What ever FL did to simulate hammer strikes physically translates to PianoTeq really well. I had to remove the "voicing" on my setup because of how PianoTeq interpreted the velocity.

In the end, the Grotrain, Bluthner, Steinway D on my setup sound the very real now, CC88 is very nice if your speakers can reproduce sounds like a real instrument, there's alot more variation to the frequencies that I think translates to more dissonance the "color" that I'm hearing that PianoTeq models.

The rest of the Piano Models, I'm not sure yet since some parts of them sound synthetic a few milliseconds after the hammer strikes.

But it would be an interesting double-blind experiment to see how few steps need to exist under those MIDI velocity curves before adding more of them makes no discernible difference to the trained ear. I'm really skeptical it would be more than 127.

I suspect that the number of steps needed is considerably less than 127. It would indeed be interesting to 'know', by some means, for example your suggestion of a 'double-blind' experiment. But, from my experience, such an experiment would be fraught with pitfalls of design and interpretation. For example, much of what we hear as emphasis depends on context. And organists, without the benefit of touch sensitivity, create stress through timing and in particular by manipulating the gap between notes.

I think it was Schumacher (of 'Small is Beautiful' fame) who said something along the lines that how certainly you could know something and how interesting a thing it was to know are inversely proportional.

I.

....

I doubt, by the way, that the hi-res velocity sensing (if that is what it is) would make the slightest difference to me. I work in Pianoteq with a restricted velocity range (ie less than 127 steps) because that is what seems to work best with the N1X, in my hands anyway. So I'm maybe working with the range 20 - 100 (can't remember exactly off-hand) and that's still plenty of resolution for me. Maybe I just play like an oaf, but I wouldn't be surprised if 20 velocity steps was enough. I certainly don't think I could reproduce 20 different 'weights' reliably.

Ian.

I've wondered about the resolution question, too. I mean, how many shades from ppp thru fff can we really control, no matter how expressive we think we're being? So your post has inspired me to create a few crude stepped velocity curves. Curiously, while anything under 8 steps sounds proportionately lumpy (as one might expect), by the time I get up to 8 steps -- one for each of the traditional dynamic levels (Velocity = [0, 16, 17, 33, 33, 51, 51, 69, 69, 87, 88, 101, 102, 114, 114, 124; 4, 4, 19, 19, 37, 37, 57, 57, 76, 76, 96, 96, 112, 112, 127, 127]) -- it starts to sound pretty damned acceptable. But it really does seem like the relatively smooth velocity curves are indeed a better handmaiden for dynamic expression, even if we can't control the dynamics with anywhere near 127-degree precision. I assume it's the continuous shades we hear that make for the more satisfying result. But it would be an interesting double-blind experiment to see how few steps need to exist under those MIDI velocity curves before adding more of them makes no discernible difference to the trained ear. I'm really skeptical it would be more than 127.

I'd just leave them be.

Amen Ptah Ra wrote:

That Kawai Novus NV10 seems to have settled for standard MIDI, as opposed to Hi-res CC#88 MIDI and unlike the Lachnit Mk23 88 Studio.  Which to the contrary along with PIANOTEQ permits the greater resolution among the two keyboards.

I myself may want to contact Robert Estrin and find out whether or not he’s aware of PIANOTEQ and wants to make his hybrid Hi-res CC#88 MIDI compatible.

Perhaps I’ll procrastinate or wait and see if he’s just got his own agenda, like Kawai.

Where big corporations are concerned sometimes you have to show some balls, man; demand satisfaction: customer satisfaction, even as a potential customer.

I just feel musicians today ain’t the eunuchs of Ancient Rome!

https://youtu.be/lxlJjcKJO9c

This is from a few weeks ago, my buddy just likes to make creepy sounds and the setup blew him away.... he was amazed that pianos could sound like that.
I added the Bose B2 Bass for deeper bass that you get from a 9ft.

https://youtu.be/0l8bM4ozhBc

I myself may want to contact Robert Estrin and find out whether or not he’s aware of PIANOTEQ and wants to make his hybrid Hi-res CC#88 MIDI compatible.

Perhaps I’ll procrastinate or wait and see if he’s just got his own agenda, like Kawai.

scherbakov wrote:

but better nouvo kawai...

what do you mean?

https://www.kawai-global.com/product/nv10/

what do you mean?

Once you get your keyboard, if you post a Hi-res CC#88 MIDI example, other forum members also will have something to compare it to some of the Yamaha Disklavier MIDI XP file examples.  Many are at this site and for them to tell maybe once and for all whether or not MIDI having Hi-res CC#88 data from the Lachnit Mk23 88 Studio can sound just as good if not better, and, as detailed as any Yamaha Disklavier recording (MIDI XP).

Incidentally, Mk4UmHa, I share your opinion about Bose system speakers.  They really reproduce acoustic instrument sounds accurately.  They came to my attention initially while I was stationed at Mannheim.  (But, before then I had been a member of a few orchestras.  Which immediately came to my mind again when I for the first time listened to the clarity of some of the Bose speakers.  They had sounded to me as though I were while in Mannheim really listening to a live orchestra instead of just speakers.)

Man, simply amazing!

Have you tried the Casio/Bechstein Grand Hybrids? They have real hammers that strike sensors. I am not a classically trained pianist, but I find that their keyboard feels better than any digital piano I have ever tried... I think someone on this forum (bm?) uses a GR500 to control Pianoteq. I wish they made a controller-only version of that!

scherbakov.al wrote:

In my opinion, if the keyboard does not have a real grand piano mechanism, then you will not get enough true feel when playing. The problem is that keyboard "simulators" usually are "sticky" or "heavy" to play. This applies to all and even the best representatives of such keyboards. This heaviness arises from the almost direct relationship between the key and the hammer. You need to "press" the key with sufficient force to the very bottom. And press all the way down. It's just that real keyboards have a lot of inertia, and the pianist usually plays with it when playing. Electronic instrument keyboards have little or no inertia. This is one point. The second point is that after pressing the hammer will be caught and holding the key down requires a very small amount of force. In simulators, holding the key on the bottom requires full force to hold the hammer in place. Therefore, simulators are tiresome for a long game, they are more "dumb" in sensations, they do not allow you to easily perform fast passages on the pianissimo, have ease of performance in quick and easy passages, and practice fast rehearsal repetitions well. A real keyboard can be quite heavy, but not tedious to play, unlike digital instrument keyboards.

Good research:

https://hal.archives-ouvertes.fr/hal-01193708

It seems to me that the acceleration graph reflects well the peculiarities of the keyboard and mechanics. It is in the relationship of the applied forces and accelerations that one must stop the gaze when researching and developing keyboards.

The optical sensors under the keys are impressive in their technology, but they cannot fully simulate the movement of the hammer, which has many complex movements due to the peculiarities of the mechanism + vibration mode of the hammer itself.
From this it turns out that the best option is to use optical sensors from the project:

https://github.com/ekumanov/cybrid/

Here you have a real keyboard, getting information from the hammer in the right place and optics that do not introduce any distortion in the measurement of movement. In my opinion, this is the best!

hee hee .. it's funny:

https://youtu.be/ZhbkRYVPkSc

but better nouvo kawai...