Just for reference, curves out of the manual:

manenbu


I did some tests.

First one is trying each of the curves. I pressed middle A on the A300, and pressed down as hard as I could.
 
Obviously, 4 is totally useless. 1 is crappy as well. I almost maxed out on 3.
Also notice how the initial velocity increases - probably the result of me getting pissed.

Next I was trying to play something (some random notes) with a "real world" attempt of aftertouch. Curve is set to 3.
 
Well, it's possible. Very hard to control the level, but I guess we are ****ed anyway. So better practice this instead of ****ing about Roland's support.
I must say that I also own an EV-5 pedal, so whatever vibrato I need goes there. AT could be sort of an "on-off" switch for me.

Good goin', Manenbu - Thanks for doing the screen shots and posts.

Curve 3 is the most sensitive, as shown in the manual's curve chart, which you've also posted.  When I try that, if I stand up and push so hard I think I'll break something, then I'll get an intermittent swoop of AT data, like in your picture, but I mean, really intermittently - just on a few scattered notes, not all of them.  But for me it's literally impossible to have any kind of reliable consistency with it.  It was less hassle just drawing it in the PRV as I did forever before this.

You've mapped AT to a pedal, good solution.  What I've done is have a special map with AT mapped to the modulation stick:

--This works for me because my primary instruments are Garritan, and most of those use CC1, or alternately, CC11 for volume.  I have a pedal for doing volume via CC11 with those instruments, leaving the mod stick handy for adding good, finely nuanced AT.  It works great, and I use it sometimes for non-Garritan instruments also.

Trying to use the spring-action mod stick for volume control with Garritan instruments doesn't work exactly for the reason that it's a spring-loaded thing.  My ancient keyboard before Apro, a $50 used generic controller beat the pants off the Apro with its mod wheel because it was free wheeling, no center detent, no spring action.  And it could be instantly programmed for any CC controller needed - a much more straightforward proposition than having to make all these various maps for the Apro.  BUT - it only had the one assignable control, the wheel.

AT shouldn't be an off/off switch, because there's a big difference between suddenly having full vibrato, and having it go in and out in degrees.  The extra spice in Garritan instruments is then to also control the speed of the vibrato, via CC17 which I have mapped to an Apro knob in my special Garritan control map.  Having various levels of AT and, if available, degrees of speed - that's what makes samples of real-world instruments come to life.

But, back to the point - AT doesn't work right in the Apro with the default connection to key pressure.  It's spotty at best.  Happily though, AT doesn't have to be only on the keys.  To me there's no problem making another sweep through a MIDI track to add AT data, or any other data, via controls on the Apro or by just digging in and drawing them in the PRV view.

Sadicus, you asked what "Sonar patch" could be used as a test - We've already proven that AT in the Apro doesn't work right, it wouldn't really be a good use of time to discover the same thing all over again.  But it shouldn't matter what patch you use anyway.  AT is commonly used in a lot of synths and their patches - In theory it should work with anything you want.

Randy B.

As everyone seems to agree, the aftertouch on the Ax00-PRO controllers is pretty much unusable, in that way too much force is required. I'm surprised the Roland hasn't responded to questions about this.
 
Looking inside my A300, the aftertouch strip under the keybed is a standard 2-conductor/pressure sensitive plastic strip, as used in most other keyboards these days. The plastic strip connects to a small PWB at the bottom left, where the controller / power connector inputs are located. There's a trimpot, VR1, just behind the expression input jack, that seems to adjust the aftertouch, but it didn't make much difference in the response: with VR1 fully clockwise, the force required is maybe a little less, and fully counterclockwise there's not any response. Probably a resistor value change somewhere in the aftertouch circuitry would solve the problem - anyone have a schematic?
 
Another approach would be to remove the aftertouch strip and use a thinner felt strip on the top where the keys press (or reduce the thickness of the existing one) and then put a strip of thin plastic under the strip to raise it back to its original level.
 
I noticed that the aftertouch circuit uses half of IC8, an 358-type dual op amp (next to VR1) as a unity-gain buffer: the + input is connected to the VR1 trimpot, the - input and the output are connected, and the output is presumably connected to the analog-to-digital-converter input. So I increased the gain of the buffer to 2, by lifting pin 6 (the - input) and connecting it to both pin 7 (the output) and pin 4 (ground) through 2.2K ohm resistors. It works fine, and I now have usable aftertouch - only half the pressure is required. But of course you shouldn't have to make this kind of modification in the first place.
df

I've messed with this for a while and I'm not sure it's mechanics. If I use AT I hear no changes in sound but the CC data is clearly being recorded. The CC data just doesn't seem to interact with the patches I'm triggering. Very strange. I have a very old JX8P and AT works great. Go figure.

dfahrner

I noticed that the aftertouch circuit uses half of IC8, an 358-type dual op amp (next to VR1) as a unity-gain buffer: the + input is connected to the VR1 trimpot, the - input and the output are connected, and the output is presumably connected to the analog-to-digital-converter input. So I increased the gain of the buffer to 2, by lifting pin 6 (the - input) and connecting it to both pin 7 (the output) and pin 4 (ground) through 2.2K ohm resistors. It works fine, and I now have usable aftertouch - only half the pressure is required. But of course you shouldn't have to make this kind of modification in the first place.
df

Did you take any photos while doing the mod? It would be useful if you could post a couple of images showing the detail. Thanks!

Hey all,
 
I came across this thread researching trying to sort the aftertouch on my Edirol PCR keyboards (predecessor to the A-Pro series) - I'd previously stuck layers of insulation tape over the aftertouch strip, which had some effect, but it's worn down over time.
 
Thank you for all the advice here, I'd have been stuck without it - dfahrner's info gave me enough (along with researching IC 358 chips) to attempt the resistor mod, and proved successful (on all 4 PCR's I have). 
 
On the Edirol, the chip is labelled IC15 as opposed to IC8, but is located near the trim-pot as per the A-Pro series.
I ensured the trim-pot was turned fully clockwise, which makes it most sensitive.
 
My soldering isn't the best (which gave me some difficulty on the first one I modded), so soldered wires to the necessary pins - green on Pin 4 (earth), red on Pin 6 (- Aftertouch Input), yellow on Pin 7 (Output). For information, the chip also handles pitchbend (Pin 8 handles power for the chip - mess with this and it stops working...)
 
In terms of resistors, I used a 220k between Pin 6 and Pin 4 (red to green), and 100k between Pin 6 and Pin 7 (red to yellow) - I found this gave more sensitivity, and there is always the option to reduce sensitivity using the trimpot or software/synth you use. Once I confirmed it worked, I taped it up with insulation tape, to the bottom of the casing.
 

 
Obviously, modding is at your own risk, but hope this helps people :)
 
Cheers,
 
Rob