Topic by YvesTh

  • Avatar
  • YvesTh
  • Member
  • Offline
  • From: France
  • Registered: 30-06-2020
  • Posts: 212

Topic: PHA-4 Keyboard (FP-10), keyboard regularity

I tried checking the velocity differences between all the keys on the PHA-4 keyboard of the Roland FP-10 ( only white keys at this time). To do so, I used a straight wooden bar to press all the keys at the same time and at the same speed. I did this test at several velocities but I didn't do it with the high velocities (you have to press very hard on the bar and I didn't want to break my keyboard). It should be noted that the comparison is done at equal speed and not at equal weight as it was done on the VPC1 keyboard. We can see a pretty good regularity all along the keyboard except for some keys. I don't think a real piano keyboard is more regular.
However it is possible to use a software like velpro to obtain a straighter line.

PunBB bbcode test

Reply by groovy

  • Avatar
  • groovy
  • Member
  • Offline
  • Registered: 12-06-2009
  • Posts: 717

Re: PHA-4 Keyboard (FP-10), keyboard regularity

Very smart, YvesTh!
How did you transfer the velocity values to the diagram? Recorded with Pianoteq and then manually inserted in a spreadsheet?
Thank you

Reply by YvesTh

  • Avatar
  • YvesTh
  • Member
  • Offline
  • From: France
  • Registered: 30-06-2020
  • Posts: 212

Re: PHA-4 Keyboard (FP-10), keyboard regularity

It is indeed very difficult with pianoteq. I used midi-ox, it allows to copy easily all the midi data. then I insert them in an excel file, process and classify the data to obtain a table.

Reply by groovy

  • Avatar
  • groovy
  • Member
  • Offline
  • Registered: 12-06-2009
  • Posts: 717

Re: PHA-4 Keyboard (FP-10), keyboard regularity

That has been inspiring. I did one first shot with my Korg B2 and used a shelf with the standard length of 120 cm to press all white keys at the same time.

I took, what I had available at the moment:

- Recording MIDI with Pianoteq.
- Importing the midifile into MusE Sequencer and saving it as part File (*.mpt) to get a raw text file.
- Importing this raw data into LibreOffice Calc and making a chart from the table.

The Korg seems to be calibrated a bit different than the Roland (after this first proof of concept) ->

https://i.postimg.cc/rwB8tL7W/Korg-B2-all-keys-pressed-with-a-plank-example1.png

Last edited by groovy (06-02-2022 00:41)

Reply by sigasa

  • Avatar
  • sigasa
  • Member
  • Offline
  • From: UK
  • Registered: 09-01-2008
  • Posts: 1,938

Re: PHA-4 Keyboard (FP-10), keyboard regularity

Thank you guys,

Very interesting

Warmest regards,

Chris

Reply by sigasa

  • Avatar
  • sigasa
  • Member
  • Offline
  • From: UK
  • Registered: 09-01-2008
  • Posts: 1,938

Re: PHA-4 Keyboard (FP-10), keyboard regularity

I'd love to see simple note for note sliders that alter the curve of each note expositionaly. As we can see from Yves Th's diagram, the differences in velocities note to note increase as the input velocities increase.

Just a thought,

Warmest regards,

Chris

Reply by groovy

  • Avatar
  • groovy
  • Member
  • Offline
  • Registered: 12-06-2009
  • Posts: 717

Re: PHA-4 Keyboard (FP-10), keyboard regularity

sigasa wrote:

As we can see from Yves Th's diagram, the differences in velocities note to note increase as the input velocities increase.

Hi,

I'm not sure, if the differences of the Roland are not in the range of randomness.

The lowest red line could be limited by the sensitivity of the keyboard, many keyboards can't transmit MIDI values lower than 10 - 15 reliably.

And we don't know, if the velocity mapping of the FP-10 had been factory default (normal, and not "soft" or "hard").

Last edited by groovy (08-02-2022 09:57)

Reply by YvesTh

  • Avatar
  • YvesTh
  • Member
  • Offline
  • From: France
  • Registered: 30-06-2020
  • Posts: 212

Re: PHA-4 Keyboard (FP-10), keyboard regularity

Hi,
Some informations :
The test I did is made with "normal factory default velocity" of FP-10
All keys allow all velocity values from 1 to 127.
It is not very easy to compare keys at high velocities because 127 is difficult to obtain with all keys ( 100-115 easily but 127 with with very very strong hit ).

If I considere average velocity is green line it seems that faster keys as "F3" looks like red curve and slower keys as "G4" like blue curve.

PunBB bbcode test

To avoid random effect I did 4 or 5 test at the same velocity and calcul the average. Differences between keys are real. But I had never heard them before I measured them. This keyboard is a good keyboard...

Last edited by YvesTh (08-02-2022 15:53)

Reply by groovy

  • Avatar
  • groovy
  • Member
  • Offline
  • Registered: 12-06-2009
  • Posts: 717

Re: PHA-4 Keyboard (FP-10), keyboard regularity

YvesTh wrote:

Differences between keys are real.

Of course, the parallel "ripples"! But sigasa and me meant the increase of that differences with higher input velocity. If this increase is significant (more than randomness), one explanation would be a non-linear velocity mapping inside the MIDI keyboard. Therefore the information was important, that a "soft" touch curve hasn't been used (by accident for example), thank you.

Probably the ripple is neglectable for the brain or easy to adapt. But for keyboards with uneven response I like sigasa's idea of a per-note-edit. - Doesn't that already exist in the Pro version of Pianoteq? I just remember the VPC-1 keyboard has a simple per-note-offset feature.

Last edited by groovy (08-02-2022 18:09)

Reply by sigasa

  • Avatar
  • sigasa
  • Member
  • Offline
  • From: UK
  • Registered: 09-01-2008
  • Posts: 1,938

Re: PHA-4 Keyboard (FP-10), keyboard regularity

groovy wrote:
YvesTh wrote:

Differences between keys are real.

Probably the ripple is neglectable for the brain or easy to adapt. But for keyboards with uneven response I like sigasa's idea of a per-note-edit. - Doesn't that already exist in the Pro version of Pianoteq? I just remember the VPC-1 keyboard has a simple per-note-offset feature.

There is no per note edit for velocity in Pianoteq Pro. However, one could perhaps use per note edit voicing which is possible in Pro. But it would be more helpful to have a per note velocity curve editor.

Warmest regards,

Chris

Last edited by sigasa (08-02-2022 19:06)

Reply by YvesTh

  • Avatar
  • YvesTh
  • Member
  • Offline
  • From: France
  • Registered: 30-06-2020
  • Posts: 212

Re: PHA-4 Keyboard (FP-10), keyboard regularity

I am aware that a correction of my keyboard is not really necessary (the differences of velocity generated by the imprecision of my fingers is more important than those of the keyboard). However I have installed springbeats-velpro which allows me to get a much flatter response) and the Placebo effect allows me to feel the keyboard well...

Reply by groovy

  • Avatar
  • groovy
  • Member
  • Offline
  • Registered: 12-06-2009
  • Posts: 717

Re: PHA-4 Keyboard (FP-10), keyboard regularity

YvesTh wrote:

However I have installed springbeats-velpro which allows me to get a much flatter response) and the Placebo effect allows me to feel the keyboard well...

:-)
Probably I would feel a Nocebo effect from the imagination, that an external per-note-MIDI-processing will add MIDI-latency and -jitter...

Last edited by groovy (08-02-2022 22:27)

Reply by sigasa

  • Avatar
  • sigasa
  • Member
  • Offline
  • From: UK
  • Registered: 09-01-2008
  • Posts: 1,938

Re: PHA-4 Keyboard (FP-10), keyboard regularity

groovy wrote:
YvesTh wrote:

However I have installed springbeats-velpro which allows me to get a much flatter response) and the Placebo effect allows me to feel the keyboard well...

:-)
Probably I would feel a Nocebo effect from the imagination, that an external per-note-MIDI-processing will add MIDI-latency and -jitter...


Groovy, what is jitter??

Warmest regards,

Chris

Reply by groovy

  • Avatar
  • groovy
  • Member
  • Offline
  • Registered: 12-06-2009
  • Posts: 717

Re: PHA-4 Keyboard (FP-10), keyboard regularity

With my own limited words:
MIDI events can have a precise constant latency in relation to real world events.
When MIDI events fluctuate around a mean latency, this is Jitter. The random character of Jitter can be more disturbing than a constant latency.

It can happen, that you are not aware of Jitter until play a MIDI environment, that is better in this regard ... it just feels more precisely and "tight".

Reply by sigasa

  • Avatar
  • sigasa
  • Member
  • Offline
  • From: UK
  • Registered: 09-01-2008
  • Posts: 1,938

Re: PHA-4 Keyboard (FP-10), keyboard regularity

groovy wrote:

With my own limited words:
MIDI events can have a precise constant latency in relation to real world events.
When MIDI events fluctuate around a mean latency, this is Jitter. The random character of Jitter can be more disturbing than a constant latency.

It can happen, that you are not aware of Jitter until play a MIDI environment, that is better in this regard ... it just feels more precisely and "tight".

Thank you Groovy. Very enlightening.

Warmest regards,

Chris