Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards A real harpsichord keyboard with midi: advice needed!

Hello all,

I am thinking of making myself a "digital harpsichord" that uses a real harpsichord keyboard.

As for the midi sensor system, I know that the PNOscan system might be good, but although I wrote to them twice asking specific questions (such as whether the sensors could be spaced a bit more narrowly to accommodate a slightly shortened octave span on a harpsichord keyboard), they didn't reply.

My questions are these:

1) Does anyone know of an alternative midi sensor system that would be good for such a project?

2) Do you know of anyone with midi sensor knowledge, who might wish to collaborate on this project?

If I can manage to make a decent, elegant digital instrument, I am thinking more could be made, to sell to others who, like me, prefer a digital instrument with a light, weighted touch, a keyboard range sufficient to play all classical music up to Beethoven, and a real wooden keyboard.

Any advice/suggestions are welcome!

Have a nice day....

David

P.S. (Please don't suggest I get a Roland C-30: I had one, and returned it, since it was not suitable in a number of respects.)

Are you trying to make a digital instrument with a harpsichord touch? If so, mechanical switches could probably be used to simulate the touch better than optical sensors.

If you just want a narrow keyboard, that can be done by having it spread at the far end like a piano keyboard.

David, that's a great idea!

I'd like to see a digital fortepiano as well.

Harpsichord keyboards are not weighted. There is no dynamic response on a harpsichord, the keys functioning as on/off switches.

If you want to make a digital harpsichord that only uses harpsichord sounds, a real harpsichord keyboard with mechanical sensors would surely work.

If you want to make an all-purpose instrument for playing fortepiano and maybe organ music as well, you should look at other possibilities. A real fortepiano keyboard is a fun idea, but I think not practical: fortepiano keyboards are considerably more complex than harpsichord ones and extremely delicate.

Well, ideally, the instrument could be used for organ sounds, and fortepiano..or for that matter, for ANY sorts of sound samples. My idea is NOT to make an instrument that has sounds, but rather, a sort of controller keyboard that sends the midi signals, but leaves all sound production up to a PC (using Kontakt, for example).

You're right, MRC: Fortepiano action is more complicated than harpsichord. I've played on them, but haven't owned one (though I have owned two harpsichords, three clavichords, and pianos, as well as digital instruments).

My idea is not really to imitate the "feel" of a particular instrument, but rather, to make a keyboard weighted lightly, yet in such a way that dynamics could be easily controlled (as they can be on a clavichord, despite a very light touch). I'm thinking about a "down weight" somewhere in the range of 10 to 20 grams at most (50 grams being more or less the standard for a modern piano).

Any ideas about where to get good mechanical contact midi sensor strips? They'd be a lot cheaper than PNOscan, though I'm not sure about their reliability.

Forgot to comment on something you said, MRC: I don't know what you mean by "by having it spread at the far end like a piano keyboard."

Anyway, I have very specific ideas as to keyboard range. As a matter of fact, I even wrote a little eBook on the subject of the perfect range for a digital keyboard for those who don't want a full 88 key model. (can be had for free at my site, http://www.dboltoncreations.com

Suffice it to say here that the best range for a 5-octave instrument would be FF to g3 (OK: 5 octaves plus a few notes), as only this range allows playing all classical works up to mid-Beethoven. (not possible on the typical 5-ocatve digital with a C to c4 range (too few notes at the bottom, too many at the top).

I was really disappointed by the C-30 when I saw one in the flesh. And the streched / looped harpsichord in our RD-700NX was kind of a let-down too. There is a niche market of ancient keyboard emulators that no DP manufacturer seems to be servicing very well at all.

I would love to see Roland combine and SN update the clavichord, harpsichord, pianoforte, and positive organ in one package - with narrower negative (white <=> black) keys, built in speakers, a !MUSIC REST!, and in a decorative case with detachable legs. That would be worth $3k-$4k USD easy and I'd be all over it.

The back of a piano keyboard usually splays out so it is wider at the back than the keyboard.

Here is the site of a company that makes reduced-width piano keyboards.

I don't think a piano-type keyboard sensor system really maps to how a harpsichord actually works. You'd need to create the sensation of a plucked string, or perhaps even more than one level of a plucked string if you want to simulate the heavier touch of two linked keyboards, which is often the case. Which means if you have a harpsichord action in hand, you also need to include the bit that plucks the string. So it's not really question of weight. I don't know of any off-the-shelf keyboard action that does this kind of thing, but on the other hand if you can sync the plucking action of a real harpsichord action with a MIDI switch for each key you'd pretty much have it. You could use short, dummy non-sounding strings to "pluck" and maybe fit it into a fairly small case. Using an actual plucked action means you'd also have to deal with the mechanical issues of a plucked action - the (presumably nylon) quills wear out and need replaced from time-to-time. There are a lot of nuances here - as the strings vary in size across the span of the keyboard, the plucked touch varies. So is that part of the simulation?

Along with real ones, I played an electronic harpsichord in college in the 80s that somehow simulated the plucking action, although it was about as good as you'd expect for that time compared to a real one.

Yeah, I know JUST what you mean about the C30. It does have built-in speakers, that really aren't bad (maybe you played on one that they had connected to external speakers?). However, the main problem was that some notes were softer than others, and there was no way to fix that. Three guys from Roland came to visit me, to convince me that my perceptions were wrong about this (since I wanted to return it), but I had collected proof, by way of recording each note individually, and determining the volume level of each: a number of notes were simply softer, not just in the harpsichord sound samples, but also on the fortepiano and organ.

In other words, some worker at Roland messed this up, and there was no way to fix it. So I returned it. A real pity, but if I pay $3,000 plus for a digital instrument, i expect the volume level to be even across the board (as it indeed was on the older Roland C20 that i owned years ago.)

I see what you mean, David. But it is not my aim to try to simulate the exact "feel" of a harpsichord, especially since the instrument should be able to play other sorts of samples as well. Being a harpsichordist myself (albeit without an instrument here in Japan), I don't think any digital could effectively simulate the feel of a real harpsichord.

Yet my views on "touch" are not dogmatic: the perfect touch is, in my view, not one that copies that of an acoustic; rather, the perfect touch is one that allows the player to most easily and effectively play the musical compositions of his choice. A keyboard touch must be, therefore:

1) relatively light (even Glen Gould preferred a down weight significantly lighter than that of a standard piano).

2) Responsive, allowing for quick repetition.

3) Easy to control dynamics. Many think that this is most achievable on an instrument with a heavy touch, but that is only because they are USED to a heavy touch. I know pianists who, for example, bought a clavichord. These have an extremely light touch: about 8 grams down weight, as opposed to a piano, that has about 50 grams (and a MUCH heavier key mass). After a few weeks, any pianist can get used to a clavichord touch, and once you do, it's easy to control dynamics, and you can play fast passages very easily as well! Will it "hurt" your piano technique? Well, Bach played harpsichord, organ and clavichord: all three instruments have a very different touch. I myself played harpsichord, clavichord, piano and digital: with a little practice, you just adapt to the different touch when you sit down in front of the instrument.

Light touch, responsive, easy-to-control dynamics, and a good key "feel"...these are the qualities I will be trying to maximize, and "build in" to a digital keyboard!

TO BDB: Thanks for the explanation! Now I know what you mean. And the link: I'll have to visit that site again: that company is on the right track!

Hi Dewster,

You say: "There is a niche market of ancient keyboard emulators that no DP manufacturer seems to be servicing very well at all." Bingo! You are SO right about that. I have written to all major keyboard companies here in Japan with suggestions, and NONE even wanted to hear them "for legal reasons" (even though I said I'd be happy to sign a paper giving them the rights to use my ideas). Totally closed-minded.

Yet consider the Vax77 (often talked about here on the forum). Two guys have great ideas, construct a controller superior to ANY made by the huge companies like Yamaha, Roland, Kawai, etc. TWO GUYS! Very often, the greatest innovations don't come from the corporate world, but by individuals driven by a passion. the VAX 77 (despite the problems they had) is a very good example of that. Now, let's see if something can be done for all those people who love historical keyboards, making a digital that is up to their standards!

The way it seems to work lately in technology fields is a handful of ideas people finally start beavering away on something that's been crying out to be worked on, and once they get down the development road a ways a much larger company buys them up. Often the startup is formed with that end result as the goal as it can be quite lucrative. Very little in the way of innovation can happen when you have ten layers of modern management (i.e. grifters) on your back.

Bob Moog single handedly pulled the Theremin back from beyond the grave by making kits that helped pay his way through college, work that led directly to his pioneering work with modular synthesis. He worked on Theremins his whole life and really pushed the field forward, even though there was little to no money in it - no modern company that I can think of would condone that kind of behavior in a staff engineer, but it's exactly that kind of monomaniacal interest that got us refined instruments like the modern piano, violin, guitar, etc.

It's kind of wild to see Moog Inc. handling the new Theremini product so ineptly, though without Bob around I suppose it's not unexpected. End of an era there.

Excellent example, Dewster!

David, may I ask how large you estimate the demand for such a harpsichord instrument to be, relative to other keyboard instruments (organ, piano)?

Kind regards,
James
x

Excellent question, James. Knowing the field of "old music" pretty well, I'd say the demand could be enough to set up a nice little business. But I might wish to go beyond that: the conception I have for this instrument would make it quite attractive to those who are into other types of music as well.

Let me put it this way: if Roland had used the ideas I now have when they made their C30 digital harpsichord, it would be a MUCH more attractive instrument, and would be of great interest to a much wider audience. As a matter of fact, back in the 90s, I used the old Roland C20 (digital harpsichord) in many concerts that I gave when living in Spain. It served its purpose very well, but it was then that I began thinking about how such an instrument could be even better. At this point, I have a bunch of what I consider excellent ideas to make something that could be truly "evergreen".

So now, I want to finally get things moving, and try to realize this dream. First step: create a prototype. but what I really need is some kind of reliable midi sensor system. Not being an electronics expert, I would prefer to seek a partner who would be in charge of this area. Hey, maybe you know someone who is into the midi part, and would like to take part?

If I were to explain the entirety of my concept, I would have to write several dozen pages, but if you are interested, I would be glad to keep in touch with you. I live near Tama plaza (denentoshi line)...and you?

Enjoy your day!

David

David,

While I do know a few engineers with experience developing MIDI hardware, I'm not sure they have sufficient interest in the harpsichord to dedicate their free time to such a venture.

To my knowledge Kawai has never manufactured harpsichords (analogue or digital), so I'm afraid I doubt there is a great deal of interest in developing such an instrument within the company.

I would suggest that Roland, with their experience of the C20/C30 would arguably be the best company to approach with your ideas. It may be worthwhile contacting their head office here in Hamamatsu to request a meeting with one of the R&D product planning guys.

Best of luck!

Kind regards,
James
x

Thanks for the thoughts, James. I don't think Roland would be interested, since the instrument I have in mind would no doubt cut the sales of their C30, which they are still selling.

Maybe you could tell any midi experts you know about my plan, to assess their possible interest. Feel free to give them my email: dboltoncreations AT gmail.com

Bye for now,

David

David, I don't think any DP manufacturer will be interested enough in your ideas to actually do anything, and even if they are there is no way they're going to collaborate with you on anything. These are closed shops that barely listen to the consumers they sell their stuff to. NIH and all that.

I believe what you want is authentically narrower keys, and this is probably the biggest issue as no one makes them, and it prevents you from using something like PNOscan which is made for normally spaced keys. For a prototype you could saw your own keys from a slab of basswood and use a metal rod as a common pivot. Or dispense with wood entirely and use common aluminum extrusions. Maybe use laminate for white key tops.

The VAX77 used magnets in the keys and linear hall effect sensors for position. You could probably do something similar with capacitance but would need a lot of relatively high speed I/O - which you could get with a low end FPGA. I would go with a continuous sensor as this lets you do literally anything with the key movement e.g. set arbitrary trip points, sense real velocity anywhere in the full range of key movement, etc.

Good luck!