dis+taste_v1

castol · Post by **castol** » Thu Feb 14, 2002 8:50 am

<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> <a name="planetz-file"></a><a href="http://www.planetz.com/Pulsar/files/mod ... .mp3"><img src="/forums/images/listen_icon.gif" border="0" alt=" Sound Example"> Sound Example</a> <a name="planetz-file"></a><a href="http://www.planetz.com/Pulsar/files/mod ... .txt"><img src="/forums/images/notes_icon.gif" border="0" alt=" Notes"> Notes</a> <a name="planetz-tag"></a>Type: Synth <a name="planetz-tag"></a>Pulsar Version: Pulsar 2.x <a name="planetz-tag"></a>Requires: Modular 2 _____________________________________ -+-dis+taste v.1-+-

dis+taste is me finally getting around to doing something with some samples of the raw
waveforms sampled from my ex sidstation. i try and borrow some features, implement
others, add new ones, or expand on them. inevitably i leave out more than a few. i think
this patch can sound similar to a real sidstation, if not even more dirty and racous.
mostly through the use of many stages of gain, a pre/post eq distortion, 2 bit
quanitizers and a frequency rate reduction. though not all used at the same time or at
full strength or mix! in this patch i got away from hooking things up to each other daisy
chain style but instead decided to try mixing signals out to mixers and combine an
effected sound with the orignal dry sound. this proves to be much more flexible. i also
implement the switches more than a few times to select differening outputs on modules
which have more than one. i figured out how to dc offset a signal, positive or negatively
and used a tip i found on the planet z site for inverting a signal in the lfos for this
patch. its a start.

i will outline what the features are and how things are routed, layed out. i will
enevitably leave some things out. this is a big patch, which is one reason why i feel the
need to go into detail about it with this readme. i had thoughts when creating this that
only myself would be able to do anything with this hulking patch, that was layed out
poorly and whose interconnection scheme was not repeatable. i straightened things
up,adopted a top down approach and grouped similar things together arranged in a manner
that corresponded to how and and what they were hooked up to. i have gone through it many
times and i am pretty sure everything is as it should be, working and routed in the
manner i describe. working with such a large patch demands an efficient use of the cable
hide and solo features of the mod2 (something i didn't use very well before this). i
found ben walkers tip on changing the colors of the cables very handy for trouble
shooting circuits.

-+-main attractions-+-

- 3 sample oscillators meant to play one of the 5 included single cycle sidtation
waveforms looped and sampled from a sidstation. these waveforms are Pulse, Saw, Triangle,
Mixed and Noise (not a single cycle). there is no portamento, oscillator syncing, or
pulse width modulation. in theory, this could be done (i believe), but not within the
modular2. unless creamware or a 3rd party built modules with these features.

all waves were sampled at c4, but for whatever reason they don't all play back at exactly
the same pitch. this leads to the oscillators slowly modulating over time when they are
tuned to each other. this behavior does not bother me. i just viewit as another subtle or
drastic form of modulation. there is a sample pool in the middle of the 4th column,
positioned so that it is possible to view each oscillator and still see all the samples
in the pool.

- 1 sample oscilator which is meant to play back a looped sample of the inherent ambient
noise the sidstation constantly generates. turn it down if you like. it has its own
seperate adsr b envelope, and is mixed together with a mix of the other signals output
before going to another linear vca being modulated by one last envelope which happens to
be a multi-segment a.

- each oscillator (minus the ambient noise oscillator) is fed into a 12db gain module,
for gain make up, adding distortion or clipping. from here it is output to the the
corresponding filter input. there are lots of places in this patch to increase
gain...mmm.

- 3 midi-syncable multi lfo's routed together so as they can be cross-modulated with any
of the other two lfos. each lfo has two supplementary circuits. one limits modulation to
only positive or negative or none such (i believe this is called dc offset). this circuit
is fed the output of its corresponding lfo and outputs ONLY to the pitch modifier for the
corresponding oscillator. this circuit is then connected to an inverting circuit, which
has the effect of flipping the waveform upside down. 2 signals are outputed from here to
the other 2 lfos mod inputs. this enables the ability to modulate an lfo with a slightly
different variation on the source lfo (inverted).

each lfos 2 mod inputs are fed a top down collection of the other 2 lfos. so lfo 1s first
mod input would be from the 2nd lfo, the second input would be from the 3rd lfo. the
second lfo's first mod input would be from the 3rd lfo, the 2nd input would be from lfo
1. finally, lfo 3 would have lfo 1 in the first input and lfo 2 in the second.

you can delay, fade in or out each lfo, and control whether it will restart on key down
or not.

the output of the lfo after running throug a dc offseter is fed into a pitch modifier. i
found that the built in pitch mod input and depth control on the sample oscillator was
pretty mild. using the pitch modifier one has much more gain over the signal which can
lead to nice, drastic sputtering pitch modulation when adjusted past half way.

- 9 dadsr multi-segment envelopes. thats delay, attack, decay, sustain, and release. the
multisegment envelopes have the added 4 powerfull features of being able to define 255
breakpoints, loop between 2 points, define the behavior of each breakpoint and also
modulate negatively and positively. these are very powerfull, in comparison to the
sidstation envelopes, which can only invert and delay. the times on the sidstation
envelopes are realatively coarse. here are the values borrowed from the manual.

Value Attack Rate Decay/Release Rate
0 2 ms 6 ms
1 8 ms 24 ms
2 16 ms 48 ms
3 24 ms 72 ms
4 38 ms 114 ms
5 56 ms 168 ms
6 68 ms 204 ms
7 80 ms 240 ms
8 100 ms 300 ms
9 240 ms 750 ms
10 800 ms 2.4 s
12 1 s 3 s
13 3 s 9 s
14 5 s 15 s
15 8 s 24 s

- each oscillator can be ring modulated by any of the other 2 oscillators or by itself. i
have the ring modulators on a switch so that either they are all actively in the audio
chain, or they are not. 6, 4x1 switches with gain are used for routing signals into the 3
ring modulators. ring modulation is post-filter. each 4x1 module is fed the first, second
and 3rd oscillators, from the corresponding filter switch outputs. the switches are as
they are labeled. 1 is the 1st oscillator/filter group, 2 is the second, 3 is the 3rd.
the first 2 modules go into the first ring modulator. the second two go to the second.
the third two go the the third. the first in each group is input to the left input on the
ring modulator. the second in a group goes to the right input in that same ring
modulator.

each ring modulator output is sent to single channel of a 4 channel mixer. so ring
modulator 1 goes to mixer input 1, ring mod 2 goes to mixer input 2....this can this can
lead to some confusion if say you have oscillators 2-3 ring modulating through the first
ring modulator.

the output of this mixer is fed into 2 daisy chained 12db gain modules, more gain...what
can i say. this is more for make up, as ring modulating i found can tend to lower volume
in some instances.

when ring modulation is switched off, the dry signals, post-filter, are fed into a second
4 channel mixer.

- 3 multi-mode filter b's. each can be controlled by a single dedicated multi-segment
envelope. this is all in terms of modulation for the filters. the output of each filters
filters, yes, filters filter, are fed into one of 3 4x1 switches which enable selection
of lp, bp and hp. in that order, from left to right.

the sidstations filters are very unique sounding, distort on the high or low end
depending on the mode. and have very limited resonance. i believe the filters are 12 db i
don't know for sure. there are 7! of them, and the frequency can wrap around on over or
undermodulation. the resonance is limited to 15 values, and the effect of turning it to
its max value is very subtle.

building a formant filter for each of the filter modes would yield a pretty accurate
representation of the filters i have read. it sounds like you just take the peak signals
of a source and recreate these peaks with a number of fixed filters at the same
frequencies.

there is an article online of some fellow who did this with the sidstations bandpass
filter and a k5000s formant filter. i haven't really read through it yet. sounds
interesting though.

- an effects section, no delay effects here thank you. depending on if one has ring
modulation on or off, the output from the switch which controls this is fed into a bit
quanitzer module. 2 signals are sourced from here. one goes into the decimator module.
the other goes to a the first channel of 2 channel mixer above the bit quanitizer. the
decimators output to the second channel of the two channel mixer atop the bit quanitizer.
the bit quanitizer in the decimator is redundant, though can be used to good effect. the
reason for two bit quanitizers is because i wanted to be able to mix in a small amount of
frequency rate reduction to a signal coming from a bit quanitizer. as frequency rate
reduction is a very drastic and distinct effect, having it it directly in the signal
chain didn't do it for me. a little is good, but having a full mix of if is not good
unless you really want that effect. for no good reason there isn't a seperate sample rate
reduction module.

from here the signal is routed 2 places. first into a 5 band free filter bank b. this
feeds a distortion module. the distortions two outputs are fed into a 4x1 switch, for
selection of soft or hard distortion. from this output the signal goes into another 5
band free filter bank b.

i remembered reading about distortion effects and the effectiveness of eq'ing the signal
going into and out of the distortion, when building the effects block of this patch. i
thought i'd give it a try. the eq's in the modular seem to suck the high end out of a
signal when they are at unity gain. adding gain in the high end didn't give it back. you
can leave the eq's bypassed if you like, most of my intial patches have them bypassed.
there will be 5 or 6 different presets of frequency setttings for each eq channel. from
here the signal is fed into, another 12db gain module. for further increased overdriving
of an allready distorted signal. the output from here goes to the second channel of a 2
channel mixer, the first channel is fed from the output of that 2 channel mixer on top of
the bit quanitizer.

the output of this mixer feeds the second channel of a 2 channel mixer underneath the
sample pool. channel one is sourced from the vca for the sid ambient noise oscillator way
in the bottom of the patch panel. from here the signal goes into a final output vca
modulated by the a multi-segment a envelope. and from this vca the signal goes to the
output 1 of the audioout module.

take a look at the included asci diagram of the module layout. it isn't super detailed
but should give a decent idea of what something does and where it gets its input and
output from.

------

i've noticed on certain settings the oscillators will bleed through to the output even
though their vca is closed and the mixers all "seem" closed. i think this is because of
having 2+ mixers of dry signals in the effects block. it kinda suprised me when i first
noticed it. in fact the sidstation has this quirk itself. whereby the oscillators allways
bleed to the output in a muted tone.

there are only a handfull of presets i've created thusfar. maybe they aren't the most
musical, but i think they sound pretty raucous and varied. i will post up some more in
the near future.

i have ideas i would and may add in the future to this patch. it needs some ruminating
time. if and when so i will post an updated patch, note what was done or implemented and
to what purpose, or effect.

i would appreciate any feedback left. you can do so here , or email me at
amfreq@hotmail.com. in the spirit of things since i'm asking for feedback i will leave
some for others.

though now that i have come to find out i can't load version 3 patches in version 2 of
the modular i won't be able to try out many new patches until i upgrade. maybe i'll just
leave feedback on older some older patches.....anyway. enjoy.

//c

[ This Message was edited by: castol on 2002-03-01 14:56 ]

[ This Message was edited by: castol on 2002-03-19 05:48 ]

[ This Message was edited by: castol on 2002-04-17 11:48 ] <a name="planetz-fileimage"></a><IMG SRC="http://www.planetz.com/Pulsar/files/mod ... o_view.jpg" BORDER="0">

Ben Walker · Post by **Ben Walker** » Fri Feb 15, 2002 4:56 am

Hi,
I haven't even tried the patch yet (can't til I get home tonight) but I just wanted to say thanks for the detail you've put into the readme. It always helps me to understand a patch if people take the time to document it properly - If I'm going to do more than just play with other peoples presets I want to have an idea of what a patch does!

I have a big sample based patch on the way myself - its pretty much done - just writing presets and documentation now.

Did you find any problems working with the sample pool? I found a few weird things, but these might be new to v3.

My patch is much more straightforward in concept to yours, but its quite a biggy.

You say 'No delay effects here please' - what's your thinking? Just that they're overused? Myself, I can't get enough of delay, both subtle and extrovert!

Anyway, more feedback once I've tried the patch.
Cheers,
Ben

castol · Post by **castol** » Fri Feb 15, 2002 9:29 am

hello ben,

thanks for your reply and excellent website.

Did you find any problems working with the sample pool? I found a few weird things, but these might be new to v3.

yes i did! i initially had problems with it recognizing my samples as valid files. according to subhuman this is a known problem with programs that write a header to a soundfile. for whatever reason the files have to be headerless.

i'm on a mac. using bias peak i was unable to load files saved from here. as this program allways adds header information into the file. the header information is stuff like sample rate, channels, encoding method and the length in samples.

using sound app to save the files i was able to load the files into the pool. these files consequently have NO header information.

thanks to subhuman for informing me of this. he actually said "extra" header information, but from what i found out above using soundhack (which allows you to view and edit header information), ANY header is enough to make the file unloadable.

on the pc i think subhuman said cooledit has the option of writting headers or not. the thread is in the problem solving section

another thing i've noticed is the file names in the pool look fine, as they should be but when i drag these samples into a sample oscillator the names are expressed as a combination of numbers and the name of the partion my luna folder is on. this behavior would bother me if i was using a large pool of samples, but with only a small number it doesn't bother me much. this behavior is the same if i drag the files from the file menu, i don't think i've tried "loading" files but i imagine its the same.

You say 'No delay effects here please' - what's your thinking? Just that they're overused? Myself, I can't get enough of delay, both subtle and extrovert!

when i said that, i was more just saying it from my point of view as if someone asked if i wanted "gravy" for my freshly steamed vegetable and brown basmati long grain rice rice dish (seasoned ever so delicately with freshly squeezed lime juice) after seeing how "plain" it was. its tempting to smother something so plain in a good sauce (even worse, a bad one), but then you lose a lot of the subtle flavor from it being rather plain (which i happen to enjoy). sure you could use something not so smoothering and taste smearing, but for sake of example i'm just using gravy which i tend to associate with delay effects. gravy is pretty well thought of as having a blanket effect on something.

in synthesis it becomes easy to build a lackluster patch and slap some "gravy" on top of it to make it sound intersting. i have been guilty of this

i am just trying to deviate from as i see it an easy default manner of doing things, finishing them up. limitation.

i'm not saying that effects don't have their place. they can not add a lot to a sound or provide that much more sonic potential and complexity. that one should try and program something to imitate say a reverb, delay, phaser, or a flange at the patch level...so they "feel" like they've pushed themselves to a limie. although it can be done.

but that adding these effects i have found tend to stifle the possiblities one could program within a patch becaue they just can be inherently easier to make something sound interesting with less effort involved.

if i in anyway offend anyone or i sound like i'm talking out of of rear end then this is where i say i am not a professional. and i could very well be speculating more so, than stating facts supported by lots of real world experience. reading has that effect sometimes.

phew!

i spent WAY to much time writting that. oh well.

btw ben, i would be more than happy if you added my patch, schemo and explanation to your page. i think though i want to add some things to my description. so i'll just mail an updated explanation to your email listed here in the near future.

one last thing, what have you used to do the schematics for the patches in your patch section?

//c

djody · Post by **djody** » Fri Feb 15, 2002 9:54 am

he castol,

i tryed the patch, but with al presets there is a huge amound of noise in the mix. I guess that that is not mend to be so i think there is something wrong with it....

I almost couldn't hear nothing but noise, like there is a noise genarator full open...

I hope Ben will check it and see if it is the same with him,

Laterz Djody

maket · Post by **maket** » Fri Feb 15, 2002 9:57 am

Cool Patch! I have on all sounds noise,does it right?

castol · Post by **castol** » Fri Feb 15, 2002 10:38 am

djody and maket:

thank you for your comments.

the noise can be a part of the sound, and yes most the presets use the built in noise.

what it actually is, is a looped sample of the noise the sidstation constantly produces (very low s/n ratio). i added this part of the patch for authenticity.
since it is only a looped sample in one of the sample oscillators, has its own envelope (which is set in all the presets to have a long release) and is mixed with the other normal oscillators pretty much at the end of the signal path with a 2 channel mixer. you can elect to turn its volume down, completely off or adjust its envelope to a behavior which is preferable.

the sample oscillator and envelope for the noise is at the botom of the patch in the 3rd column across, underneath a frequency divider. the mixer i was talking about which you can control the level of the noise with, is in the 4th column across and underneath the sample pool. the first knob controls the level of the noise and the second controls the overall level of everything else.

in a real sidstation it is almost necessary to use a noise gate because of this noise. i have thought it would be cool to build something approximating this into the patch.

best.

maket · Post by **maket** » Fri Feb 15, 2002 11:31 am

Well,Then it's very cool!
It's make feeling that cyber punk is coming!

Ben Walker · Post by **Ben Walker** » Tue Feb 19, 2002 5:12 am

another thing i've noticed is the file names in the pool look fine, as they should be but when i drag these samples into a sample oscillator the names are expressed as a combination of numbers and the name of the partion my luna folder is on.

Hmm - I've never seen that before, but it certainly happens in your patch. Not sure what could be causing that - I'll let CW know about it, but I don't think this will be a high priority fix.
Other sample pool issues I've come across are
1.) Each preset you save contains the whole sample pool. If your pool is 2MB, then each preset becomes 2MB+ in size. Not very handy.
2.) Sometimes Samples just fail to load - you sometimes get a permanent 'Loading sample xxx' name in the sample module, but it never loads. Other times it says that it has loaded, but you get all sorts of dropouts and have to restart Pulsar to fix this.

You say 'No delay effects here please' - what's your thinking? Just that they're overused? Myself, I can't get enough of delay, both subtle and extrovert

...

when i said that, i was more just saying it from my point of view as if someone asked if i wanted "gravy" for my freshly steamed vegetable and brown basmati long grain rice rice dish (seasoned ever so delicately with freshly squeezed lime juice) after seeing how "plain" it was. its tempting to smother something so plain in a good sauce (even worse, a bad one), but then you lose a lot of the subtle flavor from it being rather plain (which i happen to enjoy). sure you could use something not so smoothering and taste smearing, but for sake of example i'm just using gravy which i tend to associate with delay effects. gravy is pretty well thought of as having a blanket effect on something.

in synthesis it becomes easy to build a lackluster patch and slap some "gravy" on top of it to make it sound intersting. i have been guilty of this

I know exactly what you mean about gravy - its mostly important to make sure the patch sounds good BEFORE you put the effects on, although sometimes effects can actually be a great starting point for a patch.
I have a habit of chucking chilli sauce on whatever I'm eating, so that probably explains my love of effects too!

btw ben, i would be more than happy if you added my patch, schemo and explanation to your page. i think though i want to add some things to my description. so i'll just mail an updated explanation to your email listed here in the near future.

I'd be happy to add your patch, picture, description, etc to the site. Just send me the updated explanation when you're ready.
Also, if you could send me one or two example mp3's of the patch that'd be great too.

one last thing, what have you used to do the schematics for the patches in your patch section?

For the first couple of pictures I used the line art feature of Word, but that was a real pain as arrows don't get anchored to boxes, so whenever you move a box you have to move all the lines, etc.

Now, I use VISIO, which I have a copy of at work. It's perfect for this sort of diagram - but I wouldn't want to have to pay for my own copy out of work - I think its quite expensive.

BTW - I think the noise that other people have mentioned is really too overwhelming - on all presets I listened to the noise is actually louder than any signal - often its difficult to make anything out apart from noise - I think that this can't be right, is it? Is this something else to do with the funny file loading business I wonder.
Can you send an mp3 of what a particular preset 'should' sound like, and I'll try the same preset myself?

Cheers,
Ben

_________________
<a href="http://www.modularsynth.co.uk" target="_blank">www.ModularSynth.co.uk</a>

[ This Message was edited by: Ben Walker on 2002-02-19 05:13 ]

castol · Post by **castol** » Tue Feb 19, 2002 10:16 am

i uploaded an mp3 to my original post and an accompanying text file with the specifics (got rid of that text file which was only a copy of my post).

it should put the doubts to rest.

that i intended all that horrible noise from the begining!

though as you do i have a thought that there may be some sort of translation problem between my system and others. we will see.

i hadn't planned on emailing cw about the sample name problem, but since you say you can see it in the patch i think i'll email them as well.