multiple EG bank

[James Jackson]

Good points re the number of controls but.....

Isn't this where we start designing digital front ends to control the
analogue hardware (like, for example, the Juno 106). You can impliment patch
saving, and useful functions like cut / copy / paste parameters and so on.
I'm sure with something like a PIC (or 9) it would be possible to create
something like that, with an LCD screen for text based input (like on the DX21).

Or write a PC front end.

Or.......

James Jackson.

At 12:54 PM 12/29/99 +0100, you wrote:
>It's surely possible, and it's implemented (in software) in several
>commercial synthesizers. No problem to build something like that
>in analogue hardware, too.
>
>The big question is: Is it worth the effort ?
>And an even more important question: Is it comfortable to use ?
>In theory it's total bliss to have control over so many parameters -
>but in practice it is often not. I mean there is a good reason for these
>fourier based synths didn't have big commercial success.
>IMO the goal for developing a good synth is to get maximum control
>from a reasonable number of parameters. That's why the DX-7 with
>its FM (PM ... hi MC. !) was so much more sucessful than additive
>concepts. Mind you, the DX-7 has *few* parameters in comparison -
>and still too many for most people to feel comfortable programming it.
>
>There are other concepts of quasi-additive spectral shaping that better
>fit to analogue implementation. Vocoders are a good example. My own
>Interpolating Scanner (combined with a EMS filterbank) is another
>possible approach. Both concepts are limited to a few band pass filters
>(8 in my Scanner, something between 8 and 22 in analogue vocoders),
>but even with such a limited number of filters that would be many,
>many Attack, Decay, Sustain, Release parameters. I mean, just go and
>*write down* "Attack, Decay, Sustain, Release" 22 times. Now think
>of building it. Now think of adjusting these 88 parameters for every
>sound.
>And what have you gained ? A sound with a sophistcated harmonics contour
>that sounds - *boring*.
>Why ? It will always sound the same. Because you have not yet implemented
>anything to modulate these 88 ADSR parameters in real time. So you
>want to add velocity, after touch, random voltages to each parameter.
>How many parameters now ? I've stopped counting.
>To handle something like that, it's important to find clever algorithms
>to distribute modulations to the different channels. Graphical input
>like on the Axcel Resynthesizer is the best thing in this direction I've
>seen so far, but still it's not much fun to program (your fingers start to
>hurt
>after 10 minutes).
>
>JH
>
>> Hello
>> Has anyone heard about something like a "multiple Envelope Generator
>bank"?
>> (own definition)
>> Theoretically, a waveform is splitted into its harmonics and each one
>> receives its own ADSR curve. I thought about asociating some band-pass
>> filters (max 4) to a set of envelope generators. These EGs should have a
>> "initial delay" setting for building up complex dynamic progressions. Has
>> this been done either commercially or privately?. I think that the Kawai
>> K5000 does something like this, but of course that is digital additive
>> synthesis :P)
>> This is purely experimental and speculative, so any idea or comment will
>be
>> appreciated.
>> --
>> ==========================================================
>> Carlos Vila Deutschbein        si04697 at salleURL.edu
>> Enginyeria La Salle            www.salleURL.edu/~si04697/
>> ----------------- Barcelona, Spain -----------------------
>> ==========================================================
>>
>>
>>
>
>
>
>
>
--------------------------------------------------
James Jackson - New Era Web Designs
E-mail: james at newd.demon.co.uk
WWW: www.newd.demon.co.uk
Phone: (01438) 840993
--------------------------------------------------