From bruce@ashbysolutions.com Wed Mar 19 06:17:02 2003
Subject:Re: Leslie frequency response and clone costs

Hi Alan,

>Now, I've never owned a Leslie speaker, and while I appreciate the
>genius of what Don Leslie created many years ago, it's hard to
>believe that either the amplifiers or the speakers available to him
>40+ years ago were anywhere near as accurate as, say, a Mackie SRM-
>450.

Try 50-60 years ago! The original Leslie Speaker was made in the 40's. With the exception of speed control (2-speed, DC servo motors) and materials changes (plywood -> particle board, wood rotor -> foam rotor, different woofer and driver OEMs), nothing about the Leslie has really changed since the Model 22H.

>Can anyone comment on the frequency response of the typical dual
>rotor Leslie, and how that might compare with today's better
>amplifiers/speakers?

The Leslie 122/142 has a frequency response that rolls of sharply above 5KHz. The 147/145 has a response that goes a little higher -- 6-7KHz, with a slight bump up just before it falls off. Neither of them produce any appreciable sound above 10KHz, even if the original Jensen driver is replaced with something more full-range like and EV or Gollehon tweeter.

Certainly, the use of a real Leslie will take some of the edge off a clone's high end; the same is true when using a Speakeasy preamp before the amplifier. The more original or modeled-after-the-original components in the chain, the closer the approximation. However, I don't think that's the complete story ...

Part of the answer lies in the economics and other factors of modern keyboards. In order to accurately reproduce a 6KHz tone, we need a digital processor that runs at a higher frequency than 6KHz. How much higher? An often quoted (and abused) mathematical concept called the Nyquist Criterion says that 2x the frequency (12KHz) will do. In the audio world, that doesn't really work out well; I can tell you the frequency of the waveform with 2x sampling, but I can't tell you the shape of the wave (sine, square, etc.) To do things fairly well, you need 10x or so. To really do things right, you need at least 40 points per wave; that's 240KHz! A DSP that can process at those speeds is quite expensive, and that cost has to be passed on to the consumer, plus profit, of course. (This IS a business proposition, not pro bono work.) It's also harder to program at those speeds, because certain standard programming practices become logjams at high speeds. Figure to add at least $500 to
the cost of the clone, maybe more.

I'm sure that there are a few list members who would seriously consider a higher-priced clone with that kind of response, but let's be honest: most of us would probably opt for "fairly well," which is exactly where we are with the current batch of clones. I've heard the groans about the cost of BX-3, CX-3, Electro (especially the rack version), and Key5. I watched the recent discussions about a B4 control keyboard with interest, because I contemplated building and offering a similar product over a year ago. $500 (US) tops? You might be able to cobble together a low-grade keyboard and controller solution for something around that price, but you're not going to offer good quality waterfall keys and authentic-feeling drawbars for that money. For one thing, suppliers like Fatar and GEM are not all that interested in your business, because they know that you're only going to sell hundreds -- maybe even TENS -- of these keyboard setups. Translated into business terms: you can't
get the parts for the cost that Roland, Korg, Voce, and even Fatar (Studiologic) themselves depend on to make a profit. If I can buy the parts for say, $475 and sell them for $500, is that a great deal? Only if I can sell hundreds of them a week. If I sell 50 of them, total, over 6-12 months, it becomes a slightly profitable "hobby." (Do the math: That's $1250 total profit, for probably 2 months development and marketing effort, plus support time.) The alternative is to sell them at niche-product prices ($800?), but according to the discussions I've heard here, the current market isn't interested at that price.

The same issues are true of high-end clones. If the new B-3 sounds as good as reported -- and I haven't heard anyone come out and say that it DOESN'T sound great -- why is everyone upset about the cost? Because price DOES matter in today's market

Regards,

-BW

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