From the original SynthMaker forum (2004–2014), with its members’ names and dates.
infuzion ·
http://rapidshare.de/files/2436828/Rand ... n.osm.html
(open the link in a new tab/window, scroll to the bottom, then click on {Free}. The link will be on the bottom of the new page. Will expire after no downloading for 3 weeks.)
Synopsis: Don't connect multiple cloned "Randomizers" to the same trigger source expecting different results. Change their seeds atleast.
Trick: Try disconnecting the Trigger Thread & re-connecting the ones now disconnected directly to the BSwitch...
infuzion ·
infuzion wrote: [Synopsis: Don't connect multiple cloned "Randomizers" to the same trigger source expecting different results. Change their seeds atleast.
What is the best way of doing this? Anyone? TIA
infuzion ·
infuzion wrote: [Synopsis: Don't connect multiple cloned "Randomizers" to the same trigger source expecting different results. Change their seeds atleast.
What is the best way of doing this? Anyone? TIA
NothanUmber ·
Hi,
infuzion wrote: Change their seeds atleast.
That's the way to go I think. The most simple way is to open the generator and type some arbitrary numbers inside the SeedX and SeedY data components. (One could make these values input parameters of course, then you'd have to add a "reset" trigger input to the "initializable int sample and hold" that is triggered once the input changes)
Greetings,
NothanUmber
P.S.: Another trick would be the following: Replace the "initializable Sample and Hold" with a normal one, feed in the seed values once and implant it into the generator. Then use a manual trigger device to change the value. The last value get's saved in the sample and hold if you save the schematic so you can have different seeds without explicitly specifying them.
NothanUmber ·
Here a new version of your test that uses a slightly modified random number generator. (The advantage: You can change the seed just by individually sending some manual triggers. Disadvantage: You don't have reproducable results any more as the random number generator doesn't reset on reload anymore).
Greetings,
NothanUmber
infuzion ·
NothanUmber wrote: (The advantage: You can change the seed just by individually sending some manual triggers.
That's a neat idea!
Unfortunally, the inital problem still remains: If you have the Randomizer in your toolbox, then drag it to several areas of your schematic, & have them all connected to the same trigger, you'll end up with the same "random" outputs. Good explaination:
http://www.sawtoothdistortion.com/Artic ... mness.html
So, each individual randomizer needs to be initalized randomly, preferably on load. I can't seem to figure that part out yet.
NothanUmber ·
Hm, you could do that if you'd have e.g.
* an "after create" trigger that triggers after the component was dropped at the schematic
* some possibility to access the running time of SM (or the vst) in milliseconds, the system time... just something like that
With these two informations you could automatically generate a unique seed value for every instance of the generator.
But I also don't have an idea how one could achieve this behaviour with the current SM version without manual "desynchronization" work from the user side (either by directly specifying a seed value or by manually sending some triggers to the individual generators)...
Greetings,
NothanUmber
NothanUmber ·
To have completely different seeds after each reload one would (e.g.) require:
* access to the system time
* a "unique index" component. (This would be a component that can only be placed inside synchronized components. It outputs a different number for each instance)
That'd do it I think but as the method above it's impossible with the current SM version...
RobR ·
Just a thought - is there no way to get system time within a code component? You could use the seconds (for example) as a seed then. Of course, I could be completely off track here 😀
NothanUmber ·
I don't know if there are some restrictions on the assembler component but in theory you should be able to do system calls from assembler - but that's a little bit masochistic, begging to jorge and malc until they provide you what you need seems much more tempting to me 😉
(I don't know a way how one could get the system date (or anything else that could be suitable as a replacement) out of the "high level" code component)
Greetings,
NothanUmber
jorge ·
Creating a component to generate green random floats is very easy. If this is urgent for someone we can include it with the next release.
infuzion ·
jorge wrote: Creating a component to generate green random floats is very easy. If this is urgent for someone we can include it with the next release.
It seems some way of getting a unique number could be helpful, unless you think synth creators wouldn't mind initalizing a module after dropping it, or doing some extra programming. Any copy/pasted or toolbox-dropped random-generating modules will have the same seed, hence give equal numbers, at least initally (but will grow more random as time goes on).
jorge ·
infuzion wrote: Any copy/pasted or toolbox-dropped random-generating modules will have the same seed, hence give equal numbers, at least initally (but will grow more random as time goes on).
The component I had in mind would have a way to re-seed. Could be a boolean input or a second trigger.
infuzion ·
jorge wrote:
infuzion wrote: Any copy/pasted or toolbox-dropped random-generating modules will have the same seed, hence give equal numbers, at least initally (but will grow more random as time goes on).
The component I had in mind would have a way to re-seed. Could be a boolean input or a second trigger.
Well, I could just add a trigger to NothanUmber's remake to do that 😉 Infact I'm planning to make his even cooler B) We were hoping for something that re-seeded automaticly, like read the system clock's seconds or something similar, atleast on program load.
Tom7777 ·
Any copy/pasted or toolbox-dropped random-generating modules will have the same seed, hence give equal numbers, at least initally (but will grow more random as time goes on).
Sounds good 🙂
I already have about 5 different randomizers i have made from a while back but they probably use alot more cpu than a module designed just for it would 🙂
Wrote something here? To have a post removed or your name hidden, write to [email protected].