From bruce@ashbysolutions.com Mon May 12 15:19:17 2003
Subject:Re: Digest Number 1176

Hi Schel,

At 01:27 PM 5/12/2003 +0000, you wrote:
>Forgive my lack of knowledge concerning the actuall working
>mechanics of the key contacts, but as i understand them, the
>principal is this: 9 key switches all stacked under each other, each
>with its on wire activating a particular draw bar tone.

Yes, this is correct, but ...

>Ok heres my idea: Take a really nice fast computer, using Vstack (or
>similar VSTi host) run 9 INSTANCES OF B4 EACH
>ASSIGNED TO A DIFFERENT MIDI CHANNEL. RUN 9 OF PETKOV'S B4 ENCODERS,
>RUN THE CORROSPONDING
>KEY SWITCH TO EACH ENCODER, INTURN TO EACH INSTANCE OF B4 WITH ONLY
>THAT PARTICULAR DRAWBAR
>PULLED, ALL THE OTHERS SET TO 0. This would infact reproduce the
>effect would it not?

No, it really wouldn't. it would be an interesting exercise, but there are two things that you probably wouldn't be able to reproduce:

1) A B-3's switch bussbars are supposed to be perfectly even, allowing near-simultaneous switch triggering, but in fact they are at least a little curved and warped, especially with age. That means that each key has a slightly different sequence of timings in the key triggering. One key may trigger all nine contacts evenly 1 millisecond apart for a certain key strike, while the next one may go 1-2-2-1-1-3-1-1-1 or something like that. You're not going to get that much variation between nine copies of the B4, unless you are running other applications at the same time -- and that's a recipe for disaster.

2) When you play fast passages or "machine-gun trills" on a B-3, not all nine contacts close every time a key is struck. Sometimes only the first few switches have closed by the time the note is released. Nine copies of the B4 doesn't do this, because it CAN'T. In fact, no amount of programming magic can take a modern two-contact keyboard and make simulate this effect perfectly. Standard two-contact velocity measures the time between the two switch closures, and by definition, one of the contacts closes near the full travel of the key. Try this: Press a key on your electronic piano or synth about 1/3-1/2 the way down, then let go. I'll bet it makes no sound at all, right? Try it on a real Hammond, and you'll get a different result.

As CPU horsepower increases, I'll bet that someone may try to add in an algorithm that comes close to simulating nine-contact response in many cases, but it won't sound the same for a very light-touch technique.

>Do any other clones reproduce this effect?

Only the "new" B-3, because it has nine key contacts per key. There have been advances is location sensors in recent years, which might be used to determine how far down a key has been pressed (rather than how fast). Since it doesn't really do much for the rest of the keyboard world -- i.e., non-clone players -- don't expect this technology to show up in $1500 clones anytime soon. Even the return of the waterfall keyboard was an effort. If this were 1991 and the only "serious" clone makers were Hammond-Suzuki and Voce, Inc., do you think Fatar would have been interested enough to make the effort? I doubt it. 2001 was a different story, with a half dozen companies re-entering the Hammond clone market, but getting them interested in a follow-on product that requires a whole new electrical interface is probably another story.

Regards,

-BW
--
Bruce Wahler
Ashby Solutions™ http://music.ashbysolutions.com
978.386.7389 voice/fax
bruce@ashbysolutions.com