Need improvement for Pianoteq 6.5 for release sound of sustain pedal ?

Yes, I look forward to the development of Pianoteq's sustain pedal/dampers. Beyond the realism being a pedagogical tool, pianists sometimes use the damper's whoosh and thud sounds in several ways. Pedal accents can add quite a bit of drama to a performance. I wouldn't put this into the category of more novel effects like drumming on the fallboard, or strumming/plucking the strings.

I would guess that in order to properly reproduce all of these subtleties, the damper's effect on the strings would have to be modeled in significant detail. It would vary depending on the impedance of the sound board, of the string length, and I'm sure a host of other variables. Also, beyond the whooshing sound the dampers make when lifted from the strings (at a variety of speeds) and the thudding when dropped on the strings (at a variety of speeds), there's also the re-pedal accent -- where the dampers are lifted the same instant they hit the strings, so the resonance from the damper thud is sustained (a VERY pronounced effect).

Beyond this there is the effect of the pedal mechanism bottoming out rapidly, where the pedal/pedal mechanics produces a percussive sound (sometimes this is what people refer to with pedal accent). To some extent this could be produced in real life with a MIDI pedal. But on a real piano this sound can also be sustained, and it will resonate throughout the piano. It can be performed along with the re-pedal accent for an even stronger effect.

These aren't little things to add... it might even require something like... Damperteq!

Hello Philippe,
Was modeling for the sustain pedal already done for the model ptq Bluthner (older model)?
In the case of the above recording of the release of the sustain pedal on a Bosendorfer 200, the release was too energetic (here clumsy action of a non-professional pianist) but not exaggerated to the point of putting the pedal in abutment on the pedal's lyre. (or knowking the cabinet).
On our piano, the sound of Knock when the pedal is released - after careful observation of the mechanism - essentially the fallout of all the dampers on the strings.
I made our tuner come back especially a second time after adjusting the keyboard, especially to see the level of this knock that seemed excessive for me (even with a clumsy pianist). He told me that the fallout noise of all the dampers and its very strong resonance with the cabinet (much stronger than the noise generated by the dampers on the strings) was normal on a grand piano - because of the mechanism current fallout dampers under the effect of gravity.
I note however that on our grand piano Erard, this Knock is much more discreet, even with a bad pianist ...
On my record above with SE8 small vertical diaphragm microphones about 45 cm above the strings (actually projection of the upright pickups about 10 cm behind the dampers), the significant difference in level (over 20db) between the knock of the dampers compared to the initial "dzing" when the dampers rise above the strings, does it seem normal to you?
(NB the dampers were never changed on this Bosendorfer from 1983).
If the difference is not normal, it may be necessary that I make a reference recording on another Bosendorfer 200 (we have one at the Dijon Conservatory).
If this difference in level is normal, the separate modeling of the release of the sustain pedal - not only its pressure (and ideally a new parameter for the level of this pedal release in Pianoteq std / pro) would be pedagogically useful for training amateur pianists.

Waiting for a possible? evolution, I tried to create an additional .fxp dedicated only to the release of the pedal to use in a Daw (Reaper for example) in conjunction with a normal .fxp, unfortunately it is difficult to intercept the MIDI events allowing to identify a release of the sustain pedal, it would be necessary for this a treatment of the midi events of the controller 64 to capture the decrease in the value of the event (acceleration of the release) when the release of the pedal is poorly accompanied by the pianist . With Reaper I can only (yet?) make a filter on the current MIDI events, not unfortunatly in function last previous MIDI events from the controller...

Best regards;
Bruno

PS: Thank you for the proposed workaround with
the Kremsegg collection(Bechstein)  which works better to make a silent .fxp dedicated only to the sound of the release of the sustain pedal (used under a DAW in addition to another fxp) By inverting (for this fxp) the direction of the pedal, after modification of its curve and a little equalization, the result is pedagogically usable ... (although less convenient than a single .fxp in pianoteq standalone that could specifically set the level and the noise of the release of the sustain pedal)

Good day!
There is a dependence of the damper sound on the impedance settings, unison widths and others .. but there is no dependence of the damper sound on the position of the damper on the string. It seems to me that this moment of modeling should not be overlooked ..

Another point - when two microphones are located in the same place, the entire sound of the instrument sounds mono, but the sound of sympathetic resonance and the sound of dampers sound voluminously.

I understand Nathan, it makes perfect sense.

I hadn't played the U4 in a while, cranked up the "sustain pedal noise" and yes, there is a definite knock built in (not to be confused with the repedal accent). Aside from the knock, which is actually hard to avoid, the U4's sustain pedal noise is much smoother than the Steinway D's. If you pedal very slowly with the Steinway D, it sounds like the pedal modeling noise kicks in over and over for every new value registered by the midi pedal. This isn't really noticeable normally, because the pedal noise is very quiet and it would be unnatural to pedal so slowly. It's interesting though, when cranking the Steinway's pedal noise to the max, I think I can hear a tiny bit of added volume/damper sustain when pedaling rapidly (a hint of repedal accent), although it might be placebo. I can't actually hear anything changing when adjusting the "damper noise" parameter...

More and more I'm thinking the sustain pedal's realism is severely limited by either MIDI resolution, our MIDI pedals, or both. If Pianoteq made the sustain pedal noise more dynamic, I don't think it would be possible to control it with a MIDI pedal. With normal pedaling, I'm averaging 2 or 3 numbers registered in between 0 and 127. That's just not nearly enough resolution.

I was assuming it does this already, factoring in the speed it takes to get from 0 to 127 to 0. But a major problem seems to be the MIDI resolution simply isn't fast enough to record the speed beyond a fairly normal rate. I guess this has something to do with the rate of MIDI "ticks"? Maybe this is something that can be adjusted by software? Perhaps MIDI 2.0? Or maybe it's just a physical limitation of the sensor in MIDI pedals?

If it's not too difficult to get around this issue, then yes, I think it could open up a lot of possibilities for the sustain pedal.

Interesting, have you (by your continuous pedal) found the software senses a fast combination foot stomp and release (resulting in increased damper noise), or replicates a percussive strike to strings from a quick re-pedal?

I suspect the software Damper Noise parameter of the Pro version affects damper noise from piano keys as well as from the sustain pedal and audibly the volume from your release of it, if you ever decide to release it very quickly.

Possibly, Damper Noise as a parameter selection is a reference to damper mechanisms more so than the sound of the felt dampers themselves.

Obviously, I need some clarification!

But, my whole schlep about the sustain pedal registration in Pianoteq is that it might someday sense pedal velocity; just like key velocity has been sensed by methods implored within keyboard hardware, it could happen only inside the software from the half-pedal and continuous pedal capable units.

I see no reason, why it can’t today do that for pedals with a given virtually infinite number of potential sensory points available to software, compared to maximum tri-sensor capabilities left to keyboard keys in hardware!

Similarly, cannot software to some extant make up for a keyboard that lacks tri-sensor technology, one such as mine, whenever I attempt a fast repetition on a single key, and, especially from a depressed sustain pedal!

Personally, I consider myself a fan of the new modelled pedal noise.  To me it's fantastic!  Too bad, nobody considered to note it in a change log and consequently keep either you or me updated to any progress being made.

Certainly, I for one would like to see a new parameter affecting the pedal velocity effect on dampers specifically and possibly another varying damper felt hardness like one might control hammer hardness at a given velocity, if for PIANOTEQ PRO these too are even remotely possible.

Hopefully just as keyboard manufacturers offer a selection out of soft, medium, and hard keyboard key settings, MODARTT in its near future might do something similar just for pedal responsiveness!

I’m being now optimistic, of course.

The problem here (in the case of the modelled sustain woosh), is that the model needs a rich MIDI input. If for example your sustain pedal sends the following (poor) events:
time                        = [0  0.1  0.2  0.3  0.4  0.5  0.6  0.7]
midi sustain velocity = [0   15   30   45   60   75  100 127]
there is no way to make this data smooth, because for example at time t = 0.3 we are not able to predict the next event (here 60, but it could be also 30). Of course one could smooth the data, but at the the unbearable cost of a long delay, which wouldn't be acceptable for fast pedalling. We have tried to find a solution to this problem, to find compromises, and came to the conclusion that there is no solution able to treat a case similar to the one above, except in oversimplifying the model (for example considering only velocities 0 and 127, hence one single type of event), which would be a pity for those having good progressive pedals, sending much finer messages than above.

We have several pedals, and the Roland is the best one (gives the finest resolution). I'm not at office so cannot tell the model for now.

The Roland DP-10 is their single pedal with the flip around rubber base plate. It can be switched from on/off to continuous mode. Roland also makes a triple pedal, the RPU-3. I believe all three pedals for that are continuous.

https://open.qobuz.com/track/64961197

Great record! How well the damper is heard at the end of the track. At the end ... from the 14th minute ... when the quiet sounds begin. On bass notes, the sound of a damper is heard.

Amen Ptah Ra, sorry for some inconvenience. I will lay out the piece that I had in mind. I hope I do not violate the law ..

https://www.forum-pianoteq.com/uploads....olodos.mp3

And pleasant hammer felt ..

thanks guys.

so: with a DP-10 I will be able to make the pedal sounds sound different volume with different foot velocity correct?

can a standard usb-keyboard with sustain/controller in handle this or do I need a separate interface for this?