Skip to main content

I am using Pianoteq on a PC, hooked up via a TC Electronic Konnekt 24D to a Roland A-80 MID controller keyboard.

I have two damper pedals that support half-pedaling: a Yamaha FC3 and a Roland DP-10. Neither of them is working properly with my system.

The FC3's polarity is set so that MIDI 0 is sent when the pedal is fully depressed and MIDI 127 is sent when the pedal is not depressed. This is counter-intuitive and opposite from what Pianoteq expects from CC64. As a result full sustain is applied when the pedal is not depressed at all!

The DP-10 has the polarity correct, BUT it sends MIDI values 24 to 127. So I am getting a little bit of sustain even when the pedal is not depressed. Not good.

I have tried things like power-cycling the A-80, but to no avail.

Any help would be much appreciated.


anonymous Fri, 07/25/2008 - 19:57

Space wrote: I think you want to set the Roland A-80 to be configured for the pedals not the vst.

The DP-10 is plugged into foot controller port 3, and this port has been configured to send CC64. There is no other configuration possible on the A-80 for the foot controller ports.

The MIDI values that I am seeing are recorded on the PC, after the MIDI data has been received from the A-80 via the Konnekt 24D.

anonymous Fri, 07/25/2008 - 21:12

The only configuration that can be done is to select a controller number between 0 and 121, and a bunch of other controls like POLY on/off.

The owner's manual indicates that no other configuration is necessary for pedals like the DP-10. It says: "...continuous volume pedals allow for example 'Continuous Damper' generation...."

My best guess as to the cause of the problem is a defective DP-10. I hope that Roland will get back to me about this.

In the meantime, I have worked around the problem by using a suggestion from Pianoteq: Use MIDI-OX to re-map the 24-127 value range to 0-127, and MIDI Yoke to connect the output of MIDI-OX to the MIDI input of Cantabile (the VST host that I am using for Pianoteq). Unfortunately this is causing problems with Cubase, so is not a permanent solution.

Space Fri, 07/25/2008 - 21:24

Cubase doesn't host the plugin? I thought that was what Cubase vst plugins?

edit: Can you turn polyphony off specifically for this pedal..would that help?

I thought if I got out here and asked dumb questions maybe someone that has this 20 year old board would show up and blow me off with all their smarts...maybe in the morning.

anonymous Fri, 07/25/2008 - 21:42

Space wrote:
I thought if I got out here and asked dumb questions maybe someone that has this 20 year old board would show up and blow me off with all their smarts...maybe in the morning.

Heheh...I figured that's what you were doing. I've had the A-80 for 18 years, and recently pulled it out of mothballs to use as the keyboard for Pianoteq. The weighted action on the A-80 is fabulous, probably still the best around. And she's built like a tank. Weighs all of 30kg (66 lbs) and there is ZERO flex.

anonymous Fri, 07/25/2008 - 21:49

Space wrote: buffer settings

you can probably search for them @RO and get something that will work for your setup. Maybe set up a hardware profile to eliminate all unneeded software at bootup.

I have already done a first pass at optimising the system. Turned off all non-essential services (at least the obvious ones), disabled all non-essential drivers (including the ones for the motherboard's TWO audio chipsets!), tweaked process scheduling to give priority to background services, turned off the fancy UI presentation, and so on. I've probably got more tweaking to go.

Edit: And I have tweaked the Konnekt 24D's DPC buffer settings. It is currently set to Level 1, but I think I will have to go to Level 2. But I don't see why such things are necessary. I've got two fast CPUs, a fast FSB, gobs of RAM. And this PC is a dedicated audio workstation so there is nothing on it that is related to anything other than audio.

anonymous Sun, 07/27/2008 - 19:32

I have solved the mystery problem with the DP-10.

I took apart the DP-10 and looked at the MIDI values generated by the potentiometer when I manually rotated it through its full range. The values ranged from 4 to 127, which was obviously better than what I was getting previously, i.e., 24 to 127.

This suggested that when the pedal is not depressed something is preventing it (and hence the potentiometer) from returning to the rest position at which its minimum MIDI value is output. It didn't take me long to notice that the felt pad against which the pedal rests was too thick. This felt pad is to prevent the pedal from contacting the pedal case when fully the pedal returns to the rest position. I will report the out-of-spec felt pad to Roland.

The solution then was simple -- I used the extended blade of a box cutter to slowly shave layers off the felt pad until it was thinner by about 50%. Goodbye MIDI-OX and MIDI Yoke! I still don't get MIDI 0, but I guess the sustain at that level is practically unnoticeable.

anonymous Sun, 07/27/2008 - 22:22

All my T-shirts are inside out.
It's amazing how people who make clothes for fat people assume they are tasteless morons.

I get how shaving the pad back gives you more pot slide. But why wasn't the pedal calibrated for that sized pad?

Does Roland assume it's manufacturing is perfect every time?

anonymous Mon, 07/28/2008 - 00:26

These pedals are made in China by a contract manufacturer at very low cost. My guess is that they don't do any batch testing to make sure that the parts and finished pedals are within spec. The felt is too thick, AND the pots transmit MIDI 4 even at zero sweep! The design is actually quite clever - the shaft of the pot is driven by a cam in such a way that no calibration is needed. The felt just has to be of the correct thickness.

I have given feedback to Roland Asia-Pacific (I live in Singapore). I have no idea whether it will make any difference to the QA levels in the factories. :roll: