM-1 is a basic USB interface which has been around a long time now and I have to say that everything has run very smoothly with the exception of these two isolated, differing and inconsistent incidents. Having said that, a total driver / prefs flush and reinstall does seem like a good suggestion... however I go about doing that?!?!

I understand what has been suggested with a hardware malfunction, but my experience with this sort of stuff (MIDI splitters etc) is that they work well, very erratically or not at all. The issues I've had were on pretty simple songs of modest 3 & 4 minute length, where a couple of 9 & 11 minute ones have worked without hitch... and as I say, it's ONLY the MIDI OUT which gets affected - the MIDI IN and the Audio playback don't miss a beat

The only apps which are running are OnStage (which simply responds to Prg Ch 1-20 from an external program change instruction to close the previous Logic project without saving and opening the next to be ready to go) and Logic itself. That's all - no iTunes or anything, Airport is disabled and Time Machine has not been activated on that machine

I have not monitored the System Activity - just the CPU & Disk I/O in Logic (hardly moving either one)

I'm perfectly happy to email / upload the two projects for someone else to look at. As for finding someone else nearby with a similar set up - not a chance! We're the only act I've ever come across that uses these things to this extent!

Clive

Reply