Velocity Curves - Rev2 - Attempt at measurement and graph

[Kusho]

I tried to get a better sense of the different types. Using a plugin in Reaper I recorded velocity input for each curve. I tried to increase velocity evenly each time, but I'm no concert pianist!!
The main points are:
Minimum velocity for each curve is clear:
1: 1
2: 10
3: 27
4: 33
Etc
These were the slowest I could play those notes. So it's not just the curve that changes, but the veolcity range available.
Curve 6 has minimum velocity of 50 !!
The curves that are easiest to play loudly seem to be 3 and 8.
The actual slope of the curves is difficult to determine in such an unscientific measurement. But 3 seems very steep.
1 and 8 have the most dynamic variation it seems. I'm not sure of the difference between these two, and it seems strange that they are at opposite ends of the available settings.

[Jimmy?]

Thank you for taking the time to do this. I've always kind of wondered how they differ.

[creativespiral]

Thanks @Kusho - Cool to see a visual representation.

A few years back, this was discussed, and @cbmd mentioned some info from code comments:

Thread is here:
https://forum.sequential.com/index.php/topic,1945.msg31000.html#msg31000

Velocity:
Curve 1: linear
Curve 2: 1.5 exponential, 0 offset   * Kusho graph shows 10 offset
Curve 3: 2.2 exponential, 10 offset   * Kusho graph shows 27 offset
Curve 4: compressed, 26 offset   * Kusho graph shows 33 offset
Curve 5: custom curve, 33 offset  * Kusho graph shows 33 offset
Curve 6: 1.5 exponential, 33 offset  * Kusho graph shows 50 offset
Curve 7: 2.3 exponential, 50 offset  * Kusho graph shows 33 offset
Curve 8: compressed, 33 offset   * Kusho graph shows 0 offset

To me it appears that maybe the reported code comments might have just been offset by one for which Curve they apply to?... maybe a zero = one -- offset for the curve number?

Pressure:
Curve 1: 3.0 exponential, 1 offset * Noted as being incorrect previously
Curve 2: 1.8 exponential, 1 offset
Curve 3: custom
Curve 4: 3.0 exponential, 1 offset


Seems there are some discrepancies.   Perhaps the code comments were incorrect (or offset)?, or changed algorithms / scaling since 2018.   Would be interesting to get a confirmation of this data for both velocity and pressure - @cbmd, @pym - do you know if the curves have been changed since this was reported a few years back?  Is this data still accurate?   

This is a question that comes up for all pretty much all of the Seq instruments on forums or social... would be nice to have the official info, since its pretty hard to test this objectively without industrial grade haptics measurement tools...  I know you can always just "go with what feels the best to you", but its nice to have the tech info.   

[Kusho]

Thanks @Kusho - Cool to see a visual representation.

A few years back, this was discussed, and @cbmd mentioned some info from code comments:

Thread is here:
https://forum.sequential.com/index.php/topic,1945.msg31000.html#msg31000

Velocity:
Curve 1: linear
Curve 2: 1.5 exponential, 0 offset   * Kusho graph shows 10 offset
Curve 3: 2.2 exponential, 10 offset   * Kusho graph shows 27 offset
Curve 4: compressed, 26 offset   * Kusho graph shows 33 offset
Curve 5: custom curve, 33 offset  * Kusho graph shows 33 offset
Curve 6: 1.5 exponential, 33 offset  * Kusho graph shows 50 offset
Curve 7: 2.3 exponential, 50 offset  * Kusho graph shows 33 offset
Curve 8: compressed, 33 offset   * Kusho graph shows 0 offset

To me it appears that maybe the reported code comments might have just been offset by one for which Curve they apply to?... maybe a zero = one -- offset for the curve number?

Pressure:
Curve 1: 3.0 exponential, 1 offset * Noted as being incorrect previously
Curve 2: 1.8 exponential, 1 offset
Curve 3: custom
Curve 4: 3.0 exponential, 1 offset


Seems there are some discrepancies.   Perhaps the code comments were incorrect (or offset)?, or changed algorithms / scaling since 2018.   Would be interesting to get a confirmation of this data for both velocity and pressure - @cbmd, @pym - do you know if the curves have been changed since this was reported a few years back?  Is this data still accurate?   

This is a question that comes up for all pretty much all of the Seq instruments on forums or social... would be nice to have the official info, since its pretty hard to test this objectively without industrial grade haptics measurement tools...  I know you can always just "go with what feels the best to you", but its nice to have the tech info.   

Thanks for this.
Then maybe velocity curve 8 is the linear one, not curve 1, if they are all offset by one and bump the last one around to the top again.