Roland VH-13 emits false bow hits on chicks

[dmitri]

Can you please test the new algorithm in this version:

megadrum_20140510.zip

Disregard my previous description for MinVelocity and MaxVelocity.
MaxVelocity is ignored in the firmware.
Try various MinVelocity values between 100 and 1000.

[Alek_A]

Dmitri, I tested the firmware, it is much better. Should the false hit suppression work at the testing stage? Cause it don't by now.
- I can't get low velocity chicks. If I set MinVelocity to 1000 I can, but I have to press pedal very slowly. We should not filter out low velocity chicks, because sound from open hits keeps going. I asked for this because my chickthresh set to the level when hi hat is not closed in order to achieve best false hit suppression and velocity picking. In the new algorithm it may not be necessarily.
- Sometimes I get double chicks.

Your old algorithm picks velocity very adequate. I've implemented a little addition to it as a VST plugin and it seems to work just perfectly in my setup! I can't imagine how it could be improved.

[dmitri]

I've just release a new "official" firmware version. Please use it rather than the test version.

[Alek_A]

Sorry, no luck with new firmware. I only can get either 1 or 127 chick velocities.

[dmitri]

1. When you're in Pedal->MinVlcty, what "raw" velocity does it show when you press the pedal quickly?
2. When you're in Pedal->MinVlcty, what "raw" velocity does it show when you press the pedal slowly in 1 second?
3. When you're in Pedal->ChckDead, what "raw" value does it show when you press the pedal quickly?
4. When you're in Pedal->ChckDead, what "raw" value does it show when you press the pedal slowly in 1 second?
5. What did you set MinVlcty, MaxVlcty, ChckDead and ChkCurve?

[dmitri]

Another iteration. Please use latest firmware version 20140518.

To configure:
1. navigate to Pedal->MaxVlcty and press the pedal as fast as possible. Note the "raw" velocity value you see and set MaxVlcty just below this value. In my case it was around 900.
2. navigate to Pedal->MinVlcty and press the pedal as slow as you want it to be still registered as chick. Note the "raw" velocity value you see and set MinVlcty just above this value. In my case it was around 400.
3. navigate to Pedal->ChckDead and press the pedal as slow as you want it to be still registered as chick. Note the "raw" timer value you see and set ChckDead just above this value. In my case it was around 600.
4. navigate to ChkCrv and set it to Exp2C.

Now test the pedal for chicks. And yes, it should suppress false hihat bow triggering.

[dmitri]

It seems there is very little interest in improving Chick/Splash handling. In fact only Alek_A is interested as far as I can see. I've spent quite a lot of time in the last few weeks working on the Chick/Splash handling in the new pedal handling algorithm but to work on it more effectively I need it to be tested promptly when I post new test or official firmware versions. So for now, even though I think there is still room for improvement, I have to pause working on this feature until there are more people willing to test it and provide feedback.

[Alek_A]

Dmitri, thanks a lot! I really appreciate you for what you do.
I'm sorry that I have too little time for testing now.
I've just did as in the instruction. In the picture - four hi-hat hits followed by chicks. (chicks are 57)

velocities - new firmware.png

Sadly velocities behave wildly still. There are some room appeared where you can get velocity between min and max, but most likely you get either min or max velocity, or some velocity above 100.
MinVlcty=1, MaxVlcty=930, ChkDead=1000
Sometimes I get up to 5 chicks within a single press.
Dmitry, in my opinion debugging of new algorithm is helpless until you get this hihat.
The little interest to chick handling is because many instruments generate chicks by CC4. Probably it's easier for users to just use that instruments.
Thank you for trying to help!!

[workroom]

I can help test this if you want to see the settings I use for a hall sensor.

[mdefiore]

I can help test this if you want to see the settings I use for a hall sensor.

I can help test this too.