sticking/hanging notes

I'm making this a new thread as it concerns the full release of R4

Does anyone else get a lot of sticking/hanging notes while playing Reaktor?

I get them way too much - they interrupt all my playing on any ensemble. I tried to find the cause, and it seems to be a Reaktor issue.

I'm using a Remote 25 MIDI controller via USB on OS X, but this even happened when using the computer keyboard. Also, I tested my controller on Reason 2.5 and there was no problem whatsoever.

There's one post here that mentions it (and even john nowak seemed to experience it):
http://www.native-instruments.com/forum/showthread.php?s=&threadid=5140&highlight=stuck+notes
Also, someone posted similar experiences with the Reaktor Demo (http://www.native-instruments.com/forum/showthread.php?threadid=6621), but I'd like ot know if others experience this with the full version of R4, and if there are any solutions.

If this is a problem it should be addressed right away. It's really bad if the MIDI implementation is this rough - it makes the program quite useless for live work for me.

 

Did you try turning on reassign as per earnest's suggestion
in the thread from June you referenced? I had some problems
with stuck notes in the very first versions of Reaktor 4,
usually in the context of Reaktor crashing within Sonar
due to some of the early UI bugs, but since 4.02 or so,
I've had neither crash nor stuck note.

 

Yeah, I've tried the reassign function, but it makes no difference.

Actually, along with the stuck notes, there is also the problem that sometimes the notes don't even trigger. This happens with chords and solos - I tried increasing the voices/polyphony, but the same problem occurs. So I end up with some notes hanging there, and others not triggering.

 

If you play MIDI sequences through Reaktor, do you
also get stuck notes? If so, that would pretty
conclusively put the blame on Reaktor. Otherwise,
it could still be your MIDI keyboard/driver that's
screwing up somehow. What level of CPU load are
you seeing reported by Reaktor when this happens?

I take back what I said about no longer getting
Reaktor crashes, though. I just had a spate of
them under Sonar. It was just after I installed
the latest version of the DR-008 drum synth, though,
and I suspect that may have something to do with it...

 

It may depend on what you are doing with the notes. Perhaps if you post the ensemble with the problem some kind sould may look through it and work out what's going on.

 

I'm noticing stuck notes in the Reaktor demo that happen whether or not I'm in standalone or using Reaktor as a VST instrument.

What I'm noticing is that when I lift a MIDI key, the note is parsed as a NOTE_ON for the last note that received a NOTE_ON. Obviously, this makes Reaktor totally unusable, and happens no matter how many keys I'm playing.

 

I pretty much use Reaktor exclusively now, running through Cubase. I'd never had stuck or missed notes until I used Reaktor, although it only showed up when using three or more instruments. Also, it only seemed to happen when using SoundManager, rather than my (currently broken) Audiophile ASIO drivers. The problem is also intermittent, so i don't know what's going on.

mark
www.darklogik.org

 

Actually this happens with all ensembles!

 

Midi sequences seem to play OK, though as these were created ouside Reaktor, maybe it's not conclusive enough to proove it's not Reaktor.

The CPU load is normal for all occurences of the sticking notes - say 30 %.

I'm going to try to contact Novation and see if what they think. Maybe I'll contact NI support directly also. It's beginning to really bug me.

 

I seem to recall having seen on at least one
occasion a MIDI controller or sequencer which
transmitted NOTE-ON with zero velocity instead
of NOTE-OFF messages to terminate each note.
According to the MIDI spec, the effects should
be the same. But maybe Reaktor has a quirk in
the MIDI gate module. Which does your keyboard
send, and is it possible to configure it otherwise?

 

FWIW, I'v had no problem with hanging notes in R3 or R4, stand alone or as VSTi plugin, except when I'v messed with the P & G.
I'm using Win XP & Win ME.

Have you reported this to the NI support?

 

Yes I have now - no response yet though.

FWIW, Novation replied and said tht they've not tested the ReMote 25 with Reaktor - which seemed a bit odd to me.

 

I've now tested the Remote 25 with lots of other software - standalone & plug in - and there are no problems with it on anyting other than Reaktor.

So, that puts the fault with Reaktor, and frankly, that is totally unacceptable.

How can NI sell synth software that doesn't respond properly to MIDI controllers? Reaktor is useless to me like this! All those great ensembles, all that functionality, but no way of playing it? WTF?

Is anyone at NI listening? Hello? Could somebody wake the tech support people up please?

 

i frequently experience stuck note problems in some of my patches. i know it's not a system overload problem, as i've got a new dual 2.0ghz G5 with 2.5GB of RAM. cpu usage in the patches that get stuck notes is nominal anyways..

i'm strictly talking about using the computer keyboard to trigger MIDI notes. my MIDI device is a USB connected Edirol UM-880 -- 8 channels. i've never experience any problems with the UM-880 when using a normal MIDI keyboard -- stuck notes only occur within reaktor.

it usually happens when i'm hitting the keys fast, like when i'm playing with shorter, percussive sounds/hits that i like to play close together. the faster i trigger notes, the more likely reaktor is at sticking my MIDI note

i haven't tried this in VST mode, only in stand-alone.

perhaps this is an OS X thing?

Dual 2.0GHz G5
2.5GB RAM
OS X 10.2.7 (G5) -- updating to 10.2.8 soonish

i will contact ni support regarding this issue as well...

slag

 

I assure you the programmers are not napping, and are working on this issue as we speak. Continued emails from various people with system configurations could be helpful. I'll try to secure how the progress is coming and report back when I can get some info.

Unfortunately, I can't help anyone on this issue really because I can't test it. I don't own a keyboard. Playing notes manually isn't my thing. ;-)

I'll test with the computer keyboard though and see what I can get with that.

 

I also experience stuck notes when using the computer keyboard (Mac G4, OS X). And now i made some measurements with MidiMonitor. The messages sended are NoteOn (vel 80) and NoteOff (vel 0). That's ok, but If i hit a key very shortly in the same way, the time difference between NoteOn and NoteOff varies from 3 msec to 50 msec! I think that indicates serious timing problems within Reaktor (or possibly OS X). Thus it may happen that NoteOn is sended *after* the corresponding NoteOff which means 'stuck note'...

 

That's almost surely a timing problem with your hand rather than your computer. For you to assume you're hitting the key and releasing at exactly the same speed each time is boardline insanity. The human body is not capable of precision within 30 or 40 ms.

And for the record, I'm near-positive there are no such timing issues with OS X.

 

I would like to add, that also the computer keyboard itself may contribute to bad timings, ie. delayed NoteOn vs. NoteOff. Finally these keyboards are not designed for playing music ;-)

 

Of course.

 

Well, i just checked my 'boardline insanity' using a professional MIDI keyboard and hitting a key in the same way as before on the computer keyboard. Believe it or not: The NoteOn-NoteOff time difference varies from 40 to 60 msec.