Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Why don't manufacturers record "bad" piano sounds?

Yes, but there is subjective evidence which some pianists find quite compelling.


That's an approximation. Hammer shanks vibrate and flex in all directions, especially (but not only) in f and louder playing. High speed cameras demonstrate this without any doubt. Those flex and vibrations depend on how the key has been pushed before the free flight began: steady/constant acceleration, increasing acceleration (strongly recommended by some teaching approaches, so much so that Steinway patented a part of their action with the name "accelerated action"), decreasing acceleration, jerky push and more. It's possible to achieve the same final velocity with quite different "histories" (to use your terminology) which result in different shank flex and vibration when the hammer hits the string. Heck, even slightly different location of the hit point, which in left-right direction is important everywhere and in front-back (grand) or up-down (upright) direction is extremely important in the treble. Advanced technicians work to give control to the pianist about that, including installing shanks that are much more rigid and do not vibrate or flex as much as wooden ones (while keeping the same weight, which is extremely important). Some pianists love that, some others hate it (find it "too harsh"), and yet others consider it irrelevant. After extensive use, hammers with these shanks have "razor-thin" edges in the string markings, demonstrating perfect hammer-string alignments in all blows. Hammers with regular wooden shanks have very "broad" edges, indicating that the hammers do not always hit the string in the same point, causing a difference in timbre between those blows, even if they have the same loudness. In fact, the una corda pedal does just that in a more predictable way. I am not sure if any pianist (let alone a beginner) can consciously or unconsciously take advantage of these effects, but based on the preferences mentioned above I suspect some can. In any case, claiming that these things are non-existent is incorrect.

The truth is, the hammer's history does matter and the strings "know" (to use your terminology again). I am sure no digital and no sampled piano is able to simulate that effect (besides perhaps pianoteq). However, we don't really know whether or not that is important from a musical expressiveness perspective, even though I suspect it is, at least for some pianists.


From the last statement sounds like you haven't played an acoustic piano much, if at all.

Most of the people writing didn't understood the OP question, so here I am.

Yes, digitals always sound good. Bad sound comes from bad technique and untrained hand - so lack of control over grand piano mechanism. There's a reason why the most important thing in Russian piano school is tone creation- the thing that OP asks for, not badly prepared instrument.

Yes if one is not playing regularly on high quality grands, than playing on it may be something between fascination and sadness (first for how it sounds and what is possible, second for how I sound). In the DP sound is already created in memory and actions and far far easier to control than on a quality grand.

There are many things on the road which makes dp always sound good: firstly it's recorded good, secondly iss sound creation system is a toddlers toy compared to complicated mechanism of a grand piano, where each part makes a difference in projection.

But mostly I would call its due to lack of control over the mechanism of a grand piano. Digital (even the good ones) are far easier in this respect.

I stand corrected -- thanks! As you say, "musical expressiveness" may, or may not, depend on such effects.


Guilty, as charged, for the last 50 years. In my youth, I learned on a baby grand through high school.

Now you have me wondering if the vibrations of the hammer shaft (stronger with a percussive touch?) really do make an audible (to whom?) difference in the sound . . .

This discussion reminds me of the claims made by some audiophiles. They say that they can hear differences between cables, that you need an specific DAC to synergize with your amp, and a lot of other claims that have no support on evidence whatsoever. At the end of the day these discussions are pointless and ignore the fact that human hearing is an incredibly subjective experience. I'm sure a lot of people hear differences in tone quality in the same piano depending on how they play a key. I'm also sure these differences haven't been demonstrated and there is zero evidence to support them. I can't help myself to not believe the evidence more than the subjective experience on people's heads.

I would say such "bad-sound-when-play-bad" is too much of a simulation. After all, this is already the domain where digital is considered on its limit and one should look for a transition into acoustic when such "fine" skill is to be scrutinized.
That's the reason why acoustic is a must when crossing a certain level.
From that point on, digital only helps to practice but real end performance must be done on an acoustic IMHO.

On a somewhat related note, I recall there are DPs (and VSTs) that have velocity layers that are very difficult/impossible to reach with default touch curve settings. These tend to be at the ffff end of the spectrum and don't always sound particularly musical. You sometimes have to manually adjust the touch curve or set the DP to "light" touch to access them.


Your post was the best explanation I've heard so far as to how velocity isn't the only key factor in the quality of the tone. Presumably the hybrids with hammer sensors have this effect "built-in." And I think Yamaha actually includes "key acceleration" since the CLP-6xx series as a factor in producing the sound. At least it's something described in their MIDI implementation charts:
https://usa.yamaha.com/files/download/other_assets/1/1343831/clp725_en_mr_a0.pdf

I haven't seen anything that describes how a note changes given the same velocity but different acceleration.

Whats your experience in audiophile or studio gear and piano playing?

Acceleration measure is new : CLP7xx, not CLP6xx and not hybrids. The N1X measure the hammer hit to be sure the note should be played. The NU1X hasn’t this sensor and may produce loud notes because of this lack.

Perhaps this video could give an idea of the control of the piano sound.

Even if we have DP with full acoustic-like dynamic range and timbre changes, another interesting question then is at wich volume level it all is fully presented? At 100%, I suppose. But most of the users, according to my experience, play their DP at 70-80% volume. And I wonder how all this complicated acoustic-like behaviour is gonna be "squeezed" if you play your DP at 20% volume or even less.

I guess the so called "history" of the key stroke doesn't affect the hammer, but it affects the player. (Which then affects the hammer.)

And I guess now I need to learn to play "potato". 😉

That is not a valid argument... If I had the interest or the possibility of playing my DP at the loudness level of an acoustic, I would have bought an acoustic. It's even cheaper in the long run.

Assuming what Del Vento says is accurate, the hybrids with hammer sensors shouldn't need key acceleration sensing to do the same thing, since they are monitoring the actual behavior that the key acceleration is trying to model.

Of course there are details and nuances, e.g. where on the hammer shank the shutter sits, etc. And does key acceleration matter as much (at all?) for an acoustic with a WNG composite action and fiber hammer shanks with less flex?

Thank you, I'm glad you liked my explanation. Thanks for sharing that MIDI implementation chart, that is really interesting to see. How you say, I'd be interesting to see how this data is used, because it certainly is (since the MIDI implementation chart says that it is accepted as input -- hence not ignored). Perhaps somebody with one of those instruments can give it a try and report back?

Do you know if pianoteq uses that data? FWIW this is something that I could add to my Cybrid design



Yes, the AvantGrands do not appear to have this acceleration data in their firmwares, but I don't see why they could not have it. Even the NU1 measures the instant position of the key (which it then uses as an "aftertouch" information for damper release, like Disklavier) and could use that to infer acceleration. So I believe the "controller" part can be easily added with a firmware update. Then of course the sound engine part might be too much of a change for their MCUs or MPUs to provide a sound difference, but still.

That video is excellent (even though each one of those touches would deserve its own 10 minutes explanation), but I am not 100% sure that what he demonstrates is not equally (or at least "close-enough"-ly) applicable to a regular good digital.

Yes, the N1X measures the key position and has in theory the data to compute the acceleration. But from the MIDI stream, only the release phase send positions, then the only used data is what is needed to have half damped keys if we release slowly.

There is a dispute between pianists and piano teachers about how and if the way you touch and press the piano key affects the tone and timbre in any way beyond mere intensity.

Those who advocate that touch only affects intensity say that the impression of different tone quality comes primarily from different voicings (that is, the way you balance the intensity of each note in a musical texture to make it sound with certain quality), and maybe from other subjective impressions, coming for example from visual input, and even an placebo effect.

And there are those who advocate that touch affects tone quality beyond intensity. There are some evidences I've read in some books. For example, as Del Vento already said, hammer shanks vibrate and bend, slightly changing the point and angle of attack of the hammer. If you play more "superficially", without pressing the key fully down, this can cause a different response of the hammer than if you press the key fully down, even if the resulting "intensity" is the same. Also, if you touch a key percussively, the vibrations of the key could propagate through the action to the strings, adding some "percusiveness" to the sound. And there are maybe other variables.

But, how much this really affects sound, I don't know. I've played to teachers who firmly believed in the power of the touch to change tone quality; those pianists believe a digital piano (at least one which only captures "velocity") will never attain the full range of tone a [good] piano is capable of. But I've had at least one teacher (who is an incredible pianist) who said tone quality is simply a function of voicing. The fact is, advanced pianists spend a lot of time working on touch quality and variety. There is one kind of touch suitable for Brahms, generally not suitable for Debussy. Etc.

Edit: there even are pianists who believe playing staccato and legato, both with pedal down, sound really different from each other!

...as demonstrated on the embedded video a few messages back.

But maybe he just played more loudly when doing the staccato(?)

(Without even realizing it.)