These forums are read only, please use our new forums here.

Main :: BackTrack / MIDI Mobilizer



Active Sense signal "killing" MIDI Mobilizers?
by Ruben_cc on 2010-12-26 02:45:13

I addition to two other topics about problems with my MIDI Mobilizers, I'd like to start this new topic about a possible bug in the MIDI Mobilizer firmware.

I've had some troubles with my MIDI Mobilizers not responding to any incoming data anymore, even though the LED indicated that some data was coming in. I tried three different apps on both my iPad and my two iPhones and they all suffered from the same problem of the MIDI Mobilizers "dying" at some point. Therefore I figured it MUST be a problem with the MIDI Mobilizers, not the apps.

Furthermore, because the LED on the hardware DOES respond to the incoming signal, but it is not detected by any app, I figured it must have something to do with the firmware between the MIDI In and the Dock connector on the MIDI Mobilizer. Somehow the signal just doesn't get passed on.

By doing more extensive testing, I found out that the problem only occured on two of my three controller keyboards. Another thing these keyboards have in commong besides killing the MIDI Mobilizers is that they both send out some signal a few times per second because the LED on the MIDI Mobilizers are constantly flashing, even while not touching anything. A third keyboard (StudioLogic) doesn't send out this signal and also does not kill the MIDI Mobilizers.

After that, I connected the MIDI Mobilizer to all my other hardware (synth modules, sequencer, MIDI patchbay) and watched if the LED was flashing or not to determine which device was sending out this unknown signal and which was not.

I compared all the MIDI Implementation Charts of all the devices and soon found out that this "unknown signal" must be the "Active Sense" signal (FEH) that some devices send out every 250 ms.

So my conclusion in the end is that somehow the MIDI Mobilizers (or at least the firmware) doesn't really like the Active Sense signal and at some point even seem to crash on this.


Can anyone confirm having the same issues? Do you ever have your MIDI Mobilizer "die" and stop responding to incoming note events and does your MIDI Mobilizer also constantly have a flashing LED while connected to your other MIDI hardware?

At least I'm glad I found out what is causing the troubles, because I have a keyboard left that is not sending out the Active Sense signal, but this is an old and very basic keyboard and I'd very much like to be able to use my Edirol PCR-M80 again.



Re: Active Sense signal "killing" MIDI Mobilizers?
by willthemoor on 2010-12-26 16:08:27

I'm having the same problem - input lights continue to flash on the mobilizer hardware in response to MIDI messages but apps stop responding at some point, usually a minute or so. This includes the MMR app. This only happens with my Matrix-6.

I came to a similar conclusion - active sensing. The Matrix 6 allows you to  turn off Active Sensing and sadly, it still crashes after a short while. That  said, my SQ SixTrak works without issue and doesn't have any active  sensing options so if Active Sensings is really the culprit, perhaps there's a problem with the matrix.

If I connect the midi mobilzer to my computer (matrix->midisport  8x8->mac->midi patchbay app->midisport  8x8->midi mobilizerr->phone), it works like a charm. I'm guessing something is being  filtered but when I look at the MIDI messages via a MIDI Monitor application, it all  looks exactly the same (in vs out) so I'm not sure what it is.



Re: Active Sense signal "killing" MIDI Mobilizers?
by Ruben_cc on 2010-12-26 20:01:11

Hi Will (I guess),

Thanks for the reply. I was "fooled" the same way by my MIDI monitor software not showing anything different. I guess it just doesn't show the active sense signal because it would result in a constant stream of 4 events every second, making it hard to monitor "real" MIDI events. I really hope that the Active Sense signal is THE problem because this would possibly lead to a solution using a firmware update. I like the MIDI mobilizers, but it kinda sucks am I unable to use my best controller keyboard (Edirol PCR-M80) and one of my favorite synth modules (Korg 01R/W) with it. Lets hope the tech guys at Line 6 can take a look at this. :-)



RE: Active Sense signal "killing" MIDI Mobilizers?
by Line6Hugo on 2010-12-28 10:26:34

Some older synths (Yamaha) and MIDI interfaces have a feature called active sensing that will appear like constant MIDI flashing in your interface.

Is the active sensing hindering your ability to use the MIDI mobilizer? Are you still able to do what you are trying to do?

Regards,

Line6Hugo



Re: RE: Active Sense signal "killing" MIDI Mobilizers?
by Ruben_cc on 2010-12-28 12:16:04

Hello Hugo,

The Active Sense signal itself is not so much "hindering" the things I want to do but I think it is actually disabling the MIDI Mobilizer at some point, most probably the software/firmware inside it.

As mentioned above, the MIDI Mobilizer stops functioning entirely after a while when used with two of my three controller keyboards; the Edirol PCR-M80 and the M-Audio Keystayion 49e. It also doesn't particularly "like" my Korg 01R/W module, which manages to "crash" the MIDI Mobilizer from time to time when doing SysEx dumps (to the MIDI Memo Recorder).

I have no problems whatsoever when using my old StudioLogic CMK 149 controller keyboard or any other sound module or my sequencer.

The most suspicious thing the two controller keyboards and the Korg 01R/W have in common (after comparing all the MIDI implementation charts) is that they all send out an Active Sense signal and all my other gear is not. This is also (as you wrote) most probably the signal that makes the MIDI Mobilizer LED flash constantly, even when not touching or playing anything.

So it's not actually this signal itself that is preventing me from using the MIDI Mobilizer, but when any of these Active Sensing devices is connected, the MIDI Mobilizer just stops functioning after a short while (few minutes or sometimes not even a minute).

By "stops functioning" I mean that the LED indicates an incoming MIDI event (changes from the steady 4 times a second flashing to an irregular flashing that matches events I send to it), but the app used a that moment isn't detecting anything incoming anymore even though the MIDI Mobilizer itself IS detected. Apps I tried (and all behave like this) are Nano Studio, Music Studio and the Line 6 MIDI Memo Recorder. These all indicate they "see" the MIDI Mobilizer, but no MIDI data get "through" even though the MIDI Mobilizer itself (the LED) is clearly indicating something is going in.

So, to cut this way too long story short (trying to be complete here); since this problem occurs on both my MIDI Mobilizers, with three different apps, on both my iPhones (3G and 3GS) and my iPad but only when used in conjunction with a device that uses Active Sensing the one and only conclusion I could come up with is that the software/firmware inside the MIDI Mobilizer somehow crashes or otherwise gets disabled, preventing it from passing the incoming  signal on the MIDI in to the device on the dock connector side.

To be complete for troubleshooting's sake, I have:

Two MIDI Mobilizers with the latest updates according to the MIDI Memo Recorder app.

One iPhone 3G, not jailbroken, with iOS 4.2.1 (8C148)

One iPhone 3GS, jailbroken, with iOS 4.1 (8B117)

One iPad, not jailbroken, with iOS 4.2.1 (8C148)

Normally I connect all my stuff through an Opcode 8Port/SE interface, but for testing and troubleshooting I used direct connections also.

I switched all four of my MIDI Mobilizer cables to the MIDI-In to ensure I was not using a faulty cable.

Today I purchased a second-hand (obviously) Roland D-20, next Thursday I'll be getting a Yamaha DJX keyboard. I'll do some testing with these keyboards too when I find the MIDI implementation charts of these. Meanwhile, if you need any further information or testing from me, please just let me know! Maybe I'll make a little video showing the problem one of these days.

Thanks.



Re: RE: Active Sense signal "killing" MIDI Mobilizers?
by Ruben_cc on 2010-12-28 15:54:17

Made a quick 'n' simple video showing the problem: http://www.youtube.com/watch?v=UqplyzwsHrg



Re: RE: Active Sense signal "killing" MIDI Mobilizers?
by Ruben_cc on 2010-12-29 05:04:34

Some more info. Sorry if all these messages seem like spamming, I just want to be as complete as possible, so I am constantly testing and monitoring all kinds of things, using different hardware and setups.

Today I connected my Roland D-20 to the MIDI Mobilizer. According to the MIDI Implementation Chart the D-20 does not send out Active Sensing. It does however send out a constant MIDI Clock Signal, so the LED on the MIDI Mobilizer stays on permanently.

When connecting the D-20 directly to the MIDI Mobilizer, I can send as much program change events as I want, even while playing notes. This has no effect on the functioning of the MIDI Mobilizer and everything is working like expected for as long as I want.

When connecting the D-20 THROUGH the Edirol PCR-M80 (so introducing the Active Sense again) the MIDI Mobilizer stops functioning again after sending some program changes from the Roland D-20, either while playing, or just sending Program Changes while NOT hitting any notes.

I think this once again proves that there is some problem with the Active Sense signal COMBINED with other MIDI events (other than note on/off events).

The Active Sense by itself does NOT cause a problem. I can leave the Edirol controller connected to the MIDI Mobilizer for minutes, constantly "feeding" it an Active Sense Signal, and it still works like expected when I play some music. As soon as I start sending Program Changes with it, it somehow "dies" again. Sending program changes from other controllers does not affect the MIDI Mobilizer, so it only causes problems when send along with an Active Sense signal.

Furthermore, I've been testing all kinds of Controller Codes and this does not seems to affect the MIDI Mobilizer. But one of my controller knobs on the Edirol was sending out a Channel Aftertouch message and this IMMEDIATELY disables the MIDI Mobilizer again.

:

: Edit 1

:

Now things get a little weird...

I was thinking of a way to "strip" the Active Sense signal out of the MIDI data send out by the Edirol, but I am a little limited in the hardware available to me. Just for testings sake, I put a Anatek Pocket Merge box between the Edirol Controller and the MIDI Mobilizer.

The MIDI-In LED on the MIDI Mobilizer started to blink at twice the rate it does without the Pocket Merge box. It almost seemed like it was now receiving two Active Sense signals? However... the MIDI Mobilizer does NOT "crash" anymore when putting the Anatek MIDI Merge box between the Edirol controller and the MIDI Mobilizer!

So I am now looking for a way to determine if the MIDI Merge box is in fact stripping the Active Sense from the signal. I only have some MIDI Monitoring software on my laptop, but this doesn't show Active Sense in the first place and furthermore it takes data from the USB connector of the Edirol, not directly from the MIDI port.

:

: Edit 2

:

The problem with the MIDI Mobilizer "crashing" on a Channel Pressure event seems unrelated to the Active Sense signal. When sending a Channel Pressure event to the MIDI Mobilizer through my Kawai Q-80 sequencer, the MIDI Mobilizer "dies", even though the Q-80 does not pass on the Active Sense signal nor generate its own. Furthermore, if I play back the recorded Channel Pressure event from the Q-80 to the MIDI Mobilizer without anything else connected, the MIDI Mobilizer still stops functioning. If I set the Q-80 to NOT record Channel Pressure (but it will still pass this on when it is received), the MIDI Mobilizer crashes when the Channel Pressure is played live (and thus passed on) but not when the specific part is played back from the Q-80 (with the Channel Pressure not recorded by the Q-80).

Finally, since the Q-80 does not seem to pass on nor generate an Active Sense signal, I can not get the MIDI Mobilizer to crash when sending lots of program changes to it through the Q-80, not even from the Edirol, which does have this problem when used directly connected to the MIDI Mobilizer. Once again showing problems with Program Changes AND Active Sence and no problems with the exact same events without the Active Sense.

So at this moment I seem to be having two separate ways of crashing the MIDI Mobilizer:

1. Sending Program Changes combined with an Active Sense signal. (crashes after a few seconds)

2. Sending Channel Pressure or transport control codes. (crashes immediately)

Hope this information can add to troubleshooting this strange behavior...



Re: RE: Active Sense signal "killing" MIDI Mobilizers?
by willthemoor on 2010-12-29 11:59:41

Thanks for all of this research and posting Ruben.

If I send channel pressure directly into the line6, it crashes for me too  (from a matrix-6). If I route the matrix-6 through my computer via the  setup mentioned above, I can send channel pressure all day without problem. I don't get  it.



Re: RE: Active Sense signal "killing" MIDI Mobilizers?
by Ruben_cc on 2010-12-29 14:27:53

I don't really get it either. It's all very weird. I hope the tech guys at line 6 can reproduce this problem and solve it.

A little extra info I just found out...

The MIDI Mobilizer also totally locks up when selecting songs on my Kawai Q-80 sequencer. At first, this didn't seem related to the other problems, but then I suddenly realized; these MIDI messages causing problems (Program Change, Channel Pressure and Song Select) are all 1-byte messages, while the other MIDI events not causing trouble are 2 byte messages (Like note on/off, controller data, etc.). Maybe I'm just shooting in the dark here, but personally I find it somewhat coincidental that, so far, all data causing problems is 1 byte MIDI messages. So, since the MIDI Mobilizer crashes on Channel Pressure (one byte), I'm now going to search for a keyboard that can send out Key Pressure (two bytes) and see how the MIDI Mobilizer responds to that. :-)



RE: Active Sense signal "killing" MIDI Mobilizers?
by ricksox on 2010-12-29 15:28:41

That's really strange behavior. We were told that active sensing should not interfere with the MIDI Mobilizer's ability to interface with apps so I'm not sure what's going on.

I've forwarded your information to our QA department and to our developers so they can have a look for themselves.

I'll post back when I hear from them.

Line6Miller



Re: RE: Active Sense signal "killing" MIDI Mobilizers?
by Ruben_cc on 2010-12-29 16:56:15

Thanks for forwarding the info I found out. It is in fact "strange behavior" and at some point you were told right; the Active Sense by itself does "not interfere with the MIDI Mobilizer's ability to interface with apps", only as soon as some 1-byte MIDI events are introduced along with it (or rather in between it) something goes wrong. I'm not familiar with how exactly firmware works, but I believe it is "just" some form of software and as far as I can figure out it seems to "crash" at some point.

According to MIDI specifications the Active Sense signal should NOT be send out any longer as soon as other MIDI events are send and continue to send as soon as no other MIDI events are send out for at least 300 milliseconds. Unfortunately there is no way for me to check if the Edirol controller is behaving according to MIDI specifications, but I've been doing a lot of reading on this and many sites report that Roland is one of the few companies that is very persistent in using Active Sensing in almost all their product, so I guess by now they know what they are doing and that they are doing it right (Edirol is a Roland devision). Furthermore I think that receiving devices (like the MIDI Mobilizer) should not crash on bad data, even if it would not be 100% compliant to the MIDI specifications.

Anyway; I can now fix my personal problem by putting the Anatek MIDI Merge box I wasn't using anyway before the MIDI Mobilizer, but it would be nice if I could use both MIDI Mobilizers (I have only one MIDI Merge box) because I was planning a little project creating music with my three iOS devices all running NanoStudio. I have yet to buy a third MIDI Mobilizer, but I think I'll wait a while until the problem is solved (or just but two more MIDI Merge boxes along with it). :-)



Re: RE: Active Sense signal "killing" MIDI Mobilizers?
by Ruben_cc on 2011-01-03 04:30:10

Hi, I've got yet some more info on this weird problem which might be interesting but making it even weirder...

I connected the MIDI Mobilizer to a Yamaha DJX (PSR-D1). According to the MIDI Implementation Charts, this unit also sends out Active Sense. I'm not sure exactly -what- it is sending, but it seems a bit more than just Active Sense, since the MIDI Mobilizer LED doesn't just blink like with the other Active Sense sending devices but just stays on permanently. Apparantly the DJX is sending out a whole lot more than just an Active Sense signal every 250ms.

Anyhow, I could NOT manage to have the MIDI Mobilizer crash with the Yamaha DJX! I've send a shitload of controller, pitchbend and program change info to it while playing notes, but it just continued to perform like expected. Even NanoStudio sometimes got "confused" by all the rubbish I was sending it, but the MIDI Mobilizer itself just kept on going.

This somehow undermines my theory about the Active Sense signal causing the problems but I'm also a bit curious about the other info that the Yamaha is sending out since the LED on the MIDI Mobilizer is behaving differently. I'll be trying to hook up the Yamaha to my laptop with a MIDI monitor later on.



Re: RE: Active Sense signal "killing" MIDI Mobilizers?
by Ruben_cc on 2011-01-03 06:05:00

Again, I have some very interesting new findings on this weird problem. And once again it is full of theories, but hey, the entire world of science is based on theories, so it must be something good. ;-)

I found out that the stream of data the DJX was sending out was Clock Data. According to MIDI specifications this would mean that there should not have been an Active Sense signal after all, because this is only send when NO OTHER MIDI DATA is send out. By switching the setting from sending Clock Data to receiving it, the Yamaha DJX stopped sending the Clock Data and (according to the LED on the MIDI Mobilizer) started to send out Active Sense.

So, I started playing some notes, sending out Program Changes while doing so and kept on doing this for about five minutes. The MIDI Mobilizer wouldn't die. (???) Weird... once more undermining my Active Sense + Program Change theory.

Then I noticed on the MIDI Monitor that the Yamaha was not just sending out one simple Program Change event, but it first sends out 9 different controller values, then a Program Change, then 9 controller values for another channel, then again one Program Change (also another channel) and then once more 13 controller values on many different channels. So every single Program Change from the DJX results in a total of 33 MIDI events! I could not find any way to stop this behaviour, but I was thinking that the problem with the MIDI Mobilizer -must- be when receiving an Active Sense along with just one single-byte MIDI event.

So then I tried the other way around; have the troublesome Edirol controller behave like the Yamaha DJX. Connected the MIDI Mobilizer to the Edirol and started playing some notes. While sending Program Changes to the MIDI Mobilizer, I was turning one of the controller knobs like a madman, causing a constant stream of lots of MIDI Data. The MIDI Mobilizer just kept on going, not crashing anymore! I continued sending Program Changes to the MIDI Mobilizer and stopped turning the controller knob. Within a few seconds the MIDI Mobilizer was dead again!

So I guess this pretty much brings back the theory about the Active Sense signal followed by a single-byte MIDI message like Program Change (or the other way around). By sending a constant stream of controller data, the Active Sense must stop because there is other MIDI Data flowing at a rate higher than one event every 250 ms. This is exactly what must have been happening on the Yamaha DJX. Because it sends out 33 MIDI events at once with each program change (with the actual Program Change being in between this data), the Program Change itself didn't get followed or preceeded by an Active Sense signal. That's (according to my theory) why the MIDI Mobilizer didn't "die" using the Yamaha.

Hope this will help. :-)



Re: RE: Active Sense signal "killing" MIDI Mobilizers?
by willthemoor on 2011-01-03 21:35:35

@ruben, Just wanted to say thanks again for all this investigating and writing up.

@Line  6,  Have you been able to replicate it?



RE: Active Sense signal "killing" MIDI Mobilizers?
by ricksox on 2011-01-11 12:23:57

Hello,

We've been able to reproduce this.

Line 6 is currently putting the finishing touches on a firmware and SDK update for the MIDI Mobilizer, in order to address user issues with MIDI Active Sensing messages and developer requests regarding MIDI System Real Time messages. We plan to release the update soon, in a new version 2 of our free MIDI Memo Recorder app. MIDI Mobilizer users should watch for the MIDI Memo Recorder app update, and run it when it arrives. We're hopeful by the end of this month but we have no specific ETA.

Line6Miller



Re: RE: Active Sense signal "killing" MIDI Mobilizers?
by Ruben_cc on 2011-01-11 13:51:14

That's great news! :-D I recently tried to "kill" the MIDI Mobilizer with my "new" GEM WX2 but couldn't. I was just about to figure out "why not" this evening, but I guess that won't be necessary any longer. I (and probably others) will anxiously await the update. Great work!



Re: Active Sense signal "killing" MIDI Mobilizers?
by waytoocrowded on 2011-01-14 09:04:07

Thanks Ruben_cc for sorting out the specifics on this. I have the same problem but wasn't able to trace it to a specific signal so it was blamed on the controller. Looking forward to this fix so I can start using the device!

http://line6.com/community/thread/50796




The information above may not be current, and you should direct questions to the current forum or review the manual.