Forum rss-feed


Help: Aftertouch hypersensitivity interferes with velocity in playing.

Most Recent

written by: dhjdhj

Yeah...I looked at Bidule once but I find Max far more powerful. As soon as I'm done with an upcoming show, I will write up a new blog entry in my site on this topic.

written by: bhinton

Sun, 29 Jul 2012 08:22:34 +0100 BST

This is on a Tau - but I suppose it would apply to any model.
THe tau sends aftertouch data immediately on striking a key - even if the key is struck so softly that it barely registers a velocity value. I use velocity and polyphonic aftertouch promiscuously in my synth programming (mostly this is Zebra, Plogue Bidule, Bazille, lately a little Reaktor) to control independent parameters. Is there any way to set things so that one can use the marvelous range the eigenharp makes possible for BOTH velocity and aftertouch, but still retain the ability to play the full velocity range - carefully avoiding actually depressing the key of course - without accidentally sending aftertouch? Maybe a setting that establishes a "safe zone" for a very small zone of key travel?

written by: carvingCode

Sun, 29 Jul 2012 14:06:33 +0100 BST

What about setting the decimation rate for after touch in the Routing Matrix? Does this help?

written by: john

Sun, 29 Jul 2012 14:29:15 +0100 BST

I'm not sure exactly what you mean by this - 'carefully avoiding depressing the key' means that there's no way to actually calculate a MIDI style velocity. Do you mean that you don't want aftertouch sent for a brief while as the note starts, and then for it to return to normal after a preset time? Or do you want to be able to scale back the yaw sensitivity (or pressure sensitivity post the velocity calcuation) for a note startup time window (which if the yaw is being used to generate aftertouch would give you the same effect)? If so then it's possible we might be able to put something like this in for 2.1, would be nice to be very sure of what is wanted before writing anything though.

I hate MIDI. I can't believe that it's 2012 and we're having these kind of discussions, I find it surreal. I loved MIDI in 1987, but that's 25 years ago and we're still stuffing our performance data down a protocol designed for 8 bit microprocessors and 31.25k serial links. It's like playing Beethoven on a 4 bit audio device. The idea of a velocity at the start of a note is extremely crude and plain wrong in so many cases, it causes so many problems that we spend endless effort bodging around. Sigh. Complaining about the weather I guess!


written by: bhinton

Sun, 29 Jul 2012 18:12:51 +0100 BST

Thanks so much for the quick responses!
Side note to John - I am just blown away that YOU responded, and so quickly. The customer service and community aspects of Eigenlabs are just amazing. If only more companies were so engaged! Thank you.

@CarvingCode - no, alas, the decimation rate doesn't help.

@John - After watching your presentation on the eigenkey I think I understand the problem a bit better. I was under the false impression that while aftertouch is calculated based on pressure, velocity was calculated based only on the force with which the surface of the key is hit, independent of key pressure. Now that I see the way the key works, this issue is less mysterious, though it makes it hard to concoct a good alternative scenario!

The "delay" solution you describe would be a fantastic option - but the aliveness of the key is such a marvelous thing, I'd hate to see it mucked with in any permanent way, so I'd love it ONLY if it was an option that could be set or turned off in the software.

The idea of using yaw as aftertouch is a great suggestion - might solve the whole thing, actually. If the yaw is set to send AT in one direction only, I could simply alter the way I hit the key when the independent AT is desired (favoring the opposite side of the key from the yaw direction at first).

I'll probably work up something in Bidule eventually as well re a remapping of aftertouch values or even a delay on sending them - if people are interested (it seems at least a few use Bidule here) I'll post any layouts that result.


written by: dhjdhj

Mon, 30 Jul 2012 05:19:57 +0100 BST

I have been struggling with this problem as well as I had not come up with a decent way to infer initial velocity from the OSC data received from EigenD into my Max environment.

I finally made a hack to deal with this. I have MIDI turned on for the Alpha and the 120 keys on the Alpha have MIDI note numbers 1 through 120 assigned respectively.

So when a key is initially pressed and I start receiving OSC data for that key, I don't pass the OSC data on to the main system until I have received the corresponding MIDI Note On event for that key, at which point I now have a reasonable velocity value. That value is passed on and used as desired. OSC data is then allowed again for that key. I ignore any other MIDI data coming in (including NoteOff events)

written by: bhinton

Mon, 30 Jul 2012 06:10:06 +0100 BST

Good strategy. Are you doing this in Max?

Today's attempt at a bidule hack for it was frustrating. My first thought was isolating the aftertouch signal from eigenD and putting it through a midi delay, but as I should have known that continues to delay ALL aftertouch. A semi-useful alternative involved remapping the aftertouch response curve so that it doesn't start until around 25 and stays compressed until a midrange value - but this doesn't feel natural while playing, which is the point. I suspect I'm reducing the at signal back to low-res midi in doing this as well, though the particular patch I was using today would have masked this so I'm not sure.

I'm also now intrigued, however, by the idea of exploiting this simultaneous velocity/aftertouch reality in some new patches. Treating the key as an already charged object that becomes activated by the finger rather than as an actual key.

written by: bhinton

Mon, 30 Jul 2012 06:10:55 +0100 BST

Sorry, missed that you *did* say this was in Max.

written by: dhjdhj

Mon, 30 Jul 2012 06:14:10 +0100 BST

Yeah...I looked at Bidule once but I find Max far more powerful. As soon as I'm done with an upcoming show, I will write up a new blog entry in my site on this topic.

Please log in to join the discussions