<a name="planetz-file"></a><a href="http://www.planetz.com/Pulsar/files/mod ... .zip"><img src="/forums/images/file_icon.gif" border="0" alt=" File"> File</a><BR> <a name="planetz-tag"></a>Type: Synth<BR> <a name="planetz-tag"></a>Pulsar Version: Pulsar 2.x<BR> <a name="planetz-tag"></a>Requires: Modular 2<BR> _____________________________________<BR><BR> this explanation describes how one can push a button and get lots of differing random values outputed from s&h modules - which offset any number of differing parameters modulation inputs.
its a 3 for 3 deal for the most part in this patch, not limited to in application.
3 sine oscillators, 3 pitch modifiers, 3 sample and holds, and 3 random signal generators. sounds like the start of an international house of pancakes commercial...lol!
an on/off switch is fed a constant value of 127 (or anything for that matter, not sure of the logic behind this...it works).
the output from the on/off switch feeds the trigger inputs of the 3 s&h's. each time this on/off switch is cycled on/off (the s&h only understands on, not off....i believe) they will "capture" the value of whatever is being fed into them. in this case these signals being fed into the s&h's are 3 random signal generators, one for each oscillator.
you could also use a single noise oscillator (or single random signal generator for that matter). which would be much less of a hit on dsp but you have no control over the random signal as you do in the random signal generator.
having multiple noise oscillators i think would be redundant, maybe not.
the s&h modules are set to trigger on an "audio" signal instead of a gate signal (i haven't figured out the real difference between the two, yet). the thresholds are effectively turned off. so that everytime the on/off button is depressed, the s&h's will trigger, no matter what value is set in the constant value module feeding this on/off switch.
the s&h's input is fed by one of the random signal generators (one per oscillator) stepped output. i guess one could use one of the other outputs, effectively generating a slightly different random signal. i just thought it made sense to use the stepped output.
the level of each random signal generator sets the range of values outputed. in truth the depth value on the pitch modifiers functions in a similar manner. what would be nice is to assign a "range" of values to work within for each random output. the depth controls on both modules are i believe just attenuators. they only attenuate the "level" from the maximum value, squishing the top - unaltering the bottom.
i made a post on cludging together a range limiting solution in the modular, but it doesn't work quite right as of yet.
the rate parameter on the random signal generator sets the rate of change in random values. it may prove to not really influence the random nature of all the signals...i'm not sure, but i did think this would lend itself to giving more of an opportunity to get differing random values from each random signal generator. by setting the rates differently.
the rest of the patch is pretty self-explanitory i think. press a key and hit the on/off button a few times, wahlah, randomly changed semitone offsets for each oscillator. each "value" is interpreted as a semitone increase or decrease in pitch.
<font size=-1>[ This Message was edited by: castol on 2002-03-24 04:51 ]</font><BR><BR><a name="planetz-fileimage"></a><IMG SRC="http://www.planetz.com/Pulsar/files/mod ... ection.jpg" BORDER="0">
Example of Random Value Offset Change
thanks. i had run across doing something of this sorts in one of the excellent nord modular tutorials, but i didn't envision a use for it until i was working on that 8 sine oscillator (additive) patch i mentioned. this idea t just sort of came out of the blue as i was tuning the frequency multiplier modules for each of the oscillators (ugh!) trying out different things for some presets.
i think it was in the s&h tutorial (makes sense)..but i don't recall exactly.
i posted the links to two urls which have a whole bunch of them in a previous reply to someone in the modular talk section.
best.
<font size=-1>[ This Message was edited by: castol on 2002-03-23 10:07 ]</font>
i think it was in the s&h tutorial (makes sense)..but i don't recall exactly.
i posted the links to two urls which have a whole bunch of them in a previous reply to someone in the modular talk section.
best.
<font size=-1>[ This Message was edited by: castol on 2002-03-23 10:07 ]</font>
Here's one of your 2 links, reading them now, very interesting stuff!
http://www.cim.mcgill.ca/~clark/nordmod ... k_toc.html
The other was http://www.clavia.se/nordmodular/Modularzone/index.html
Thanks a lot man, I'm finally diving into modular
http://www.cim.mcgill.ca/~clark/nordmod ... k_toc.html
The other was http://www.clavia.se/nordmodular/Modularzone/index.html
Thanks a lot man, I'm finally diving into modular
yah, all those tutorials provide a MASSIVE base of interesting things to do and experiment with, in modular synthesis. the cw tutorials are....eh...well...
its just the differences between the two systems inhibits doing quite a lot things in the tutorials, or rather i just don't know how to interpret them from nord modular to cw modular.
we need <a href=" mailto:johnbowen@bigplanet.com">john bowen</a href> to teach cw modular classes (he has expressed this as something he would like to do) ! heh. everyone inundate him with emails expressing they would be willing to pay a marginal sum to partake in them!
we would have a leg up on the nord modular users....no? well..i dunno, it is hard to even begin to match the amount and depth of info in those tutorials. still, it would be cool i think.
if anyone doesn't know who john bowen is have a read <a href="http://www.zargmusic.com/bp001data/">here</a href> for a varied and interesting (synthesizer) history on him.
i only found out about him when i was looking into creamware and their 3rd party developers. pretty cool, i don't know him personally but the few times i've exchanged email hes been extremely friendly.
<font size=-1>[ This Message was edited by: castol on 2002-03-23 09:28 ]</font>
its just the differences between the two systems inhibits doing quite a lot things in the tutorials, or rather i just don't know how to interpret them from nord modular to cw modular.
we need <a href=" mailto:johnbowen@bigplanet.com">john bowen</a href> to teach cw modular classes (he has expressed this as something he would like to do) ! heh. everyone inundate him with emails expressing they would be willing to pay a marginal sum to partake in them!
we would have a leg up on the nord modular users....no? well..i dunno, it is hard to even begin to match the amount and depth of info in those tutorials. still, it would be cool i think.
if anyone doesn't know who john bowen is have a read <a href="http://www.zargmusic.com/bp001data/">here</a href> for a varied and interesting (synthesizer) history on him.
i only found out about him when i was looking into creamware and their 3rd party developers. pretty cool, i don't know him personally but the few times i've exchanged email hes been extremely friendly.
<font size=-1>[ This Message was edited by: castol on 2002-03-23 09:28 ]</font>
Good idea, Castol. I think John abandoned the Modular Tutorial classes because of limited response. If enough of us are willing to pay the $25 (or whatever) fee for this project it may be the financial incentive to revive the idea. This is an opportunity to learn from the master. Let's let him know how many of us would subscribe.On 2002-03-23 09:22, castol wrote:
we need <a href=" mailto:johnbowen@bigplanet.com">john bowen</a href> to teach cw modular classes (he has expressed this as something he would like to do) ! heh. everyone inundate him with emails expressing they would be willing to pay a marginal sum to partake in them!
/dave
its good to see another person express interest. now could there be perhaps 30 more of you? when i talked to john, this is around the breaking point in number of interested people willing to pay before he'd consider thinking seriously about doing the classes.
i might be up for campaigning and publicising the idea of these around the web. but on second thought....i don't have the resources to put much effort into that at this point in time.
maybe someone else would though?
i did some more fiddling around with this idea i posted and the patch i originally envisioned it for. now i've got it implemented satisfactorily
not exactly how i wanted initially, but instead of how i initially thought i'd have to give up a feature (transposition of all sine oscillators) i've now got both working side by side, happily. it can be a bit fiddly if you only want one and not other and their full range of values. not exactly hard to do, but fiddly none the less.
i have another crazzzzzy idea that could switch between these two functions, but it would be very elaborate and just plain silly for what it provides.
i did a similar sort of thing in my dis+taste_v1-1 patch with the external input routing through the filters and effects section. if one is interested to know what i'm talking about.
i posed the question in my original posting of one random signal generator/noise source -vs- multiples of these. the later (multiples) i thought would be needed to trully get differeing random values for each s&h. one would think that the same random/noise signal is going to each s&h, and this is partially true in the value and rate sense of the matter....but the nature of software modular synthesis means that nothing happens at exactly the same time (like how midi is). so there is a short delay between each signal coming from one random/noise source and each input going to each s&h.
i forget the name for this behaviour, but again it is something covered in.....guess what? the nord modular tutorials
i guess this can get one into trouble when dealing with lots of modulation massaging/routing. where one needs things to happen in a certain sequence of events (some events have priority over another, and if one knows these rules they can choose the event of actions). apparently it can be compensated for by adding small amounts of delays strategically in the patch. that is as far as my understanding goes.
for the most part these delays aren't noticeable or anything to really have to bother with.
so....one really doesn't need one random/noise source for each s&h, in order to get differening values for each. by the sheer nature of these delays you will allways have differing values being fed into the s&h's. though it is logical to conclude that these delays once everything is patched up are constant. so maybe having a source of random/noise per s&h still has some merits.
in this patch i'm just settling on one simple noise source for all 8 s&h's. just because this one part of the patch takes up a considerable amount of space beneath the main working surface (doesn't need much attention once its working), as well as dsp. although these types of modules are really quite light in dsp, they do start to add up in large numbers.
/me proofread his post before posting
/end long post mode....
best.
i might be up for campaigning and publicising the idea of these around the web. but on second thought....i don't have the resources to put much effort into that at this point in time.
maybe someone else would though?
i did some more fiddling around with this idea i posted and the patch i originally envisioned it for. now i've got it implemented satisfactorily
not exactly how i wanted initially, but instead of how i initially thought i'd have to give up a feature (transposition of all sine oscillators) i've now got both working side by side, happily. it can be a bit fiddly if you only want one and not other and their full range of values. not exactly hard to do, but fiddly none the less. i have another crazzzzzy idea that could switch between these two functions, but it would be very elaborate and just plain silly for what it provides.
i did a similar sort of thing in my dis+taste_v1-1 patch with the external input routing through the filters and effects section. if one is interested to know what i'm talking about.
i posed the question in my original posting of one random signal generator/noise source -vs- multiples of these. the later (multiples) i thought would be needed to trully get differeing random values for each s&h. one would think that the same random/noise signal is going to each s&h, and this is partially true in the value and rate sense of the matter....but the nature of software modular synthesis means that nothing happens at exactly the same time (like how midi is). so there is a short delay between each signal coming from one random/noise source and each input going to each s&h.
i forget the name for this behaviour, but again it is something covered in.....guess what? the nord modular tutorials
i guess this can get one into trouble when dealing with lots of modulation massaging/routing. where one needs things to happen in a certain sequence of events (some events have priority over another, and if one knows these rules they can choose the event of actions). apparently it can be compensated for by adding small amounts of delays strategically in the patch. that is as far as my understanding goes.
for the most part these delays aren't noticeable or anything to really have to bother with.
so....one really doesn't need one random/noise source for each s&h, in order to get differening values for each. by the sheer nature of these delays you will allways have differing values being fed into the s&h's. though it is logical to conclude that these delays once everything is patched up are constant. so maybe having a source of random/noise per s&h still has some merits.
in this patch i'm just settling on one simple noise source for all 8 s&h's. just because this one part of the patch takes up a considerable amount of space beneath the main working surface (doesn't need much attention once its working), as well as dsp. although these types of modules are really quite light in dsp, they do start to add up in large numbers.
/me proofread his post before posting
/end long post mode....
best.
.more modular tinkering....
the randomicity in the example provided is NOT repeatable in presets
oh well, still cool nonetheless. i just love the sound of sinewaves. they can sound really creepy and alien when you have enough, or not.
i've not had any experience with additive synthesis really outside of this. i hear its quite an intensive process, programing something like a k5000. 128 partials (i don't think they are limited to sine waves?), with pitch and amp envelopes per partial...eh. just tuning 8 of them is quite a bit of work.
its kinda funny to think this patch i'm making exists as a SINGLE compact module for the most part in the nord modular, eh. that might be a good addition to add to the "things to add to the cw modular" thread.
the randomicity in the example provided is NOT repeatable in presets
oh well, still cool nonetheless. i just love the sound of sinewaves. they can sound really creepy and alien when you have enough, or not.
i've not had any experience with additive synthesis really outside of this. i hear its quite an intensive process, programing something like a k5000. 128 partials (i don't think they are limited to sine waves?), with pitch and amp envelopes per partial...eh. just tuning 8 of them is quite a bit of work.
its kinda funny to think this patch i'm making exists as a SINGLE compact module for the most part in the nord modular, eh. that might be a good addition to add to the "things to add to the cw modular" thread.
...for whatever reason? eh. i don't think i can keep up this pace much longer.