Author |
Message |
sebber
Joined: Aug 27, 2004 Posts: 501 Location: Berlin
Audio files: 3
G2 patch files: 33
|
Posted: Mon Dec 01, 2008 3:08 pm Post subject:
Alvin Lucier recreation |
|
|
Hi, I'm trying to recreate a score by Alvin Lucier. There's a tape with oscillators in it and in the description it reads:
"Oscillator 1 rises from E4 to E6 at 18'' per semitone, descending back to E4 at 22''."
I tried a lot with the LFO Shape Osc, but I don't get a steady movement - it's faster in the beginning and slower in the higher register.
I also don't like the quite audible changes in the tone. There are still too many steps, it's not really smoothly rising and falling.
Finally I have no clue how a constant module modifies the direction/rate ratio. I know what it does but I can only guess _how much_.
Mayby it's a good idea to count 24 times 18'' up and 24 times 22'' down, instead of doing the whole thing at once, but I don't now yet how to realize that. Any ideas in either or another direction? _________________ Fish don't swim. They dive. |
|
Back to top
|
|
|
ian-s
Joined: Apr 01, 2004 Posts: 2669 Location: Auckland, New Zealand
Audio files: 42
G2 patch files: 626
|
|
Back to top
|
|
|
sebber
Joined: Aug 27, 2004 Posts: 501 Location: Berlin
Audio files: 3
G2 patch files: 33
|
Posted: Tue Dec 02, 2008 4:31 am Post subject:
|
|
|
Unfortunately it's a bit more complex than that. The whole "swing" from E4 to E6 and back takes 16 minutes. It goes up a little faster, taking 7'12'' at a rate of 18'' per semitone and down 8'48'' with 22''. I don't need the semitones themselves, it's one big sweep.
There's another sweep in the opposite direction, but that's easy enough if I find the solution for the first one.
I've got two problems:
the sawtooth doesn't go steadily at a pace of 18''/semitone but is faster in the beginning and slower in the higher register. I don't understand that, I expect a linear behaviour from a sawtooth. The note quantizer is just in there to prove the point, it shouldn't be there.
The other thing is that I do hear certain "steps" in the sweep, which I don't want to have in there.
Finally I've got modulators for both direction and rate in there and don't really know how all these interact. Well, I know what they do, but I don't know if I can really calculate exact times with them or if I should rather look at them as being knobs on a virtual analog synthesizer.
All this leads me to think, that doing it in one big sweep might not be the best solution, but I'm a bit stuck right now.
I remember having read about the longest possible sweep or so (Jan??), does anybody have an idea what my brain is trying to tell me?
Ian-s, you're right, the ratio is 9:11, that's 45% to 55%. I don't know what IPS is.
Description: |
Trying to get a real nice sweep over two octaves in 16' time. |
|
Download |
Filename: |
LucierX.pch2 |
Filesize: |
1.58 KB |
Downloaded: |
1480 Time(s) |
_________________ Fish don't swim. They dive. |
|
Back to top
|
|
|
sebber
Joined: Aug 27, 2004 Posts: 501 Location: Berlin
Audio files: 3
G2 patch files: 33
|
Posted: Tue Dec 02, 2008 5:28 am Post subject:
|
|
|
Ah, found the main mistake myself. It wasn't the Osc's fault that the sawtooth wasn't linear, but that I sent it through an X-fade that was set to exp.
One question remains: how exactly does the direction input on the OscShape work? In the help files it says:
Osc: 200Hz, dir input: 64; Osc=200Hz
Osc: 200Hz, dir input: 0; Osc=0Hz
OK, I get that. Is there a way to calculate exactly this:
Osc: 200Hz, dir input: (126(bidir)); Osc=198,42? _________________ Fish don't swim. They dive. |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Tue Dec 02, 2008 12:16 pm Post subject:
|
|
|
sebber wrote: | I remember having read about the longest possible sweep or so (Jan??), does anybody have an idea what my brain is trying to tell me?
Ian-s, you're right, the ratio is 9:11, that's 45% to 55%. I don't know what IPS is. |
A little over 7.5 hours I think .... being 64 semitones below 699 s, or 699*(2^64)^(1/12), maybe even slower by inputting less than -64 into the rate input of an LFO.
I think Ian read your " symbol as inches and not as minutes, in that case IPS would be Inch Per Second, or tape speed.
Are the steps you are hearing really there or is it a psychological phenomenon? I did a little test where I have two oscillators frequency modulated by the same signal (an LFOShapeA in asymmetric triangle mode) but one going through a sample and hold set to sample about twice per minute, I then hear a very gradual change in the beat frequency ... whereas when I run the swept one on its own my ears seem to quantize ... weird
The dir input is a linear frequency control input, but 126, 127 and 128 settings are a bit odd at times, as step is missing. I checked it for 64 and 32 and 48 (32 -> 48 :: 7 semitones, 32 -> 64 :: 12 semitones).
Will attach my test patch, probably you'll want to set it faster as it's really going sloooooooooooooooo .....
.... ooooooooooow
now
Description: |
|
Download |
Filename: |
slow_sweep.pch2 |
Filesize: |
1.48 KB |
Downloaded: |
2216 Time(s) |
_________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
sebber
Joined: Aug 27, 2004 Posts: 501 Location: Berlin
Audio files: 3
G2 patch files: 33
|
Posted: Wed Dec 03, 2008 3:50 am Post subject:
|
|
|
I heard the tone/sound (not the pitch) change in certain time intervalls (ca. 0.3s) when the LFO that changes the OSCs pitch went through the logarithmic x-fade. It's not there when the x-fade is linear. I thought I could literally hear the machine calculate. (OK, now, wait, not so fast, slowly, here she comes, gee, what a beauty, there's another one, lift it up, lift it up, slooowly)
I now have the ShapeLFOs at 44% / 55% respectively and the rate input is controlled by a constant of 48 (112). That's pretty close to what I want, but I sat there with a pen counting seconds for a while. _________________ Fish don't swim. They dive. |
|
Back to top
|
|
|
|