Real-time MIDI Input Issues

• Feb 18, 2018 - 19:10

I trying to use the manual version of real-time, but I am having issues with the insertion of the correct rhythm. The piece is in 4/4 time and I am trying to write in a bass line. I am using a Casio WK-7500 as the MIDI Controller. I am using a assigned my pedal for the real-time advance feature. I start at the beginning of a bar, with real-time (manual) activated and set to a quarter note. I press and hold the key I want, I press my pedal and the rhythm I am left with is a whole note. I noticed if I set it to a half not it gives me 2 whole notes, and if set to a whole note it gives me 4 wholes notes. Any help with this issue? Has anyone else had this?


Comments

My guess is your pedal is sending multiple MIDI messages, maybe it is sending on multiple channels? Does your keyboard have a split or layer function, perhaps? Disable it and try again.

Yep I get exactly the same problem. I've just connected my Yamaha DGX620 "Portable Grand" to my computer and MuseScore for the first time, and that was the first problem I encountered! I thought I could cure it by setting the minimum note length to half what I needed (eg quaver instead of crotchet) but then of course that just inserts two quavers instead of one crotchet, HahahahahAAAAAARGH! So it looks like your MIDI instrument is issuing 4 events per press, and mine is issuing 2 events per press.
I will have to dig DEEEEEP into my instrument manual to see if I can cut out the extra events, but actually it would be extremely helpful if MuseScore could anticipate this problem (which seems to be quite common, now I've researched it a bit) and simply include an option to convert multiple simultaneous pedal-press events into a single one. So it could just be a box into which you type the number of events your pedal issues when you press it, and then MuseScore could use that number to ignore the excess events for that particular keyboard's pedal :o)
We will have to rely on the coding gurus to sort that one out for us!
Best wishes,
Tony.

In reply to by Gigabeast

Until the app is recoded or I find a way to kill the extra events my pedal is sending, I will probably have to resort to using the fully automatic Real-Time MIDI entry mode and try to keep up with the metronome, heheheheh! I know I can slow the ticks down, but whatever speed you've set it to, you have to keep up with it...
OOOOOOOOOOOH I found another workaround which does actually work on my keyboard - you don't have to use the pedal to create the Real-Time Advance event, you can use any key on the instrument. I tried setting it to the bottom-most key in the preferences section where you record the even that's going to be used to cause the Advance, and now instead of pressing the pedal to create the Advance, I press the bottom note on the instrument and it advances the beat by one unit instead of the pedal's two. Yay! That's me sorted until the coders can strip out the pedal extras then :o)
Tony.

Hi, I have the same issue but my pedal is not sending the event two or four times but hundred times. This causes Musescore adding about 20 empty measures on a single pedal press. This is because when I press down it goes from 0 to 1, 2 ... 127 and then from 127 to 126, 125 ... 0 when I release the pedal. My digital piano is capable of full pedal dynamics so it's not a simple on/off switch.

In your case I suspect that your pedal is capable of two sates (full and half) and it is sending an event for each state. You can verify this by testing your MIDI device with a MIDI monitor tool (I used midimonitor.com with Microsoft Edge to figure out what was happening).

This is also why my MIDI Bluetooth adapter causes crashing all apps (Musescore, Synthesia, ....) when I press the pedal, because the amount of data is insane for the wireless transfer rate. So I'm restricted to wire connection only. Unless I replace my full-dynamic pedal with a basic one I can't use it for the manual real-time mode.

Please don't blame me for waking up a five year-old thread. I'm convinced that every single idea is worth to be shared. Google took me in this thread when I was looking for of an explanation. This thread helped me a lot in finding the real cause.

In reply to by tazzit24

Hi tazzit24!
The thread is old but the problem is still very current! Your contribution is very valuable, especially as you have taken extra steps to try to diagnose what's going on. I'd never considered the pedal to be the issue, I assumed it was my keyboard sending too many signals, and I'd made the foolish assumption that my pedal was just an on-off switch, hehehehehe! I've had a look at your suggestion of Midimonitor.com and went there in Chrome only to be told "your browser is not compatible, please use Chrome" - D'oh, I'll fire up the keyboard later and have a go - if Midimoitor.com won't play with my system then there are others to try. I'm lucky I have 3 pedals I can try, at least one of which I'm pretty sure is just a switch as I use it for turning pages on my antique MusicPad portable score display screen so it would be very handy if I could just plug that in when I want to record music in MuseScore :o) Best regards, Tony.

In reply to by Gigabeast

Aha, I used https://volcanomobile.net/midi-web-tools/midi-monitor-online/index.html and paused it after a press-release of the pedal (because there's a stream of data that comes out of the keyboard when it's not doing anything), and it turns out on my system the following MIDI information is received by my computer:
Datalength 03
Code 176 - controlChange
ControllerNumber 64
ControllerValue 0 (off) and 127 (on)
Channels - both 000 & 001 for each change,
Yep, that's what's happening then, With each press of the pedal the keyboard is sending MIDI codes for a controlChange (code 176) for controllerNumber 64 (pedal) with a value of 127 on channels 000 AND 001, and then the pedal release is sending the same info but with a value of 0 on channels 000 AND 001.
By comparison, when I press a piano key, the data is for noteOn (code 144) and only sent on channel 000 for pressing (data includes key number and velocity) and releasing (data is same key number and velocity=0)
So, I wonder if MuseScore can be set to ignore Channel 001? That would fix my current setup - I have a suspicion that changing the pedal will produce the same result unless it's got some electrickery in it to produce MIDI signals on two channels and I rather doubt that! It seems the code for releasing a key is still noteOn but with a velocity of 0 which seems fair game although in that case noteChange might have been a better name for the code hahahaha!
If your pedal/keyboard is chucking out 127 amplitude readings every time you press it then I guess the only solution is to use a switch pedal instead, but if it's chucking out one reading on 127 channels then it might be possible to ignore all the channels except 000, I don't know, I haven't dug into MuseScore yet to see if I can turn off channel 001 :o)
Tony.

In reply to by Gigabeast

HI Tony,

Very interesting. I made the test and unlike the previous tester I also get duplicate events on two different channels. My Roland FP-4 is indeed configured to use Ch1 (000) for upper tones and Ch3 (002) for lower tones. If I change this settings on my device then the online monitor indicates the new chosen channels as well. So I think the webapp is reliable. So when you press the pedal, your Yamaha device is sending the control on all both active channels. I checked your device's user manual and I couldn't find any option to control the output channel (it seems you are limited to Ch1 and Ch2). In my own device I can select the same channel for both Tones (I tried Ch5) but the device keeps sending duplicates "004". It seems that our digital pianos are not the best suited for MIDI control purposes.

Here's the (truncated) log when I press the pedal with the default MIDI config (Ch1 -000- on Upper Tones and Ch3 -002- on Lower Tones) ;
code: 176, type: controlChange, channel: 002, controllerNumber: 064, controllerValue: 127
code: 176, type: controlChange, channel: 000, controllerNumber: 064, controllerValue: 127
...
and so on for the other 127 pedal states.

So even if I plug an on/off switch pedal I will face the same issue as yours.
It makes me sad but at least we know why this is happening. I don't think that a Musescore setting could save us from this unwanted side effect. Thank you very much for all the tests you made.

Have a nice day,
Giancarlo.

In reply to by tazzit24

Hi Giancarlo,
I doubt it's beyond the programmers at MuseScore to create an optional filter so that only one pedal channel is used and the other data discarded. And being open source of course others could come to the rescue with a plugin I expect. Wouldn't help you with your proportional pedal, but a cheap switch pedal would then work as mine would without the single duplicate :o)
I wonder if my keyboard (Yamaha DGX-620 is the on I use with my PC) instruction manual can help me stop the second channel sending. I'll post an update here if I succeed!
Best regards,
Tony.

In reply to by Gigabeast

Nope I failed with the instruction manual. Tried the various PC connection settings but I either get no output at all other than clock, or I get two channels of pedal output.
From another very old thread, I wonder if https://das.nasophon.de/mididings/ would be a solution, to filter out the extra channels, though that looks like a steep learning curve and an internal MuseScore fix would be infinitely preferable :o)

Do you still have an unanswered question? Please log in first to post your question.