I believe this is roughly an LPG. Note that the EnvAD affects both the filter and the amp. The filter is LP6, so only first-order (an RC circuit). The Osc and Scope are just for testing.
Talking 'west coast', I'd really love a function generator: control over rise and fall time, rise and fall shape, CV input to engage cycling ('looping') or one shot mode, and a bi/unipolar output option. Integrated EOR, EOF gate outputs would be the icing on the cake (end of rise, end of fall). Something drawing from the likes of Make Noise Maths or Frap Tools Falistri
It's a bit tricky to roll your own in drambo at the moment (or it was last i tried several months ago). The current envelopes don't quite have the time range or control over stage shape.
I think source of uncertainty had some unique distribution / probability curves. Something like biasing random output to low/mid/high values. So it's more than just pure random. I tried to build something like it a few months ago, but gave up with the distribution curves 😅
Would be curious what solutions there are for this!
Correlation" is basically a cross fader between the noise-synced triangle wave and the output of the sample and hold. At zero, the output changes randomly from one value to the next based on the external clock input with no correlation to the previous voltage level. At the max setting, the voltage doesn’t change at all, as it is just sampling itself. In the middle, the output is influenced by its previous states, with the Correlation position effectively defining how far the new random value can deviate from the old one—a simple, but profound way of implementing control of just how random the output is
the inputs are complex random clicks and the expected output is ähh lets say more of that but altered
im reading something with much bigger internal blah blah ...
English isn't my native language and I understand shit about math
this is the math/logic part I dont understand : what does this say?
"if n=1, then the n+1 output can produce two different values (as 1+1=2), and the 2^n also can produce those same two values (as 2^1=2) ... If n=2 then the n+1 output would choose between three values; the 2^n output, however, would choose between four values (2+1=3, but 2^2=4). With 3 selected, the n+1 is 4 and the 2^n is 8, etc."
Ok. That sounds pretty clear. A couple of questions:
How high is "n" likely to be? 2, 5, 20, 100?
How are you selecting the output from the number of available values for the "n+1" case? Is it a random draw from 0..N+1? Presumably you'll use the same method for 2^n. Do you actually need the value of 2^n to make the selection?
2^n used with a shift register usually means that each output has a different weighting: 1,2,4,8,16... so if you multiply each oitput signal with the respective number and add them all then you have a shifted binary to decimal converter.
Comments
I believe this is roughly an LPG. Note that the EnvAD affects both the filter and the amp. The filter is LP6, so only first-order (an RC circuit). The Osc and Scope are just for testing.
try this for starters
an LPG is my wishlist, too.
(i have no idea how to replicate the stored random stuff from source of uncertainty )
but it also offers noise and s/H ...
I think what happens
it takes a sample from random noise and stores the result in buffers (bbds)
and then does some hopping over in the buffers read out ? no?
Thanks guys I knew that someone smart would have some drambo workarounds! Thank you! Still fiending for those vactrols though! :D
hm, LPG is a very musical wish burned into one module
VCA+Filter combo mode
if signal is louder signal is brighter if its low in volume its a little muffled ...
(this imitates what "real" instruments do)
on every synth in the world you have to set this up yourself ( I wonder why? its so logical to do that ...)
with an LPG you have to do nothing
you just plug it in turn it up or down with something - and it does its thing
its not really meant to be a filter in the classic sense, I guess
I think of it as "musical vca" (the somehow "slow" reaction from vactrols adds to this)
but there is one disadvantage : usually lpgs dont close totally. (sure dont want that, a vca that doesn't shut up ...)
Talking 'west coast', I'd really love a function generator: control over rise and fall time, rise and fall shape, CV input to engage cycling ('looping') or one shot mode, and a bi/unipolar output option. Integrated EOR, EOF gate outputs would be the icing on the cake (end of rise, end of fall). Something drawing from the likes of Make Noise Maths or Frap Tools Falistri
It's a bit tricky to roll your own in drambo at the moment (or it was last i tried several months ago). The current envelopes don't quite have the time range or control over stage shape.
close but no cigar
u can trick around by switching from loop to one shot via morph or scene morph ...
"end of this envelope part" is indeed interesting
@aleyas Sounds much like an extended Env AD Mod.
@aleyas I built one month ago, not perfect but it works.
@ZhouJing
looking at your patch: oh interesting. I should use slew. 🙏🏻
thx for pointing this out. :)
In the new version there is a shift register
so „source of uncertainty“ should be doable.
something like this?
?
i don’t have the module I can’t look what it’s output looks like
I need to look closer into how exactly the flip flops are used in that module
but it’s something like this
im not sure if it’s jumping randomly over the Flipflops
or if there is more logic involved ?
i guess there is …
more some other time
have fun
I think source of uncertainty had some unique distribution / probability curves. Something like biasing random output to low/mid/high values. So it's more than just pure random. I tried to build something like it a few months ago, but gave up with the distribution curves 😅
Would be curious what solutions there are for this!
What do you mean with distribution?
more or less clicks ?
playing with the filter does that
If I turn tone down it gos brrrrr the more I turn it up, is more sporadic click click silence silence silence click
i need to read up a little and what he is doing with the flip flops , i bet there is some trick with read this output read that output involved
The distribution stuff is just some other noise spectra than white
Aha
Correlation" is basically a cross fader between the noise-synced triangle wave and the output of the sample and hold. At zero, the output changes randomly from one value to the next based on the external clock input with no correlation to the previous voltage level. At the max setting, the voltage doesn’t change at all, as it is just sampling itself. In the middle, the output is influenced by its previous states, with the Correlation position effectively defining how far the new random value can deviate from the old one—a simple, but profound way of implementing control of just how random the output is
Ok
that sounds like much more Chaos
Hah 😁😎
It’s this. (Without the crossfade)
🤔 so far not bad
I need help with the math and the logic now for the stored random thing
how to do the 2^n. Thingy ?
I always fail at This 🤷🏻♂️
It’s some math with the outputs of the shift Register
i suck at math, I can’t do the simplest things
logic is fine, but math me no clue.
help
Ach who cares,
instead of 2n I use axb
we are after random things. 😃
Maybe I can help with 2^n. What does it actually do? What are the inputs and expected outputs?
the inputs are complex random clicks and the expected output is ähh lets say more of that but altered
im reading something with much bigger internal blah blah ...
English isn't my native language and I understand shit about math
this is the math/logic part I dont understand : what does this say?
"if n=1, then the n+1 output can produce two different values (as 1+1=2), and the 2^n also can produce those same two values (as 2^1=2) ... If n=2 then the n+1 output would choose between three values; the 2^n output, however, would choose between four values (2+1=3, but 2^2=4). With 3 selected, the n+1 is 4 and the 2^n is 8, etc."
Ok. That sounds pretty clear. A couple of questions:
How high is "n" likely to be? 2, 5, 20, 100?
How are you selecting the output from the number of available values for the "n+1" case? Is it a random draw from 0..N+1? Presumably you'll use the same method for 2^n. Do you actually need the value of 2^n to make the selection?
i have 8 hm different streams of random numbers that fall out of door 1,2,3,4,5,6,7,8
i dont understand what the math says
do I need to calculate blah plus 1 and compare it to the other numbers or something ? or compare numbers and then do if it then ?
it must be something very simple but I dont get it
currently I do door 3 x door 5 (this is interesting but not what this says it should do)
damned I always went to the mens room smoking when I had math classes. I should have payed attention. 🤦🏻♂️
2^n used with a shift register usually means that each output has a different weighting: 1,2,4,8,16... so if you multiply each oitput signal with the respective number and add them all then you have a shifted binary to decimal converter.
I believe you're just meant to select one of the doors. So, for n=3, n+1 selects one of doors 1..4, 2^n selects one of doors 1..8.
@lala Could you maybe post the document you're using, or a link to it? Maybe if we can see the context, we might understand what it means.