A cleaned up version of the 303 frame...
It still has the quirky slide section..maybe i do a even more stripped version...
Its only important to have it this way to transfer lines from a real 303..of cause it would be possible to do any kind of seq array aslong we have a gate time..slide activation...accent activation and of cause pitch information.
This is called a frame to allow to keep soundengines and sequencers as seperated modules to develop something that has a max performance factor
while still having resources on the G2 left...
I also include a maxed version of the 303 sequencer that allows aswell 303 style writing with pitch mode and real time note insert aswell as having a step seqencer for pitches that can be shifted in time or reversed and punched into the main pitch sequence...
Even when this extra functions allow faster sequence creating and experimentation they are way too expensiv to have live a decent polyphony left for other tasks than 303 emulation...
Therefore i think the big version is a sequence development tool while the small one is for performance versions of the patch...
Its possible to parameter paste info between the main seuqencers.
still try to avoid the real work in rebuilding the soundengine
and as a last add beforer going to the real questions..
The 303 frame 05 with the big sequencer ...for the lazy ones
Btw:
the information about boosting LF on accented notes appears to be accurat and important.
And one more of the real background infos to master this challenge...
this 3 audiofiles show impact of the accent to the filter.
again a devilfish that allows settings that are not possible with the standard 303.
Waveforms before filter look close to what they should be..exept that the sqare is far from perfect and is varying in pw over the tone range of the 303.
Filter knob is full up... saw...no resonance... no env mod...
gatetime one quarter note
no accent and full accent and full accent with long decay...
its interesting that is the long acc decay note the event is longer than the gatetime...
Ok here an important one that might give an bigger insight why all simulation aproaches fail at one point...
listen to that... Has the 303 2 filters? or how is it possible that we still hear the base tone?
This is recorded directly out of the filter...we hear the accent sweep circuit working, on the devilfishs self oscilation mode...a higher resonance setting than possible on the standard 303...
what is going on here? you hear the strong sustained tone? this is the oscilator...the rest is what the accent sweep circuit makes out of it..
in one as it seems totally independend filterpath !
I think this recording is without envelop mod with a medium accent decay setting...
rec with envelop mod later
Any opinions?
accsweepBoostRes.wav
Description:
Filesize:
986.86 KB
Viewed:
2605 Time(s)
Last edited by 3phase on Wed Mar 08, 2006 4:51 pm; edited 1 time in total
its interesting that is the long acc decay note the event is longer than the gatetime...
how is that possible?...
The 303 mod env is fixed to 200ms for accented notes, devilfish adds a pot so you can have long accent decay. When accent signal is on, mod env is routed via accent amount pot to both vcf and vca. I think.
On the 303, the accent signal is only connected to two CMOS switches. The first routes the mod env to the top of the accent level pot, the second shorts out (or disables) the env.decay pot, forcing the decay time to minimum (200ms).
hmmm..looks like...and this is an important detail..
because all statements are against the present of the mod envelope to the vca control.
this might be the part that adds lowend to accented notes as the guy with the impulse responses was stating...
I think its important to uncover this little things...
they sum up to the difference that made the 303 allways better than the clone attempts.
Especially the response to the accents is one of the points where clones usually loose against the original...it just dont sounds as sexy as with a real 303
btw: thanks your input is very welcome
Even when most people think the 303 is a lame item...
The difficulty patching it shows thats not yet fully discovered.
A pro one was easy to patch... I wonder if Sheridan is still around ..
He was/is brilliant in modelling natural sounds.. I wonder how he would do a 303... But he would also just get the static moment...without the real structure you never will cover the range of possible sounds of a 303...
and this is the special magic about the machine... allmost any knobsetting is usefull and directly communicates with the others...
Yes its definitely connected to the VCA, through R119, R120 and D27. Also the wiper of the resonance pot that feeds back into the filter is also summed into the VCA input, so with increase resonance, amount of filter to vca increases.
Your sample posts are interesting, a lot more clipping on the saw than I expected. It is almost half wave rectified. They also show time constant of resonance lag on env.mod.
The mod env decay curve, as it effects Filter is expo but not quite. Interaction between the shape of the basic env and the (very simplistic) expo converter on the filter. It seems to morph from expo to linear over time.
Yes its definitely connected to the VCA, through R119, R120 and D27. Also the wiper of the resonance pot that feeds back into the filter is also summed into the VCA input, so with increase resonance, amount of filter to vca increases.
Your sample posts are interesting, a lot more clipping on the saw than I expected. It is almost half wave rectified. They also show time constant of resonance lag on env.mod.
The mod env decay curve, as it effects Filter is expo but not quite. Interaction between the shape of the basic env and the (very simplistic) expo converter on the filter. It seems to morph from expo two linear over time.
huhh.. i see, you see a lot
does that say that we have increasing impact of the resonating filter signal on the amplitude modulation of the vca?
And other question..because thats where i am just working at...
the last audio sample of the full resonating filter with accent sweep action..
what do you think about that? usually the filters i know dont let pass much signal when they are resonating so heavy...
in this example it sounds like having the original wave without resonance upfront..sounds like to seperated signals to me...
any ideas how this is caused? cant beliefe that roland spend for 2 independent filters...
3phase, I can't wait to hear your 303 emulations ... RedOne will be unpacked tonight !!!
/Dasz
I havent posted a new emulation try yet...just the frame and a few starting steps... i havent reached the state of modelling the sound yet..
still collecting informations..with results
just the enevelop with dynamic decay time and the lowend boost can bring it much further to exitement.. and i dont know yet how the things G2ian is suggesting might work. Its good to be able to verify this against the real 303 next to me...that is sadly no devil fish...everything what makes it closer is correct.. Therefore result of the last days..
The lowend needs to be boosted an accented notes..its defenetly happening on the 303...you just dont realize because its rather subtile but gives the accents more 3d presence.. next version will feature that..
And Ians add shows me a better way than crossfading between eq´s..
this is never part of a real 303...that the lacking mod envelope has a different freq response / less lowcut than the direct signal seems much more likely to me.. I ve to try it ..
Its necessary to get the complete picture before i start and than finetuning based on a wrong model again...
The 2004 version had the problem that tuning in one direction was loosing on another...a proof that the structure was wrong. Last edited by 3phase on Wed Mar 08, 2006 5:39 pm; edited 1 time in total
Your sample posts are interesting, a lot more clipping on the saw than I expected. It is almost half wave rectified.
do you think this is clipping? i was overdriving the filter far beyond this can happen on a 303 and the waveform actually got in shape.
in my earlier emulation attempts i had the same waveform just by filtering..lowcut and mid boost around 450 hz was part of it...
So i was thinking that the funny waveform is more a result of unlinear amplification..
is it possible that certain circuits that bend the freq response get a flatter response when driven hard?
does that say that we have increasing impact of the resonating filter signal on the amplitude modulation of the vca?
No, the first gang on the resonance pot (the one that controlls the amount of feedback to the inverting input of the filter) is also connected to the VCA audio input (R121,C21)in parallel to the normal vcf -> vca connection. This compensates for the bottom end loss of the moog ladder at high resonance. Also might explain that last sample.
I see both vca inputs go through a cap which might explain the HPF look to the basic waveforms.
do you think this is clipping? i was overdriving the filter far beyond this can happen on a 303 and the waveform actually got in shape.
in my earlier emulation attempts i had the same waveform just by filtering..lowcut and mid boost around 450 hz was part of it...
So i was thinking that the funny waveform is more a result of unlinear amplification..
Yes your right, just low cut or HPF makes that shape. It matches other real 303 samples I have.
Joined: Dec 23, 2004 Posts: 14 Location: MD USA
G2 patch files: 11
Posted: Thu Mar 09, 2006 5:09 am Post subject:
303 Model
Guys - I'm travelling for the next week - can't wait to hear how far you guys get in that time.
The patches I posted cut the VCA to close to 200ms on accents, but I think there is some interaction on slides above 200ms - I could be wrong.
There are some brilliant minds out there, I'm sure collectively we can get close to the real thing - unfortunately I don't have the money for the real thing - so the samples are really useful.
Posted: Thu Mar 09, 2006 8:56 am Post subject:
Re: New 303 model
8BitVertigo wrote:
3 phase
Have a go on this one - I haven't tried to start from scratch, but this sounds pretty good. I remodelled the filter (borrowed heavily from a filter Rob Hordjik had posted) and smoothed out the accent sweep envelope. There's a couple of variations to play with, screams something along the lines of the real thing.
Let me know what you think.
this one truly walks like a duck and quacks like a duck!
most fun i've had tweaking a 303 emulation in years... that sound!!
Joined: Dec 23, 2004 Posts: 14 Location: MD USA
G2 patch files: 11
Posted: Thu Mar 09, 2006 9:01 am Post subject:
303 Emulation
I've had a few late nights myself - hours lost tweaking this one as well.....kind of wish I had the real thing but keep telling myself these magic red boxes must be capable of getting close.......
Sven has a good point about the MIDI not being up to the task for controlling the voice slot. This version uses two bus lines, one for pitch, the other carries the gate, accent and glide control using a DA/AD to encode/decode the separate signals.
The voice module in slot A is a little simpler as Morph groups seem to make more sense emulating the more complicated interactions. The mod envelope has a shaper, so that the curve rolls off closer to the shape from the samples. The filter has the same routing of the resonance feedback to the VCA input as shown on the schematic.
Only three variations, turn the knobs on the parameter page.
The envelope sounds indeed more like it than just a plain one..
I will try it ...want to do some experiments now with
the enevelope and the negative offset...
i wonder if the envelope for the filter mod might be reversed in control voltage? i ve to cut more audio examples... but its interesting that increasing envelope depth on accented notes, the ones with the highest filter mod..only lowers the frequencie the filter falls to...
The only thing that makes the filter go even higher is the accent sweep circuit...
I just got a printed service manual...shit...2:30 allready here...a bit late to start patching...but i think its my turn again to post one version
the filter looks shit complicated..this two traces to the vca from the filter......
when the resonace is off one of them goes to ground while when its up booth traces are like one... the value of the coupling caps doubles this way..
is that only for that reason? puhhh... what a miracle box...
Ok ..i try to give it some more research today..but in the moment i doubt that i really have the time to get thru with this... Threfore i post a few more important audio snippets that show what is going on and might help others that are in research of the item.
The behaviour of the EnvMod Knob of the 303 is the most important.
As i stated before the unique point about this envelope mod is that it sounds somehow very ruberish... I thought the decay times get longer when its up..
But by analyzing this audio samples its even more complicated...
As more up the EnvMod pot is the lower the frequencie gets the env falls to.
This indeed has the fx that depper modulations appear to have a longer decay... as you can see with the samples its almost twice as long...
But..Close to the max position the filter frequencie the envelope falls to gets so low that the decay shortens again....
Explains pretty good why the EnvMod knob on clones appears to be somehow dead in relation to what is going on on a real 303...
I ve to experiment with this again now.. I am still a bit curious if the envmod knob really just gives a negativ offset...Or if this is rather a function to the envelope that reacts dynamic during the sweep...
One more 303 frame..this time a more economic version. But still trys to aim to behave like a 303.. Its possible that the final gets even more minimal.
But i think this one is a good development frame now because its not too overweighted but still has all 303 functionalitys and is prepared to hold extra sound fx within the 303 slot... or to use the keyboard for other things while having focus on the 303 controls...only thing missing...the original 303 sound..but its getting closer.
while the sequence is running the keyboard is adressed to an extra sound module that can be replace with midimodules to drive external sources.
when you press button A1/8 on slot C the machine gets into the 303 selct mode. now the left octave on the keyboard is doing the transpose while the right octave is selecting patterns. Octave shift in the middle...
Its still buggy because in the moment you go into the 303 edit mode the last keypressed is still present and interfears with the 303 control...
Posted: Thu Mar 23, 2006 1:15 pm Post subject:
Final 303 Frame
Edit..Sorry for confusion... I post another version as 303 Frame 1.00
as a clean version that just 100% has 303 functionality without control gimmicks..
It was necessary to redo it because the real 303 resets any transpose commands on patternchange... somehow important when you want to do that same on a G2 as you would do it on a 303.
This emulation effort is for me mainly to replace the real 303 on stage...
therfore it need the same behavior...waste of resources or not.
I turned the conception back to all controls via midi remote..
The yellow modules in the FX area slot C...
I placed a 303 dummy in the VA aswell...as a reference..but later this will be the place where the final engine will be placed aswell.
The VA of slot C is prepared to recive the seq information from slot D aswell, so its a development area for the 303 emulation...
But can be later used as a place for a independent polyphonic synth sound..as its done here just to demo it...
In this version you can change patterns with knob 1B5 or you switch knob 1B8 to 303control on..than the keyboard is removed from the VA and acts as pattern selector (octave 3) or Transpose (octave 1)
Booth functions are quantized to the syncpuls that is defined in the corosponding seq variation in slot D...
so...done...
next step is to get a better filter and the envelope balistics wright...as well as overworking the structure following the 303 schematics...
Joined: Dec 23, 2004 Posts: 14 Location: MD USA
G2 patch files: 11
Posted: Fri Mar 24, 2006 7:15 pm Post subject:
303 - W/3 Phase Accent Sweep
Added the Accent Sweep circuit 3 phase provided - nice one mate - works a treat. Not sure if I'm using it entirely correctly but it sounds like it's doing the trick.
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You can download files in this forum
Forum index » Clavia Nord Modular » Nord Modular G2 Discussion
FAQ
Calendar
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
Chat Room
your input is very welcome
