Topic by stillwaiting

  • Avatar
  • stillwaiting
  • Member
  • Offline
  • Registered: 18-09-2021
  • Posts: 6

Topic: comparing audio curves Roland FP-10 and Pianoteq

Hi all,

I have Roland FP-10, which action I like, but don't like the sound. On the other hand, I love Pianoteq sounds, however always was struggling to tune it up to get good playability, especially for pieces that require pretty wide range of articulation and dynamics (e.g. Debussy: Golliwog's Cakewalk).

I spent hours and hours playing with the velocity curve, dynamics controller, equaliser, reverb and became desperate enough to compare audio curves of outputs produced by my default FP-10 sound and by Pianoteq (replayed MIDI of what I've recorded during FP-10 recording, in parallel). To my surprise, I noticed the following anomaly:

FP-10

In FP-10 output (see image above), I can clearly see that the point A is the most loud of all the sounds.

However, if I take Pianoteq's curve of it, I can observe completely different picture.

Pianoteq

As you can see, the dynamic range that Pianoteq produces has completely different structure. Here the same point A is far from being the most loud.

I'm wondering if it is some limitation of FP-10's MIDI controller? Am I doing something stupid? Currently in Pianoteq I'm using the default flat velocity curve and 60 as dynamics...

Thank you.

Reply by MeDorian

  • Avatar
  • MeDorian
  • Member
  • Offline
  • Registered: 28-06-2015
  • Posts: 240

Re: comparing audio curves Roland FP-10 and Pianoteq

Hi stillwaiting. It looks like the Limiter in Pianoteq is changing the peaks, and your dynamics set to 60 is changing the overall shape.

60 dynamics seems high, good on you though for trying this.

Nick

Last edited by MeDorian (18-09-2021 23:30)

Reply by MeDorian

  • Avatar
  • MeDorian
  • Member
  • Offline
  • Registered: 28-06-2015
  • Posts: 240

Re: comparing audio curves Roland FP-10 and Pianoteq

To add to the my last comment, if you are trying to match Pianoteq to the Roland FP-10 internal sounds (volume, dynamics etc), then try Pianoteq dynamics closer to default, 40 or lower, this might be a closer match. I still prefer the lower dynamics settings, 33 for example on some older Pianoteq models like the Erard and Frenzel. Remember to look at the output meter in Pianoteq, the Limiter can be seen working (or over working) from a line that appears from the right hand side of the display.

Nick

Reply by stillwaiting

  • Avatar
  • stillwaiting
  • Member
  • Offline
  • Registered: 18-09-2021
  • Posts: 6

Re: comparing audio curves Roland FP-10 and Pianoteq

Thanks, Nick.

I tried to turn off the limiter and played with the dynamics, no luck at all.

I dug a bit deeper and checked what are MIDI velocity Roland FP-10 sends to Pianoteq and then matched it with the FP-10's audio curve, and....

FP-10 audiocurve with MIDI velocity

How on Earth could note with velocity 95 sound quieter than with 86?

Given that the action/playability of FP-10 is subjectively good, I'm now inclined to think that its MIDI interface is crap... unless I miss something.

Last edited by stillwaiting (19-09-2021 01:41)

Reply by stillwaiting

  • Avatar
  • stillwaiting
  • Member
  • Offline
  • Registered: 18-09-2021
  • Posts: 6

Re: comparing audio curves Roland FP-10 and Pianoteq

Or maybe there's something to deal with the off notes that could be tuned?... Cause I believe the previously pressed notes contribute to the spike, however for Pianoteq curve it looks as if it might not...

Last edited by stillwaiting (19-09-2021 01:47)

Reply by YvesTh

  • Avatar
  • YvesTh
  • Member
  • Offline
  • From: France
  • Registered: 30-06-2020
  • Posts: 263

Re: comparing audio curves Roland FP-10 and Pianoteq

Hello,
I also play on the fp10. Like you I like the action but I don't like the internal sound very much and I prefer by far the pianoteq sound. What internal velocity curve of the fp10 do you use? This has a big influence on the midi message sent to pianoteq.
On the other hand I don't think that an audio level curve as you measured it is the best method of comparison, the final level is only one element of the sound and the multiple harmonics of a note (at very different levels) are not represented on this curve. You would have to establish the spectrum of each note in parallel.
Currently I use (0,0)(15,27)(30,47)(50,62)(70,75)(90,90)(127,127) on pianoteq with the internal "standard curve" in fp10 but it is very subjective and can change this curve sometime.

Translated with www.DeepL.com/Translator (free version)

Last edited by YvesTh (19-09-2021 09:04)

Reply by MeDorian

  • Avatar
  • MeDorian
  • Member
  • Offline
  • Registered: 28-06-2015
  • Posts: 240

Re: comparing audio curves Roland FP-10 and Pianoteq

I would first go to file in Pianoteq (top left) audio/midi and look at MIDI events. Play some notes and see what the lowest value is, and the highest. If 0 value and 127 are achievable the FP-10's velocity curve is fine, if not, as Yves Th says, 'what internal velocity curve on the FP-10 are you using?'. Maybe it's default curve is not ideal.

Regarding the Limiter, it's there for good reason, to prevent audio clipping. If using a high value dynamics, 60 for example then the limiter will be more active. You could though try to reduce the volume in Pianoteq until the Limiter is less active.

I would use the preset values for dynamics until you are sure the MIDI values are as they should be. Also, using a high value dynamics isn't an indication as to ones own playing skill level, if Glenn Gould was alive today using Pianoteq, I really doubt he would set the dynamics to 100.

Nick

Reply by MeDorian

  • Avatar
  • MeDorian
  • Member
  • Offline
  • Registered: 28-06-2015
  • Posts: 240

Re: comparing audio curves Roland FP-10 and Pianoteq

The MIDI values and audio volume differences on your last post stillwaiting will be the Limiter. It can make a velocity value of 95 quieter than 86 if the Limiter is already limiting an audio section. It is not instant both in attack and release.

Nick

Last edited by MeDorian (19-09-2021 19:10)