Thank You Duke, giving this support to a project with a very exciting perspective.
As a complete enthusiast layman I cannot imagine, how big it would be as a project, but I am preparing ourselves to start with it (unfortunately, not before about 2-3 weeks time, my son has to finish semester and after then he must undergo a smaller shoulder operation before we can start).
I certainly wish your son the very best with his shoulder operation!
I don't think this will be a terribly difficult project. You'll need to make two boxes, each roughly 400 mm tall by 270 mm wide by 180 mm deep. Plywood would be great. Each box will have one speaker cut-out on the front, and one on the back. You'll also be installing input terminals (perhaps in a "cup"), and a small port. Fancy joints not needed - butt joints are fine. You'll also be soldering some internal wiring and/or crimping on some connectors.
But the result should very exciting either way: if the outcome shows this way we can't achieve a significant sound quality boost, I would regard this as a relevant information about state of the art DP technology - just more so if it should turn out, that there is a large potential for commercial DP development....
Well, I certainly don't have plans to do this speaker commercially. The most expensive part of manufacturing a speaker is assembling the enclosure, and in order to be competitive, a low-cost speaker enclosure almost has to come from Asia by the container-load. I am in no position to invest that much money into such a venture... but on the other hand the chance to build it yourself and save money may make it attractive to digital piano players.
I promised you some theory. Briefly, we are going to try to approximate the sort of sound field an acoustic instrument (in this case a piano) creates in a room. The key ingredient of that sound field is a well-energized, spectrally correct, fairly diffuse reverberant field. Most small speakers do a decent job with the first-arrival sound, but a lousy job with the reverberant field, whereas acoustic instruments do just as good a job with the reverberant field as with the first-arrival sound. By using a second fullrange driver to send a fullrange signal off the back of the cab, dedicated entirely to the reverberant field, we hope to recreate a bit more of the "feel" of a live instrument.
Here's an article I wrote for an online magazine about the full-scale incarnation of this concept. The intended audience is the high-end home audio world, but most of the principles apply to amplified acoustic piano:http://www.hifizine.com/2010/06/the-controlled-pattern-offset-bipole-loudspeaker/
My first commercial product embodying these principles received a "Golden Ear Award" from The Absolute Sound magazine, one of the top two high-end audio print magazines in the US.
I'll open also a new thread for this.
Sounds great. When you are ready to begin, shoot me an e-mail, firstname.lastname@example.org . I don't always check in here regularly, and I don't want you to start without me!
Also if you have any questions or concerns before you get started, feel free to contact me there as well, or here in this forum if you prefer.
I'd like to outline some of the pro's and con's of small fullrange drivers, in hopes of giving you plenty of relevant information before you start spending money.
To recap, here's the driver I recommend:http://www.parts-express.com/pedocs/specs/264-914-tang-band-w4-1320sif-specifications.pdf
The main advantages of fullrange drivers are:
1. No crossover. Crossovers are usually the most technically challenging part of a speaker design, and no crossover means great coherence comes automatically.
2. Relatively high bang-for-the-buck in the lower price ranges.
3. As long as they're not pushed too hard, fullrange drivers tend to have very nice midrange.
The main disadvantages are:
1. The conflicting requirements for covering the top and bottom end of the spectrum mean that compromises must be made, and they are usually most apparent at either end of the spectrum (the on-axis curve of this speaker looks good at the top end, but it will be beaming badly in that top octave).
2. Loss of clarity at high sound pressure levels. When the cone has to move far to reproduce bass signals, that degrades the clarity of mid and high frequency signals.