Awesome Groovy, very ingenious.

I wonder if there is anybody who knows what velocity a 50gram weight/100g weight/200g weight etc.. should trigger when placed at front of key and allowed to drop in a consistent manner on a well regulated grand piano??? This would certainly aid any velocity curve creation. Maybe this could best be found out on a hybrid piano, a Yamaha Avant grand for example???

Kindest Regards,

Chris

+1 for "ingenious" !   It lets you _compute_ the physical key velocity, as a function of horizontal wheel velocity. 

a) There's no industry-standard definition for "well-regulated grand piano".   Manufacturers do things differently, players' tastes differ, and piano techs have different preferences.

b) You'd have to be careful to measure the velocity _at the moment of releasing the hammer_.  Grand piano actions (I think) don't have constant resistance, as the key goes down.

c) I'm not saying the experiment isn't worthwhile.  But don't expect Piano A to give the same results as Piano B.

.     Charles

I wonder if there is anybody who knows what velocity a 50gram weight/100g weight/200g weight etc.. should trigger when placed at front of key and allowed to drop in a consistent manner on a well regulated grand piano??? This would certainly aid any velocity curve creation. Maybe this could best be found out on a hybrid piano, a Yamaha Avant grand for example???

Kindest Regards,

Chris

I geeked out about velocity again . I wanted to know the relation between the "mechanical" speed of my keys and its "electrical" representation known as MIDI-velocity. I don't know any methods for this and so I re-invented the wheel ...

Wow using a wheel is an awesome idea , very impressive !

Hi there,

I geeked out about velocity again . I wanted to know the relation between the "mechanical" speed of my keys and its "electrical" representation known as MIDI-velocity. I don't know any methods for this and so I re-invented the wheel ...

Rolled over the keyboard of my Kawai ES3 I got columns of note-on velocity values. Together with note-off for each note and the timespan between those events I tried to calculate the angular speed of the turning wheel at each note-on. Required were the diameter 185 mm an the trigger-points of my keys, that I measured with a ruler to be -9 mm for note-on, -5 mm for note-off, and around -12 mm for key down completely:

With some trigonometry and the combined values of five roll-overs at different speeds (slow -> medium -> fast) I got a non-linear response to the "mechanical" speed of the keys:

I don't know, whether this non-linear response it typical for the majority of keyboards with rubber-bubble-contacts or just for Kawai models like this ES3 (or if I just did the wrong math). It would be interesting if something is known around this topic. 

Thanks for your attention

I guess you could say the subject of MIDI velocity is "touch and go."  I would say what is more important is how musically and expressively you are actually able to play the instrument.  The is not way I know of measuring the "velocity" on an acoustic piano.  The keys either respond well or they don't.  With electronic instruments, the subtlety of response can be elusive.

I geeked out about velocity again . I wanted to know the relation between the "mechanical" speed of my keys and its "electrical" representation known as MIDI-velocity. I don't know any methods for this and so I re-invented the wheel ...

Rolled over the keyboard of my Kawai ES3 I got columns of note-on velocity values. Together with note-off for each note and the timespan between those events I tried to calculate the angular speed of the turning wheel at each note-on. Required were the diameter 185 mm an the trigger-points of my keys, that I measured with a ruler to be -9 mm for note-on, -5 mm for note-off, and around -12 mm for key down completely:

With some trigonometry and the combined values of five roll-overs at different speeds (slow -> medium -> fast) I got a non-linear response to the "mechanical" speed of the keys:

I don't know, whether this non-linear response it typical for the majority of keyboards with rubber-bubble-contacts or just for Kawai models like this ES3 (or if I just did the wrong math). It would be interesting if something is known around this topic. 

Thanks for your attention