| previous by date | index | next by date |
| previous in topic | topic list | next in topic |
Hey Scott,
Right, "one bit" patterns allow the Klee to act like a normal sequencer, right down to the gate bus. All of that flies out the window once more bits are added.
Yes, the random input will give you a lot of variation in a sequence. The first Klee I built used a CD4006 shift register and had a four bit output, and relied entirely on random input for programming. The pattern inputs were thought of originally as a way to be able to program or return to a pattern without using a pencil and paper, plus, obviously, it provided more output bits.
In fact, I rarely use the pattern bits for programming (they are handy for starting out with a number of bits). Instead, I use them more to "remember" a pattern that occurs through other means. Surprisingly, I rarely use the random input for programming, though it is nice for auto-running random stuff. The most interesting variations I come up with have to do with switching between 8X2 and 16X1 and Invert B on and off. 8X2/16X1 keeps the same number of bits active, but shuffles their positions, which provides a different pattern. Invert B adds (or subtracts) bits of the pattern. Sometimes I will use the random input with nothing applied to subtract bits. Random can also be used to add, because a little mentioned feature is if the ref level is all the way CCW, it provides a constant '1' to the input of register A.
When I get a good pattern going, I'll stop things and flip the switches that are located beneath an LED that is illuminated to "ON" and switches that are below an LED that is unlit to "OFF", which always allows me to return to that pattern after flipping things around even more.
The Klee boards are quite open-ended if you take a look. There are a number of things I'd implemented on the breadboard, but didn't put in just because the panel was getting too busy. If you look at the switching for Random and 8X2/16X1 you'll see they are actually switched by a CD4053. Originally, the idea was to be able to remotely switch the modes via that IC. I wish, I wish, I wish I would have put in those extra comparators and done that, because I use that stuff all the time. "Invert B" came along later, and you can see it is actually switched by the switch hardware itself, but an extra transistor might handle that with no problem at all.
Anyway, back to the boards - you'll see that you can put just about anything in between the pattern switches and the CD4034s that you want. Your idea is quite intriguing - I am not at all adverse to anybody improving things or modifying things or adding things or doing boards that do such things and selling them. Hell, I'd probably stick'em in my Klee =0).
The way the schematics are written makes them a bit hard to follow - they were used directly to manufacture the boards, hence all the "ports" instead of just drawing the controls in. But, if you look, you'll see all kinds of opportunities to do things that one might want to do. My suggestion would be to build an original then see if its worth it - everything it can do is not apparent even from the description of the circuit. I'm still thinking of new things it'll do.
If you haven't seen it, the "Klee 101" thread discusses the functionality a bit more:
http://electro-music.com/forum/viewtopic.php?t=25057
Ideas for modifying the Klee are in the "Pimping your Klee" thread here:
http://electro-music.com/forum/viewtopic.php?t=25584
BTW, one nice thing I've never made a sample of is that the Klee can be clocked at audio rate - the gates and triggers drop off fairly soon, but the CVs just keep chugging along. It would probably sound fairly righteous through a Neural Agonizer.
Take care,
Scott
-----Original Message-----
From: Scott Juskiw
Sent: Sep 22, 2008 11:07 PM
To: ModularSynthPanels@yahoogroups.com
Cc: scottnoanh@...
Subject: Re: [ModularSynthPanels] Klee Third Run Poll
Yes, I see that, thanks for pointing it out. The random input appears
to allow programming one bit into the shift register (serial
programming) . What I am hoping to achieve is programming any of the
bits (even all 16 at once) without having to manually set the
switches. The documentation refers to this as parallel programming.
After a second pass through the docs, I believe that what I'm
suggesting is that all 16 channels have the provision for a random
input, not just channel 1. Not sure if anybody but me would find this
useful.
Also, in my original email I was confusing "programming a new pattern
via the 16 switches" with "playing back the currently loaded pattern
via the 16 switches" which is wrong (the switches are for programming
a new pattern). Sorry about that, I should have read the manual twice
before posting (measure twice, cut once).
On 22-Sep-08, at 8:59 PM, Paul Lord wrote:
>
> Scott Juskiw <scott@tellun. com> wrote:
> One thing that I see as a limitation is that it appears to require
> manual switching of the 16 pattern switches in order to set a new
> pattern or to enable the 16 channels of the current pattern to be
> summed to the output. I don't know if anyone has brought this up
> before, but I think it would be more useful (for some crazy people
> like me) if the pattern switches could be changed via external
> voltage control.
>
> Keep reading the docs :) If I'm not mistaken, that's what the
> random, random gain, and reference input/knob section is for!
>
> Paul
> --
> Could very well be mistaken, but suspect not.
>________________________________________PeoplePC OnlineA better way to Internethttp://www.peoplepc.com