Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Something nearly ALL digital pianos get wrong.

I confirm the above holds for Kawai CA63 as well. On one hand it's surprising that one mentioned this 'issue' as far back as 2008 and yet the problem still occurs [while some DPs didn't have this problem in the first place], but on the other hand this is what you get if you don't use physical modeling - some acoustic behaviors do not derive automatically and have to be programmed in. This is for me a non-issue, I don't expect sampled DP to be 'perfect'.

Lol. Classic type of assumption people make who have no experience modeling things on computers. One doesn't just describe the piano to the computer and automatically get all the emergent subtle properties of an acoustic piano. Inasmuch as modeled pianos more accurately mimic this type of behavior, I'm sure it was carefully thought out and purposely added. Modeling is a painstaking procedure that produces gibberish the first zillion times you try it and you have to go back and tweak it ad nauseum, carefully adding any important features until it kind-of sort-of mimics the desired behavior.

The sampled pianos (hardware and software) may come up short, but it's just because they have lazy or understaffed programmers who haven't put the work into getting ostensibly unimportant details like this right, not because the pure modeled approach is a panacea for all piano behavior or "automatically" gets piano behavior right.

Pardon the apparent huff. It's a bit of a peeve of mine when people ascribe sci-fi abilities to computers and programs that exist in the real world, so I always speak up when I can.

One thing I notice in the acoustic piano: from the note F5 onwards the notes are sustained (non-dampened), which means you don't have to step on the pedal to sustain the notes.

Only a few DP's I tested have this feature, like the Kawai CN24, KCP80, KDP80. None of the Yamahas I tried - YDP 141, 161, 181, P105, CLP440 - have this feature. Neither does the Casio PX-735.

Let's have your input on various other models, too.

That's odd. I've seen the opposite in digitals: high notes not damped.

I'm surprised too as I've not yet come across one that doesn't. This includes various Casio's, Yamaha's, Roland's and Kawai's. We get quite a few newbie posts on here complaining about this 'defect'.

Haha! My dissertation is about piano modeling, so yeah... But anyway, what I meant was the fact than once you create the model, the interactions between various elements of the model become more natural. Like you model a vibrating string, then you can easily add 2 or 3 strings per note (and thus, say, simulate the 'soft' pedal). Or if you model the damping system, then it doesn't matter if it's key off or partial pedaling, it's the same system underneath. With the sampled piano, you can't have a pianoteq-like option with "All keys on the piano have 3 strings", or you need special samples for the soft-pedal [or anyway, do a 'cheap' version like take out the high frequency of the samples, etc].
Creating the model, that's a different story and of course you have to think where you want to cut corners and where you want the model to be as close to reality as possible.

I'm sorry, but that's just plain rude! If a piano graduate doesn't play Rach3, then that doesn't make him lazy. And I'm sure that in no matter what profession you are, laziness is not tolerated by the management!
I don't know if with sampling you can do all the things you can with modelling, but what I'm sure of is that for these 'extra-special-cases' it takes a lot more effort to do it in the case of sampling (and thus this could be the reason why money is not invested into this, not the 'lazy programmers').

It would not be practical to try to do with sampling everything that can be done with modeling. You can model things like hammer wear, damper wear, infinite lid positions, infinite velocities (or at least 127 with standard MIDI), etc. Whether these things are all necessary/desirable, or make up for other areas in which modeling currently may fall short, is a different discussion.

As for the phenomenon being discussed here, regardless of whether the tone generation were based on sampling or modeling, I think it comes down to a programming issue, of how to interpret the interplay between key up, key down, pedal up, and pedal down.

Meh. From the perspective of the consumer, lazy programmers, incompetence at any level, and underfunding (justified or not) are observationally equivalent. I don't really care which is the case (and I doubt that you know). I'm not inclined to give them the benefit of the doubt considering the large numbers of years that pass with little progress in DP technology and the fact that some pianos have the desired features while other do not...for years and years. Attributing it to a technical difficulty of sampled technology seems odd, considering the fact that some sampled pianos do behave correctly. This is one of many behaviors that gets discussed here that some or most digitals do not do correctly and doesn't seem to get fixed.

Look, I'm sympathetic with people whose job is hard and/or unappreciated, but my experience with digital piano technology leads me to expect better than what we observe. That's the general sense of this whole thread, by the way (ancient, though it may be).

@bogs, would be really interested in your diss. Could you send me by PM ? What are the models you use for the string, etc ?? I am doing a lot of structural dynamic but in a different business.

ap55

My V-Piano does this perfectly, though of course it's modeling technology. A sterner test would be to silently hold down different notes (i.e. not the ones you've just struck and then released) while keeping the pedal down, then releasing the pedal to see if the sympathetic vibration from the sounding strings have transmitted to the ones you 'didn't' play, but are now undamped (because you're holding them down). It works on the V-Piano, just as on a real piano, but I suspect you won't get this on sampled DPs.

Incidentally, this latter device of sympathetic string vibration causing an eerily soft sound from undamped strings is exploited by Schumann in Carnaval, Op.9 (the 7th chord at the end of 'Paganini', to lead back into the reprise of Valse allemande).

My new Roland HP-505 passed the test.
And it has the sympathic (here called "string-") resonance feature.

OMG! Im so disappointed now! My KAWAI CS6 failed the test Yamaha P155 passed it though!!!

The CA95 fails the test (above) when using both pedals:

The sustain, or, the middle one.

@jscomposer,

I am now taking this issue directly to Kawai (as you have done) to see as to what they have to say about the sustain pedal resonance issue.

Not likely that much will be done, unless there is a way the pedal and resonance behaviors can be fixed within a software update.

If not fixable, current owners may have no recourse.

Will have to wait until a new model arrives.

It actually works on the RD-700NX, although only for the Supernatural AP and EP tones.

It also works when I use it to control Galaxy Vintage D. Cool.

I actually do use this feature, mostly when I forgot that I need the bass note to sustain after I release the pedal

CA51 with Galaxy Vintage D and Vienna Grand (Bösendorfer) have it, but itself not.

For me a typical case for a minor SW-Update (CAx5).

Waiting for a model change would be a non-solution, being the CAx5 a brand-new model (2-3 Ys).

Perhaps something to consider for .v2 model refresh, but before You can get it, it is probably another 6 months when not underway already.

Manufacturers must have it difficult nowadays with all of the pressure from us forum-folks with our public scrutiny...

Wouldn't be the physical modeling really the solution. What can you expect from companies that are not really eager to can get proud enough to put physical modeling on the market, but pay for some acoustic technician to do sampling measurements. Thats the way if you was not willing to pay engineers and programmers to do excellent work. So all the behavior we expect could be modeled and if there is an open interface for programming there will be many open source developer to contribute to the perfect piano. There will be supid business, but no passion in DP development. Good to have Roland v-piano and soon hopefully viscount physis piano

This post looks like it makes sense when you read it fast, but on careful parsing, I realize I have no idea what you are saying.

@gvfarns: If there is a physical modeling, the piano will be well defined in a software-code. If this code is open, so an evolution of this code in the backward direction will likely be avoided, because what happens is transparent. The same situation appears if there is competition in a traceable situation (software modeling is usually traceable, if the code is available, or at least for the company who did it). For sampled based piano there is a lot of competition, but no traceability what really happens, because the evolution of sampled based piano is only on adding with some workaround new features.

But may be you are requesting with your question for a conclusion. In this respect you are right concerning my missing question mark for the first sentence: "Wouldn't be the physical modeling really the solution" ??? So my post was more a question, which can be answered. The actually conclusion for me I tried in the last sentence, which should not focus on a special brand, but in fact that I think it is good to have the first physical modeling pianos on the market and the horizon with above respect. If this market will grow hopefully in future we will not need to discuss about some shortcomings of different features for sampled based pianos.

Or you like to see it that way, that my contribution is therewith closed because self-containing. You like you might agree or post a counterrevolutionary opinion. On one point however you are right, it does not contribute to an answer to former question and complaints, but put them in a more global light.

@ap55 - Are you mesmerized by Physis, or just an astroturfing marketer?