Hey 5pinDIN,

Thanks again for the continued support with this matter. I’m not sure how to filter CC#11. When I choose the option to filter it out it stops all data being transmitted from the pedal. As for the second output, I believe I had the “MIDI Mapper” output device enabled within MIDI-OX (I assumed I needed it enabled).

I might be missing something, but I just can’t seem to find the option to specifically filter a CC message on a desired channel.

Many thanks.

Sorry, I only filter System Realtime messages myself. Apparently you can’t specify which channels to apply CC filtering to. Let’s forget about filtering CC#11. Instead…
Options > MIDI Filter
At the right of the Filter window under “Channels” , click “All”, then uncheck “1”. You had previously mentioned using only Channel 1. By having only that channel unchecked, all data from the remaining channels should be filtered out. Hopefully that will allow only Channel 1 data to pass, with CC#11 messages on that channel already being remapped to CC#4.

Hey 5pinDIN,

I’ve done as you have directed, but I’m not entirely sure everything is correct. I had the idea of screenshotting the various settings and devices - maybe it will be helpful?

Many thanks.

Click thumbnail to see full-size image

The screen captures look about right to me. Are you now able to play the MM8, use the FC7, assign the Foot Controller (CC#4) as desired on the XS, and have it work correctly? If not, what’s happening?

Hey 5pinDIN,

it seems that the pedal is now fulfilling the desired function, but it continues to cause large amounts of latency whilst playing notes and depressing the pedal rapidly. Also, I am still unsure on how to record into Reason with this current setup. It appears when I set up an external MIDI device in Reason the pedal defaults back to only controlling the volume - bypassing any changes made with MIDI-OX.

Many thanks.

Azluna - 27 August 2018 04:21 PM

Hey 5pinDIN,

it seems that the pedal is now fulfilling the desired function, but it continues to cause large amounts of latency whilst playing notes and depressing the pedal rapidly.

That may be a MIDI buffer issue. I noticed that you have a checkmark for System Realtime messages in the Filter. Yamaha keyboards tend to send F8 (MIDI Clock) and FE (Active Sensing) even when not required. Those messages can cause buffers to overrun, so filtering them from the data stream can sometimes be helpful. If you’re not doing any sequencing from the MM8, you might try filtering all Realtime messages, and not just from the display.

You might also experiment with the number of buffers and their size. The default settings aren’t necessarily the best, and larger buffers don’t always improve things. Sometimes more smaller buffers works better.

If the latency issue isn’t resolved by what I mentioned, you might want to ask at the MIDI-OX forum…
http://www.midiox.com/cgi-bin/yabb/YaBB.pl?board=query

 

Azluna -

Also, I am still unsure on how to record into Reason with this current setup. It appears when I set up an external MIDI device in Reason the pedal defaults back to only controlling the volume - bypassing any changes made with MIDI-OX.

Many thanks.

I don’t use Reason, so I can’t be helpful with its operation, but I suspect that proper routing of data via MIDI-OX (not Reason) may be the key. You may also be able to obtain guidance at the MIDI-OX forum.

By the way, if an event processor (connected to the output of the MM8) was used to convert Channel 1 CC#11 messages to CC#4, and to filter the remaining channels, the Reason setup you currently use should not require any change. It also shouldn’t introduce any significant latency. You might contact , explain what you need to do, and verify that.

Hey 5pinDIN,

I’ve tried altering the buffer settings as well as filtering out realtime messages from the data stream but to no avail. I’ve taken a look around the web, but I’m not having much luck finding any help with regards to latency and MIDI-OX. The latency only ever occurs when using the pedal rapidly in conjunction with the MM8. 

I’ve attached an image below that might be helpful. It seems to show the Note On messages being interrupted by the pedal’s operation - causing the Note Off messages to happen much later than they should. I’m not sure if that’s correct at all, but the faster I operate the pedal the larger the distance seems to be between the Note On and Note Off messages.

Also, I’ve noticed that despite having all but channel 1 disabled, as well as filtering CC#11 to CC#4, the MIDI Port Activity monitor is still showing incoming data on channels 1, 2, 3, 9 - 16 when operating the pedal. Perhaps this is the issue?

Thank you again for the continued support. I think I will consider purchasing the MIDI Event Processor you recommended if this latency issue cannot be resolved.

Many thanks.

Click thumbnail to see full-size image

Azluna - 28 August 2018 02:20 PM

Hey 5pinDIN,

I’ve tried altering the buffer settings as well as filtering out realtime messages from the data stream but to no avail. I’ve taken a look around the web, but I’m not having much luck finding any help with regards to latency and MIDI-OX. The latency only ever occurs when using the pedal rapidly in conjunction with the MM8.

OK.

 

Azluna -

I’ve attached an image below that might be helpful. It seems to show the Note On messages being interrupted by the pedal’s operation - causing the Note Off messages to happen much later than they should. I’m not sure if that’s correct at all, but the faster I operate the pedal the larger the distance seems to be between the Note On and Note Off messages.

What you’re seeing is normal MIDI message priority, but…

 

Azluna -

Also, I’ve noticed that despite having all but channel 1 disabled, as well as filtering CC#11 to CC#4, the MIDI Port Activity monitor is still showing incoming data on channels 1, 2, 3, 9 - 16 when operating the pedal. Perhaps this is the issue?

...you made me do the math - and it looks like you’re correct.

MIDI data speed is 3,125 bytes/second, which means 0.32 milliseconds per byte. Each CC message is three bytes long - Status, Data 1, Data 2. That results in just about 1 ms per message.

You don’t need the pedal data on anything other than Channel 1. However, each message sent by the MM8 will take 1 ms, and it’s also sending on 10 additional channels, thereby adding 10 ms for each pedal value sent. The TIMESTAMP in the screen capture is time in milliseconds (in hex). The difference between the pedal events is 11 ms, corresponding to 1 ms for each channel (1~3 and 9~16).

So while the unwanted message on channels other than number 1 can be filtered out, the timing is still delayed by them.

 

Azluna -

Thank you again for the continued support. I think I will consider purchasing the MIDI Event Processor you recommended if this latency issue cannot be resolved.

Many thanks.

Unfortunately, an event processor will have the same issue. The problem is indeed due to the MM8 sending pedal data on so many channels.

I have to admit that this is a situation I’ve never dealt with before. I was aware that the MM models had MIDI issues, but I hadn’t realized the degree. The sheer amount of data from a pedal, even when operated rapidly, is not a common occurrence.

It looks like unless there are undocumented commands for the MM8, it isn’t a good controller for your purposes. You might try contacting Yamaha Support and see if they know anything about this problem.

Sorry that it took so long to fully understand what’s happening, and I’m glad you were persistent.

Not to open another “can of worms”, but I wonder if the same excessive pedal data is sent if the USB TO Host connection is used instead of the 5-pin connection.

Hey 5pinDIN,

I appreciate the rapid feedback and support you have offered throughout my issues. I have to say it is a little sad to hear that it is likely an issue with the MM8 itself, but I also feared that it might have been the culprit all along. I will continue to look around for a potential fix and keep you posted if I find one. I’ll be sure to try connecting the MM8 in a different manner using USB as you suggested.

I’d like to also personally thank you for the insight into MIDI as well - I have definitely learned a lot more about it thanks to you and your efforts.

EDIT: Success! I think I have finally resolved the issue!

I was about to throw in the towel, but I decided I would try one last time to fix the issue before taking a break from it. I sifted through each setting on the MM8 and the Motif-Rack XS for some time in the hopes that I had missed something - it turns out I may have found the offending setting.

Now, I am not entirely sure why this is the case (because the MM8 manual does not seem really explain this at all), but it turns out that having the Pattern Out setting enabled was causing the MM8 to transmit data on channels 9 - 16 when operating the pedal. When I double checked MIDI-OX, I realised that the data was only being transmitted on channels 1 - 3 with the setting disabled.

All I could find in the manual regarding it was as follows:

“This setting determines whether keyboard performance data is transmitted via the MIDI output (ON) or not (OFF).” (P. 78)

“Determines whether pattern data will be output via the MIDI output (ON) or not (OFF) during pattern playback.” (P. 65)

Seeing as the pedal (at least to my understanding) had nothing to do with “pattern playback” I assumed the setting had no bearing on the pedals function. It turns out it did. I can’t help but feel a little silly that I did not think to go through every setting sooner.

Either way, I cannot thank you enough for all of the support you have provided me 5pinDIN. I would not have gotten as far as I have if it were not for your guidance. I appreciate it.

I had also been thinking about why the pedal data appeared on Channels 9~16, and I found the Initial Setup section on page 79 of the Owner’s Manual. There’s a “NOTE” at the end of that section which mentions 16 MIDI channels.

I was about to research that further to see if there was a setting on the MM8 to prevent it from sending multitimbral data, hoping that would stop it from sending pedal data on other than the first three channels. However, before doing that I logged on and saw your EDIT.

Congratulations on finding the “Pattern Out” ON/OFF setting! With pedal data only being sent on three channels, the time per value sent will be only 3 ms as opposed to 11 ms. That should certainly speed things up. By the way, does “Song Out” ON/OFF also affect pedal data?

Although I didn’t have all the answers (the MM models are a bit of an enigma), I’m glad that between the two of us things have turned to the better in the end. As to any assistance I may have provided, you’re very welcome. I hope you’ll now be able to more fully enjoy the Rack XS.