Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Kawai CA95 keyboard action vs. others

Generally I am not keen about reviving an old thread, but, I went shopping for a new DP recently and had the time to play several CA95 and CA65 in several piano shops. Before touching them, I wasn't able to really understand this thread, but playing the CA95/65 as well as playing with the Kawai key action model displayed in the store (consisting of 2 white and a black key together with all the action components), and comparing them to other DPs and APs nearby, I think I understood what this is about. So, I hope to be able to provide a bit more clarification to future readers of this thread:

(1) The "keys seemed to take forever to rebound after playing a note"-effect mentioned by the OP occured on all CA95/65 I played as well as with the key action model. I believe that it's not just a production quality issue affecting only a few units, it probably has been designed to be this way: Once you release a key, it doesn't immediately snap back to the original position, but wobbles around for a short time. I have not seen this effect on acoustic pianos I have played recently, and not on most other DP's as well. My old Kawai MP9000 doesn't do this either.

(2) This does not mean that you can't play quick repetitions on the CA95/65. Just hit the still wobbling keys and they will sound as they should.

(3) But, the wobbling is real and visible and can be felt when playing the keys in quick repetitions, as well as felt as slight vibrations when playing keys close to still wobbling keys. Some might not care, while some might find this annoying. (I did.)

(4) The reason for this wobbling: When observing the key action model, it seemed that the hammer is "hitting back" on the key, causing the key to wobble, and the whole system to require a short time until coming to a complete halt. When observing Yamaha NU-1's upright piano action in the transparent demo model in the piano shop I went to, it seemed that it's "real" upright action is designed to make this impossible. I assume it is similar for Grand Piano actions, as well as the old Kawai DP action in my MP9000.

(5) So, this is neither about the length of the keys, nor about playing technique. (Well, okay, at least not to me, or, not from an engineering viewpoint.)

I suppose that for whatever reasons the Kawai engineers of the current actions (sadly) either don't care about this (I assume they have several metrics on which new designs are evaluated, and maybe this just isn't in their decision criteria?) or had other priorities when designing the action of the CA95/65.

So I guess future buyers just have to try for themselves to see if they notice this and find it annoying or not; I hope I have been able to clarify a bit what the op was about (unless I have completely misunderstood it :-)) and what to look for.

Thanks takura, we love discussing key actions on the forum.

What about Rolands muffled sound compared to Kawai

I have also observed this previously: there is a thin felt on the key block to cushion hammer impact while falling back. And also speculated about the designing decisions behind. My thought to this, that as there is no fixed join between the Kawai's hammer part (unlike most other DP keyboards), this arrangement could help to gain hammer momentum (the rest of it after bouncing back on a flexible foam ) to accelerate the key block itself back to its initial resting position, while bringing the hammer to a smooth stop at the same time.

Perhaps this side effect you are describing was assumed by R&D either as a negligible one or even as desirable as it could be felt as an added vibration feedback to the finger, perceived perhaps as another added realism for a less experienced buyer?

[EDIT 2013.05.12] Key rebound emulation seems to be more a feature, s. following posts.

Stabilizing key return is perhaps one more target for future improvements GF actions (beside more pronounced escapement simulation and playing out of the jack feature.)

This same design is to be found on both RM3, RM3-II (VPC1, CA15) and GF (CA95, CA65, CS10). One question is how far is it discernable on the VPC1, e.g?

While I think this effect could have been present with RM3(II) as well it might have been mitigated there by the fact the key part behind fulcrum was longer with RM3 (relatively to hammer length and to the main key lever part, but even as measured length). The distance between fulcrum and the felt cushions is short.

GF:

RM3:


Would be interesting to see a real-life comparision GF/RM from one who definitely can identify this wobbling behavior.

the wobbling you can se also hear about 4:18 and few seconds later

http://www.youtube.com/watch?v=0KZknBsQ0ZQ

I wonder what the reason behind this thorough dissection really is... Just out of curiosity, I have looked up a video of playing on a Steinway grand piano and I believe the "wobbliness" is present there as well. You can see and judge for yourselves.

As has been stated before, there are 3 sensors for each key with the express purpose to allow for easier fast repetitions/legatos. Whether this particular type of key action works for a given player is at their own discretion but the GF action has clearly been designed to mimic the action of a GP and if the video above is any indication, this particular trait is not limited to just the GF action.

I am the happy owner of the CA-65, and yesterday I had the opportunity to play several different Kawai grands including the SK-2. The "wobbliness" is very apparent on all of them as well as on the Steinways I played. So I guess it is a feature more than a fault. Actually I'm amazed how close the CA-65/95 action is on the real grand action.

Forget all this wobbliness talk. It's perfectly normal on a real piano. A key has mass and when struck, it has momentum. That momentum does not just disappear at the end of a stroke - it must be brought to rest via damping forces coming from the felts. There will always be a brief oscillation in the key while that energy is absorbed and the key is brought to rest.

Wobbling is natural, however I think the question is the level of wobbling. And yes, each grand has it.

Never thought about this before, but I've just checked on my V-Piano - yes, when playing staccato, the keys do wobble when they come up on sudden release.

I had to conclude, that apparently no real CA95/65 owners have complained about wobbling keys yet. (Not even PV88 :-), or some people who traded in their CA95, some only to regret it later).

Reminds me a little on the birth of my first son. I was present at child delivery, and the first thing I could realise was some irregularities, deformities how his toes are grown, which I found a minor but lifelong affliction. The delivery nurse didn't seem to bother much - and my son is meanwhile 17 Years old and has no defects at all - at least with his toes not.

The "wobbliness" occurs naturally in acoustic pianos so it should be no surprise that this is emulated in digital piano actions. I have just briefly tested the V-Piano, Kawai CA95, Clavinova CLP-990M, and, all three of these digitals (that I own) display the trait.

Looks like the V-Piano's keys wobble the least on fast repetitions although they still wobble just the same. This wobble appears to be an up and down bouncing motion of the key and not a lateral one from side to side. I would call this effect a "bouncing" key and not a "wobbling" key.

Also, this does not affect repetition speed on the CA95:

https://www.box.com/s/m2yf56x2926dyprrpl7s

I think synth(organ) keyboards don't wobble, and some of the older weighted keyboards(like my old yamaha dgx-620) have less wobble. Personally I find my CA95's touch is just right. It is not a real grand, but the feel is close enough.

I see, I could have been more careful in choosing words. :-)

(1) PV88> I would call this effect a "bouncing" key and not a "wobbling" key.

Yes, my apologies, "bouncing" seems a better word than "wobbling" to me, too. (English is not my first language.)

(2) I completely agree that all pianos will wobble/bounce due to physics, and, as kapelli has written, it's a question of the amount. Some (including me) see the CA95/65 to wobble/bounce more than "other" APs/DPs, some don't. (Do not agree that therefore this is meaningless to discuss.)

Engineering-wise, I guess this could be measured as
- time until it stops (i.e. becomes not realizable)
- number of swings until stop
- frequency of swings (e.g. in Hz)
- size of wobble/bounce in vertical direction (i.e. measured in mm).

A comparison close-up video of keys in different pianos could clarify, but it will be necessary to hit the keys similarly, and that will be difficult. Just putting your finger off a completely depressed key won't probably be sufficient as it seems that the hammer is "hitting back" in (at least) the CA95/65 (not sure about others). I don't have a piano playing robot available to me (I know of university labs that have), and I don't think the piano shops I have access to will let me drop (calibrated) weights onto keys while recording that with a video camera. :-)

(3) Concerning the possibility of the hammer hitting back on the key and causing some more wobbling/bouncing in the CA95/65 than in other key actions: my thanks to Temperament for discussing further and putting up the pics. (The CA15 with similar action to VPC-1 does wobble/bounce more than I'd like it to, so that's currently the only concern I have about upgrading to VPC-1.)

Although it's still just an assumption and not proven that it really does so in the CA95/65, if someone knowledgable could confirm or negate whether hammers can hit back onto keys in real piano actions, I'd gladly appreciate to learn.

Completely agree ... if I bought a CA95 I myself might forget about this and be satisfied with it ..... or not. Anyone here who recognized this before buying and found it annoying, but then got used to it? :-)

Another analogy that comes to my mind: I believe most Toyota-owners are satisfied not only with reliability but also with the haptics and feedback to driving operations (steering, braking etc), but that doesn't mean that all BMW-drivers bought their cars only for the badge prestige and that there is no perceivable difference in driving responses. Measuring the difference and judging whether it is large enough to be perceivable is engineering, deciding whether or not the difference is significant to you everyone has to do for themselves.

So, a real measurement instead of subjective opinions (including mine!) would be nice to have. Just noticed that I could borrow the high speed cam in my lab to take comparison videos; if I record enough samples I might be able to just pick those with similar velocity for each piano, without having to control the velocity of pressing the keys using weights or robots. Will take me about a month to complete though.

You can see a piano action in action here , also check YT as there are some videos there.

Look at this video, after 3:00 there is good depicting of key rebound behaviour



As it seems key rebound emulation is rather a feature than a bug. It can even help with repetitions, because the experienced player can make use of the little rebound energy when starting the next striking the key.

The importance of this initial position rebound will further put in perspective by using 3 sensors simply by the fact, that during fast repetitions there isn't necessary to return to starting key position. (As a matter of fact nor is it with 2 sensor systems either, only the starting point to start a repeated note is higher.) Its importance is only with true staccato repetitions.

What is more interesting, how much rebound force is present and will be felt during fast repetitions. Here are completely other mechanics emulating the functionality of the very complex Renner piano mechanics.
In original piano action there is mainly the felted back check is absorbing the energy of the falling hammer - a complex combination of friction and spring bending processes.

(But the emulating mechanics of a simplified DP action are complex enough with many parameters: foam elasticity which has a dampening effect on hammer impact, hammer geometry and weighting, felt placement and thickness, etc...).

There can seemingly be nothing more simple than the action of a traditional clavichord (no hammers at all) but even with the design of my clavichords were a lot of tricky decisions involved. Instrument makers did guaranteed no scientific measurements and calculations some 3 centuries ago when this keyboard instrument class was established. But they invented pivot point offset of black keys or weight compensation and variable lever geometry, and a lot of other things, which will be marketed now as the newest breathtaking innovation of the newest DP-generation! (Well, Kawai and the Yamaha hybrids and their meanwhile long ago discontinued CLP990 have already applied pivot point displacement, the others will follow next, to be sure.)

What really counts is how good, realistic, playable the DP actions feel. However some of us in this forum have just a special interest to know how clever and detailed their technical implementation is, what features were actively taken in consideration and how much optimized in the action design process.

To me it sometimes seems more like chasing ghosts and tirelessly looking for flaws that actually are not there.

^ This.

For my part quiet not. It is for clarifying questions about technology, both acoustic and its the digital emulation, seeking for a in-depth understanding of principles, the level of perfection in technologies. Asking questions and getting as I think thorough and unbiased answers to these or at least balanced opinions about the topic is useful for the general understanding and contributes perhaps to the development of technology.

One possible problem is that we are a completely mixed public,
some people seeking for simple buying advice, and they may be not so optimal served by too much scrutiny here, they would need more certainty for their decisions.

But the time is near when I'll retire from this forum, so You can be relieved - at least of my hair-splitting in some posts.
I could learn a lot by both reading and writing, but the learning curve is just beginning to get a little bit flat at last.

Well, I'm all for research and understanding of the underlying designs of the various key actions out there but I get the feeling that more often than not our lack of deeper knowledge gets in the way and we sort of start to agree on that there must be something wrong with this or that aspect of a key action's behavior while the truth is that it's perfectly fine and that it's only our limited insight that has mislead us.

I do agree that this may prove harmful to people who maybe don't understand or care about all those details but who get a false impression that the key action of a piano is bad even though it actually is not.

Oh, and just for the record, my post was not directed specifically at you, Temperament. It was just something I had noticed in the past several months of my membership here on this great forum.