"Speed" of rotary knobs

[mattrixx]

I'm still not sure what I noticed is real or not, hence the question. However, I don't have a real problem with the way it is working at the moment in ProTools either.

[yoerik roevens]

anyone else noticing when using the eq3 plug-in knob 11 (mf q/freq) acts extremely jumpy?
Most likely related to this: http://duc.avid.com/showthread.php?t=384170

[mattrixx]

Just checked here and is working as it should. No jumpiness

[yoerik roevens]

sure? are you testing with the eq7 plugin? (other plugins work fine here as well) Do you experience the bug as described where 'Q' and 'freq' are not shown when you toggle between them?

[Audioartist]

Originally Posted by J-S-Q

This raises the question.... does the feel/acceleration of the knobs get dictated by the Eucon software itself, or by the individual DAW software that is being used?

I don't think it is like that on the S3L, so it is probably the eucon software.

[J-S-Q]

Originally Posted by Audioartist

I don't think it is like that on the S3L, so it is probably the eucon software.

Well it may be different on the S3L but that doesn't necessarily mean that it's all down to the Eucon software. What I mean is... do the knob kinetics differ between Pro Tools/Cubase/Logic etc.? Or are they all identical (because it is 100% dictated by Eucon itself, and NOT the individual DAW implementation of Eucon)?

[Michael Carnes]

I'm sorry that I can shine any light on the particular question. But on the general topic of knob ballistics I can share some experience and a funny story. The story's old, but I doubt things have changed much. Knobs are attached to optical encoders that output a very small value when the knob is turned. In my day it was only 4 bits. This didn't actually give you a physical position for the encoder shaft. Instead it could only tell you which way the shaft had turned. And if you didn't check very often (every few milliseconds) you could get the wrong direction out of it. All you had to do was turn the knob too quickly. This means that you've got to experiment with spinning the shaft to make sure you can get an accurate reading at human rates. I always wanted to clamp the shaft in a drill chuck, but they wouldn't let me.

The next thing you had to do was integrate this change over time and compare this to the range of the parameter you're trying to control. You don't want the knob to be so sensitive that you overshoot values when there are only ten positions. But you don't want to have to crank, crank, crank when you're trying to cover a range of 1 second by milliseconds.

The ideal solution to this is ballistics--start out slowly and then pick up speed if the user turns more rapidly or over a longer period of time. I learned this was easier said than done.

The first time I ever did this myself was when I wrote the operating system for the old PCM-80. I started out with the mechanical idea of the flywheel. I managed to create an ideal software model of the flywheel and quickly learned it wasn't the right thing. The more you turned the knob in one direction, the faster the value would change. It took maybe 10 seconds to get it up to speed--just like a flywheel. Then if you stopped turning, the thing would take another half minute to grind to a halt. Or if you started turning in the other direction, you could slow it down more quickly. But your mind would trick you into almost feeling the resistance of the flywheel as you fought against it. It was fun to play with in a stupid sort of way. It didn't take long for me to figure out the appropriate heuristics to make it work the way it should. But I still get a laugh out of the memory. Perhaps it's a good thing I don't get around the hardware much any more!

[mattrixx]

That's fantastic.

[J-S-Q]

Just thought it was note worthy that this is now officially the longest thread in the S3 forum.

Come on Avid... you can do this for us! YES you CAN! :)

[teedeecee]

Originally Posted by Exponential Audio

It took maybe 10 seconds to get it up to speed--just like a flywheel. Then if you stopped turning, the thing would take another half minute to grind to a halt.

Could we add this "Flywheel Ballistics Mode" to our request? ;-)