Hammer time...

What you say makes sense to me.

The difficult part may be getting the right elasticity (or recoil of the hammer from the string) across all 88 keys.  I suspect that it may vary considerably.  I haven't heard of any research that investigated how much a hammer bounced from a string.  Seems to me this might be critical.

The advantage a grand piano has is that gravity pulls the hammer away from the string as soon as the hammer has hit the string and recoils away, likely preventing any repetitive bounces.  And gravity never seems to fail (as do the springs in an upright).  A ball bounces repetitively on a hard surface because gravity keeps pulling it back down.

Glenn

I've thought of this before. But piezo transmits the vibrations of what it's attached to. If it's on a piano soundboard, then it'll transmit those vibrations. For example: http://barcusberry.com/product.cfm?ProductID=9

So if you were to use something like piezo-equipped rubber stops in place of strings, I'd wager it'd sound like rubber. LOL

Another idea I had was to use small piezo-equipped springs in replace of strings. But I'm not sure that'd be any more useful, practical, or realistic.

No... you are talking about piezo pickups as used to transduce audio vibrations directly into an electrical audio signal.  I'm talking about piezo triggers that are connected to a MIDI system similar to MIDI percussion triggers.  In this case, only the "volume" of the attack/strike is used and that is converted into a MIDI velocity range accordingly.  The piezos and their rubber mounts can easily be isolated from each other as well, so that an adjacent hammer has no effect.  The piezos are very inexpensive and the impact surfaces would be easy to tweak as well as, of course, all the MIDI translation parameters.  But mechanically, there is much less to go wrong than with an electronic keyboard with all those delicate rubber contacts and exposed circuits.  I just don't know if there is a product all ready out there that could be retrofitted... some 88 key controller that could have it's key strikes replaced by a piezo voltage spike.  I'd tackle a prototype myself, but first I've got to finish my MIDI pinwheel.

This formal article argues that touch does have an effect, since the strings aren't the only source of the sound--the touch doesn't affect the hammer hitting the strings, but instead affects the sound of the strike on the keybed (which gets transmitted to the rim and harp) and the resonance of the keys themselves.

(You have to download the pdf version to see the diagrams.)

http://74.125.47.132/search?q=cache:MBl...&gl=us

(So now I want to be able to control the key resonance and the extent to which the keybed transmits freqs to the rim and harp...Apparently "the stiffness of the felt washers
under the keys can influence the attack sound noticeably.")

Glenn,

I edited my post several times, and you may have replied before I added some information--that the article argues that vibrations from the keybed get transmitted in varying degrees to the rim and harp (depending on the octave and the stiffness of the felt beneath the keys), so that there is a low but audible "tonal" aspect to the non-string sound. (Some people might also argue that even if the strike noise was just noise, that noise is part of the overall sound, and is thus part of the tone...And drums are often tuned.)

Sorry for all of the edits--I posted too soon, before I'd really absorbed all of the information in the article. There's a lot to absorb, and I'm not at all sure I understand the article fullly. Just stumbled across this article while trying to learn more about transients, and thought it would be of interest in this debate.

Edit: I worry about the article's discussion of bridge vibrations before the strike. Wouldn't the tension of the strings, when they were in place, limit or prevent any bridge movement before the strike?

Jake:

Is there any way of getting the missing figures in the paper (six in all)?

They seem to be quite important in fully understanding what's going on.

Glenn

Sounds like we are talking about some form of acoustic aftertouch.  I think I saw Bigfoot and Nessie this morning, too....  No, really, I would think the best way to prove if there is an audible effect would be simply to record, on your own piano, a single note, with and without post finger rolling/wiggling/pressing/hocus-pocusing, and then compare the results.  Now it would be cool if piano manufacturers could implement an intended aftertouch effect -such as some form of tremelo- and have this be a purely acoustic/mechanical effect, but I'm not sure how they could do it without jeopardizing the integrity of the instrument.  A giant wang bar assembly on the soundboard might put off the purists.  Of course we can do all sorts of things with our MIDI controller, either with aftertouch or additional continuous control pedals, ribbon strips and infra red sensors.  But back to the personally customized, easily repaired, percussion-based MIDI keyboard controller... I think it's doable for under a grand.  (..or at least under the price of a grand.....)