BrainModular

any chance for Novation Launchpad support ?
is a cheap and good controller.

http://www.novationmusic.com/products/m ... /launchpad

it could also attract more ppl to usine.
and more sells i guess ...

yes good idea but as I know

• Old-school MIDI control
Utilising Automap, Launchpad can operate as a standard MIDI controller, creating custom MIDI maps with 72 of its buttons transmitting either CC or MIDI note data.

so it's probably easy to create a command patch?

i'll get it next week and try to make a patch for it

cool!

dont miss the programmers reference:
http://www.novationmusic.com/support/st ... ce.pdf.zip

imho, automap isnt necessary? automap cant change the the internal midi configuration. if one needs to remap things, bsorks midi mapper would do it too, i guess.

may i ask what kind of patch u want to do? i am very interested in how to make use of 12 colors and especially bi-directional communication.

what should a 'command patch' do exactly?

will also try, as soon as i get the launchpad: nomone: mask launchpad as monome (and use all monome apps)
http://post.monome.org/comments.php?Dis ... e=1#Item_1

Just for the record: I took a quick look at the Launchpad programmers reference, and the MIDI mapper doesn't help with this at all.

yup give reports!

dam, still waiting for my launchpad..

cool new application, stepseq for Live:
http://www.novationmusic.com/news/

inspirational!

received my launchpad today
got a nasty problem that lcould be a problem with my setup or a problem with Usine midi...

launchpad needs to receive midi data to make visual response. to play launchpad as a keyboard, launchpad IN goes to launchpad OUT.
the problem: if i hit loads of different pads, sometimes strangely other pads are lit, or the played pads arent lit off.
this does not happen with midiOX and Phrazor. therefore somehow midi data gets corrupted in Usine?

anybody can confirm the problem?

remark:
the monome emulator (nonome) doesnt run very well here yet..

http://little-scale.blogspot.com/2009/1 ... encer.html
easy to try: little step sequencer, max runtime needed.

So i'd like to know if this novation launchpad works all right with usine, as i might buy one, please if someone can tell in a not to technical way...?

Zz Brr wrote:So i'd like to know if this novation launchpad works all right with usine, as i might buy one, please if someone can tell in a not to technical way...?

launchpad is not a traditional midi controller.
if u press a button on the launchpad nothing seems to happen. its lights can only be controlled by incoming messages.
also u cannot edit the launchpads internal midi configuration.

i still have the problem mentioned above. there is a workaround if i use MidiOX and virtual midi drivers, or Usine as VST (tested with minihost).
works: launchpad_IN -> Usine -> midiYoke_OUT -> midiYoke_IN -> MidiOX -> launchpad_OUT
what does not work properly: launchpad_IN -> Usine -> launchpad_OUT
anybody can confirm this problem?

i believe there is a severe bug with Usine if there is a kind of 'virtual midi feedback loop' (i remember other threads..).


-> u can use the launchpad with Usine.
Usine has fantastic programming facilities alike to C++ or MaxMSP launchpad hacking, just easier imho.

very strange?
What kind of message are sent by the launchpad?
and what kind of message it receives, if for example we want to light a cell?

yep , i'm curious too...

Maybe not related, but I am seeing that the Remote (send out) midi data from Usine appears to be very thinned out to only provide the necessary final positions to outboard controllers (it looks like the control on the remote device is stuttering when it is receiving a CC from a mapped Usine object) . Perhaps it is this thinning of data that is the cause of the Launchpad not updating it's LED's properly?

I also wouldn't be suprised if there were some internal midi problem in Usine with all the strange midi things that have happened to me but I can never properly recreate or I don't know enough about how it should work to prove that anything is actually wrong

Edit:
Ok, here is some possible proof that there is an internal midi feedback loop in Usine.

Take either of these VST midi monitors, MidiMonitor at http://www.rs-met.com/software/freebies ... r_v0.5.zip
or the Pizmidi Midi monitor at http://thepiz.org/pizmidi/pizmidi.zip and with no midi inputs or outputs enabled in Usine, insert either plugin all by itself with no midi connections into an empty patch, press play in Usine and then press stop.
You should now have Ch1. CC #123 (All Notes Off), Value 0 feeding back endlessly and possibly crashing Usine.
Does anyone else get the same results?

senso wrote:very strange?
What kind of message are sent by the launchpad?
and what kind of message it receives, if for example we want to light a cell?

sent and received: straightforward note On, Offs and CC's.
there is no midi thru, so there is no 'physical midi loop'.

gurulogic wrote:Take either of these VST midi monitors, MidiMonitor at http://www.rs-met.com/software/freebies ... r_v0.5.zip
or the Pizmidi Midi monitor at http://thepiz.org/pizmidi/pizmidi.zip and with no midi inputs or outputs enabled in Usine, insert either plugin all by itself with no midi connections into an empty patch, press play in Usine and then press stop.
You should now have Ch1. CC #123 (All Notes Off), Value 0 feeding back endlessly and possibly crashing Usine.
Does anyone else get the same results?

tried and got the same result. the monitor vst is flooded with all-notes-off CC messages.
the same happens if i press Panic instead of stop.
thanks for working out this test., gurulogic.

Take either of these VST midi monitors, MidiMonitor at http://www.rs-met.com/software/freebies … r_v0.5.zip
or the Pizmidi Midi monitor at http://thepiz.org/pizmidi/pizmidi.zip and with no midi inputs or outputs enabled in Usine, insert either plugin all by itself with no midi connections into an empty patch, press play in Usine and then press stop.
You should now have Ch1. CC #123 (All Notes Off), Value 0 feeding back endlessly and possibly crashing Usine.
Does anyone else get the same results?

This problem comes from the fact that if no midi in is connected to the midimonitor VST, and only in this case, the ALL_NOTE_OFF is always sent to the vst. It appears only in this precise case and will be solved in the V5.

Then, MidiMonitor crashes but it's not really a bug of usine. Simply, Midimonitor doesn't like high speed series of ALL_NOTE_OFF.
From my point of view It's not a sign of an internal loop.

Another important Info:
the remote update process is made every seconds (to be lighter in CPU as possible). So the remote way is not adapted to update lights of the LaunchPad. For that try to use the MidiOut Module.

I guess that helps explain things for me. So is it safe to assume that any plugin that has a midi input must have its midi input connected to prevent strange behavior..at least untill v5?

One other thing that has made me think internal midi feedback is that if I load up two instances of Molar VST and send midi from one instance to the other via direct midi connection, the instance with no midi input is also affected. This doesn't happen in Live 7 but I am willing to say this is just because Usine and Molar are strange together.

Also when I seem to have had stuck midi buffer loops but I can never recreate this...(maybe the same as the no midi connected problem?)

Anyways, lets just see how everything works in v5

senso wrote:This problem comes from the fact that if no midi in is connected to the midimonitor VST, and only in this case, the ALL_NOTE_OFF is always sent to the vst. It appears only in this precise case and will be solved in the V5.

Then, MidiMonitor crashes but it's not really a bug of usine. Simply, Midimonitor doesn't like high speed series of ALL_NOTE_OFF.
From my point of view It's not a sign of an internal loop.

thanks for the clarifications, and thanks for the upcoming fix, senso

senso wrote:Another important Info:
the remote update process is made every seconds (to be lighter in CPU as possible). So the remote way is not adapted to update lights of the LaunchPad. For that try to use the MidiOut Module.

nice to learn about that lag.
i do use the midi out module.

the situation:
i quickly hit a single launchpad button -> no problem.
i quickly hit a few launchpad button -> other buttons get lit.
very strange: there seems to be no logic, the same behavior does not seem to produce the same output (in terms of 'ghost' notes and velocities)
the trace window does not output additional midi data. but if i route the launchpad input to my hardware midi synth, i can hear the 'ghost notes'.

however, right now, i use 'midiOut' by insert_piz_here: http://www.thepiz.org/dl/?p=vst instead of Usine midi out. works.

@gurulogic:you should not focus on the ALL_NOTE_OFF bug because it's totally insignificant and without any consequence on 99.999% of vst...

however, right now, i use 'midiOut' by insert_piz_here: http://www.thepiz.org/dl/?p=vst instead of Usine midi out. works.

Interesting. They are two investigation directions:
1) use an unpack sctript before the midiout module.
2) use a small Midi-delay. to send midi codes one by one separated by few ms.

senso wrote:

however, right now, i use 'midiOut' by insert_piz_here: http://www.thepiz.org/dl/?p=vst instead of Usine midi out. works.

Interesting. They are two investigation directions:
1) use an unpack sctript before the midiout module.
2) use a small Midi-delay. to send midi codes one by one separated by few ms.

investigation report:
1) in the 'midi In' module i switched on 'unpack'. yai, there we go!
2) 'midi delay' script does not help.


senso, u made my day - once again

edit: further investigation:
the 'unpack' script is needed if i connect a piano roll to midiOut and create some midi notes at the same point in time.
without unpack script: usine and launchpad freeze. rebooting launchpad and usine is necessary.

loud thinking: there might be a general problem with generating/sending multiple midi messages in one block calculation?
i didnt encounter that problem with virtual midi ports (Usine -> midiYoke -> midiOX -> launchpad). maybe (some) midi-hardware drivers simply cannot handle packed midi data? and if so: the unpack routine hardly affects channel messages (note, cc, etc.) but it could be bad to unpack system messages (clock, MTC)?


btw, any other launchpad (ab-)users here? would love to discuss launchpad-usine apps.

short note:
it might be the case that my suspicion about a faulty Usine midi implementation is wrong.
other people (standard Live users) seemed to have problems with hanging lights too.
maybe the problem is launchpad-specific. if so, novation will most surely release an update - or people might have to use Usine to work with their launchpads
discussed here: http://www.novation-launchpad.com

good news, for me I mean!

when i get my head around usine a little better i will have to connect my launchpad and nocturn and see how it works with usine:)

you could check the patches from Etnhix73
check videos here :
http://www.youtube.com/user/ethnix73#p/u/3/osaIApQ3Gsg