Prototype with infinite pivot lenght

RIP Then, by moving the seesaw back and forth (from pianist's view: left and right in the video) one could make this thing change the real touchweight of the keybed, not just the simulated one via the response of the audio! Wow! Isn't that amazing?

Not so fast, I think... You change the touchweight, but you also change the leverage of the key (ie, how much of key vertical down-displacement translates into vertical up-displacement of the action). So, you can only monitor the actual key movement to create MIDI events, since the action movements will change with each adjustment, and the calibration would be all over the place.
You can already do this in an acoustic grand action: if you change the screw-head type of capstan by a longer metallic piece, you can move the keys back and forth to different positions: that will also change leverage and alter touchweight. But now your hammers would be either flying off because of too much momentum, or would be still because the angular displacement of the whippen is not enough to make them move.
Boy, am I a sceptic or what???

vagfilm Boy, am I a sceptic or what???

Being skeptic is great in general, and typical of scientists like you are. Otherwise one would just waste all their time in conspiracy theories and creative nice-sounding pseudoscience.

And in this case, you are correct (but there is a twist). In fact this "touchweight regulation" won't work on a real acoustic action or even on a hybrid with hammer sensors (unless one adjusts only the sensors which is too much work).

However in this case there is no reason that the seesaw below the key moves a specific amount: any amount is fine. As in most digital pianos, the sensors can be in the key and the "action" is there only for giving some "right" feel. AFAIK, all the digital pianos with "fake hammers" have key sensors, not hammer sensors, and this one will be no different.

RIP Being skeptic is great in general, and typical of scientists like you are. Otherwise one would just waste all their time in conspiracy theories and creative nice-sounding pseudoscience

Not only that... You only have time to do a small number of experiments, so you need to choose wisely and be sceptic of your own experiments during your planning, otherwise you work a lot and end up with nothing to show... I believe that the most important thing I teach my grad students is to question their every move, and be the first critics of their experiments and data; then, once they are convinced that there is nothing wrong in their work, they should never let anyone question their results without the others doing the same amount of data probing that they have done already. Be sceptic, and then be proud...

Lego piano

spanishbuddha

Will it offend anyone if i were to say it looks like it qualifies as a hybrid?

Going back to this sort of stuff, here's a guy who also was serious about having a "too short of a pivot" while still wanting the action have a smaller footprint

https://youtu.be/S_vWSY5fdwQ?t=50

As you can see he also uses the parallelogram idea, however he makes the "double pivot" to move, which forces him to use a spring, which we don't like 🤣
I am still wondering why nobody designed, patented and sold something like this: it does not look any more complicated than existing folded actions, while at the same solving the major problem some of them have with playing towards the back of the key (Privia?) or with excessive size/weight (VPC1?)

RIP
The finger will exert force at a certain point of the key and the inertia of the hammer will act with the opposite force not necessarily at the same point, it will force the flexure to bend in some wrong axis and I doubt very much that it will be possible to design such flexure that will be stiff enough (comparable to a wooden key) in all axes other than the one it is supposed to move, certainly not from any plastic.

I think the concept of "virtual pivots" in DP actions is pretty interesting. Both Kawai and Casio seem to have their own patented actions designs that uses a virtual pivot, which in theory, could allow a longer pivot action in a much more compact form.

Kawai's Patented Action: https://patentimages.storage.googleapis.com/f9/70/ee/a50b5a173b43eb/US12136403.pdf

Casio's: https://patentimages.storage.googleapis.com/b9/26/d3/1f2d1c75b76544/US12315474B2.pdf

I hope Kawai or Casio is on their way to releasing a model involving their own virtual pivot actions soon.

Mind these are interesting but I see some disadvantages with the Kawai one (and the Casio one I can't really understand it, maybe the back of the key slides through some guide...?):

The only fixed fulcrum is that of the hammer. And the one for the key is a floating one, linked to the hammer, so when the hammer raises, it pushes that floating fulcrum slightly upwards, thus creating a virtual pivot point. But that means that the key is permanently linked to the hammer which is known to be a bad thing for digital piano actions, you can't just tap the key to accelerate an independent hammer. Instead, both the key and the hammer need to move together.

Now, regarding the Casio one, as I said, I can't really understand how it works, I will have to dig into it more and I'm a bit tired today for this, but what I can see for sure is that the hammer is again linked permanently to the key, so it's the same critique I have.

Other than that, it's an interesting idea and I hope we can see some progress towards virtual (longer) pivots in compact actions. But more than anything, I'd love seeing any real escapement in digital piano actions. I guess by this point, it's already clear that we don't mean simulation of the escapement bump, but rather the complex detachment of the hammer from the key leverage which also changes the leverage ratio, and the backchecks catching and stopping the hammer rebound which removes the forceful pushback regular digital piano actions exert against your fingers.

Mind

I've got the infinite-pivot CME xKey but it's not a hammer action, rather a set of space-bars.

I recall earlier versions that have never materialized. The Casio one looks too thick to be marketable.

CyberGene I'm kind of interested to know if there are DP action examples that shows how if

"the key is permanently linked to the hammer which is known to be a bad thing for digital piano actions"

as I think I may not be fully understanding your description?

What I currently understand about the Kawai virtual pivot design, is that compared to a "standard" design (e.g. their RH3), they added/attached a "second arm" connecting the standard hammer to the front of the key (the second arm itself is also connected to the main chasis I think), and the geometry of the second arm as well as where they connect all of them together dictates how much the fron the of the key dips relative to the back (which is sort of the same as pivot length).

The Casio's "cam follower" or the guide on the back probably traces out the trajectory of how the key will move as you depress it. So it's kind of like a "2d" key guide I guess?

In both designs, I don't think the key is fully rested on the hammer.

From the figures and diagrams in the Kawai one, it gave me a feeling that it's more "production-ready", and not just a concept level "in theory this might work" design.

Mind I'm kind of interested to know if there are DP action examples that shows how if

"the key is permanently linked to the hammer which is known to be a bad thing for digital piano actions"

as I think I may not be fully understanding your description?

I believe the wooden Kawai actions are those where the hammer is independent from the key. The easiest way to test this is to put your thumb against the front (vertical) wall of the key, under the lip. Then with the other hand to strike the key but use your thumb to prevent the key from bottoming out, ideally moving just a few mm. If the hammer can still press the switches and thus produce sound, then this is an independent hammer, that is not hard linked to the key. On most folded plastic action, what presses the switches is not the hammer, but rather the key, which is why the hammer and the key are linked in such a way that they move together (the hammer cannot move forward without the key moving forward), in order for the key to be able to press the switches.

CyberGene But more than anything, I'd love seeing any real escapement in digital piano actions.

I totally agree and think about the same. I believe the back check in the acoustic action is the most important. Here's a slow mo video I found a while ago comparing the key return speed between a non-checked vs checked (hard strike) hammer: https://www.bilibili.com/video/BV1Lz411v7MQ?t=1078.9 timestamp 17:59 to 18:14

I'm not sure if one could design an action without detachment or without true escapement to allow some kind of checking (storing the hammer energy to aid return speed of the key), which is why real escapement mechanisms is a must if you want faster key return like on an acoustic grand.

The only "simpler" design I could think of would be incorporating some kind of (maybe a shrunk version of) hickman action (as recently used by Mike Filip on their DIY janko project) to an action similar to Kawai's GFIII. But at this point I'm just speculating and this may turn out to be a bad idea.

Unrelated small "rant": I recently discovered from browsing (and not 100% certain) that the NV10/NV12's action doesn't even feature the full pivot length of the "extended key length" version of the MIII action which is used by SK-EX, SK-7, SK-6, GX-7, GX-6, where the pivot length is longer by about 3-5cm(?) However I'm not sure how much meaningful difference to a pianist there actually is, given that both versions have pivot lengths exceeding what most DP actions offer (~23cm+), and that there are only diminishing returns for each cm of pivot length you add above some point.

CyberGene Oh, I see what you meant. Based on the folded actions that I have on hand, the PHA-4, PHA-50, PHA-100(or hybrid grand), and the Kawai RH3, all seem to have a non-linked hammer. Interestingly the Yamaha's GT and GT-S have a linked hammer design. I'm still uncertain whether a non-linked vs linked hammer have significant advantages in certain areas, since in most scenarios of playing (at least from what I could think of so far), the key return is always pushed back by the hammer (unless you grab the depressed key and manually return it faster than the hammer can "follow up"), so whether a hammer is linked or not in a DP action, it doesn't appear to make too much of a difference (?)

Mind I'm still uncertain whether a non-linked vs linked hammer have significant advantages in certain areas, since in most scenarios of playing (at least from what I could think of so far), the key return is always pushed back by the hammer (unless you grab the depressed key and manually return it faster than the hammer can "follow up"), so whether a hammer is linked or not in a DP action, it doesn't appear to make too much of a difference (?)

Yes, you have a point. I imagine that if the key is balanced in such a way that it returns by itself without needing the hammer to push it back, it would feel a bit more like an acoustic. Another effect I’ve felt in the Kawai wooden actions that I assume is a result of this, is a certain feeling of moving mechanics under your fingers. On most other DP actions there’s a single fluid motion which is a bit fake and weird IMO, compared to an acoustic piano where you feel so many things moving. But that is a purely subjective feeling.

_sem_ I've got the infinite-pivot CME xKey but it's not a hammer action, rather a set of space-bars.

Yeah I just realized that actual space bars are indeed "infinite-pivot". Very unfortunate for simplicity that hammer actions needs weights and/or counterweights.

CyberGene

CyberGene Another effect I’ve felt in the Kawai wooden actions that I assume is a result of this, is a certain feeling of moving mechanics under your fingers. On most other DP actions there’s a single fluid motion which is a bit fake and weird IMO, compared to an acoustic piano where you feel so much things moving. But that is a purely subjective feeling.

I do see what you're talking about. I do agree that DP actions do seem fake in terms of their motion.

On fast repeated notes in an acoustic grand action (single note action model), I do indeed feel more of this "certain feeling of moving mechanics under your fingers" like you described.

Mind I think the concept of "virtual pivots" in DP actions is pretty interesting.

Yes it is, so much that it had its own discussion on this forum where I encourage you to re-post your message

#175

As far as I understand these patents, I think they are doing something similar to what I demonstrated with LEGO in post 23 of that thread perhaps with the subsequent improvements that @dore_m and I discussed privately to make the virtual pivot length anything between infinite and the actual pivot. Not sure if anything that we did classified as prior art and would invalidate the patent, but it would be quite interesting to see the concept implemented commercially.

@CyberGene should we remove this section of the discussion out of this thread and moving either to that thread or to a new one? I don't want to derail the discussion too much on this (already huge and unmanageable) one

CyberGene stopping the hammer rebound which removes the forceful pushback regular digital piano actions exert against your fingers.

That's the main improvement I'd like to see in digital piano actions.

Way back when I bought this Clavinova I mostly left my Kawai upright acoustic untouched.
Months later when I went back to the upright for a bit I was reminded of this shortcoming in the digital action.
The difference was startling, although it should not have been.

I doubt there will be a proper escapement in any digital piano that's not a hybrid.
The regular digitals are just too small to fit a complex action.
And I think it would be cost prohibitive.
And the complexity and wear would result in more frequent and more expensive maintenance ... just the sort of thing that we don't want.