Is it possible to do this...

Andrew C
2008-12-10
2013-05-23
  • Andrew C

    Andrew C - 2008-12-10

    Hey Nick, just wondering,

    Is it possible to recreate Jens Johannson's lead sound using one of the emulations in bristol?

    For instance, do any of the emulations have an assignable LFO?

    http://www.progsounds.com/2005/08/21/jens-johansson-lead/

    Andrew

     
    • Nick Copeland

      Nick Copeland - 2008-12-15

      Hi Andrew,

      From the text the suggest that this was done with a Poly-6. Bristol emulates one of these (-poly) so you should have the mix of saw and PWM square available. The 'Assignable LFO' is just an LFO with a route selector (I think), which is in the emulator, and judging by the text this could well have been the Polysix which means this is the only kind of LFO it had anyway.

      The filter is advised as a 4pole lowpass with some resonance and it is driven by the envelope to give the cutoff some touch sensitivity. The current bristol filter is quite a warm 4 pole resonant filter.

      The key envelope is suggested to be fast attack, full sustain, and some decay (add a little decay to give the patch some atmosphere). If he did use the Poly-6 this is unlikely, he probably just used the envelope bypass switch for the amp. The original only had one envelope and if it was sent to the filter it was often the case that the amp would be driven by a gate rather than the single envelope. The Bristol envelope here should be touch responsive such that the harder you play the more the filter will open up.

      Now the bristol poly-6 does all this so you could have a go. I am not sure if the Bristol hard sync will do what you are after, it is there bu its something I have been working on again lately to strengthen it. Since it is digital it might flutter a little since the trigger for a resampled oscillator tends to move slightly rather than be right on a sample.

      The bristol also has two envelopes, one for the filter and one for the amp. It has a gate bypass option as well so that it can mimic the original synth.

      Another point discussed was to use mono mode to trigger all six voices at the same time. You will have to see how that works for you, its the 'Mono' button on the emulation however the transient CPU requirements triggering all six voices may give you overruns with the default filter - try with the -lwf option, it will save some CPU cycles at the expense of warmth.

      If you want to use the on-board effects then try settings 1 and 2 at low intensity. It looks like Jens used a distortion which was not a part of the synth. Perhaps see about chaining bristol into rakarak with Jack?

      It would be interesting to hear how you get on. Any chance of posting the resulting memory file?

      Regards,

      Nick.

       
    • Andrew C

      Andrew C - 2008-12-15

      I seem to kinda be getting a bit close to the sound, but this part leaves me cold:

      Now choose an LFO with a sine waveform, set its range around 0.10Hz (very slow) and assing it to Pulse PWM.

      while i'm setting the frequency, delay and level low and and assigning it to VCO, it appears I need to make the LFO LFO-ise the PWM?

       
    • Nick Copeland

      Nick Copeland - 2008-12-16

      Hi Andrew,

      The Polysix had a single oscillator which generates different waveforms plus a suboscillator.

      The fatness of the sound came in part from its use of two LFO. You have one LFO section with rate/delay/level which can drive the VCO (vibrato), VCF (wah wah) or VCA (tremelo). There is a second LFO dedicated to PWM - this are the PW and Rate settings of the VCO - select waveform 4, a pulse wave, and then apply PW modulation as you need it. You probably then want to select a subosc setting of 1 or 2 to fatten the sound out.

      Both of the LFO are the same from Bristol's perspective, so both will give you a range of about 0.1 to 50 Hz however the PWM LFO has a fixed waveform, the 'assignable LFO' has a selection of waveforms.

      Bristol implements the subosc as a separate oscillator however its tuning should be 0 cents of the main VCO. Even if you add some '-detune' to the emulation it should not separate the subosc from the main osc, this was intentional since I am sure the Polysix only had a single VCO with a logical divider circuit for the sub frequencies (the subosc waveforms are square waves which is what would come out of a logical divider circuit). I could add in some detune between them, that would fatten the sound out however you should get similar results from the FX setting 1 or 2 at low intensities.

      Regards,

      Nick.

       
    • Andrew C

      Andrew C - 2008-12-16

      Wow, thanks for that eye opener! Maybe I should read the manual of the polysix. I thought that the freq under the VCO was some flanger thingy, and I was like 'WOOT FLANGE!' when actually it is the PWM LFO! :P

      Also, I take it that, if I were to use the -detune option at startup, both the main oscillator AND the sub-oscillator would be detuned by the same amount?

      I'm considering assigning the first LFO (that you cannot turn off?!!) to the VCF, this gives the sound a nice sort of effect.

      Also, the sound sounds incredibly weak in the higher octaves. Though this may be remedied by just using the sub 2.

      Anyway, I'll tweak it some more tommorrow. :-D

      Thanks for the help Nick!

      Andrew

       
    • Nick Copeland

      Nick Copeland - 2008-12-16

      Hi Andy,

      The synth has a preset sequence for most assigments:

      VCO->Filter->Amp->FX

      The audio signal always passes through this sequence. The LFO can be sent to any of these to modulate them, there is an envelope each for the filter and amplifier but you can bypass the envelope on the ampliifier (with the GATE switch) and you can adjust the amount of filter envelope to the filter (with the ENV setting in the filter).

      Yes, if you use -detune it affects the main and sub oscillators alike. What it does is detune each voice, or rather note you play (only slightly if you use values between 50 and 100) and this acts like temperature sensitivity that affected analogue synths where different voices would drift by slightly different amounts.

      If the sound is weak in the higher octaves then you might want to increase the filter KBD setting, this allows the filter to track the note you play and open up at higher frequencies automatically. Also, if you have a high 'ENV' setting on the filter then the filter frequency also depends on how hard you hit the keys since it is a velocity tracking envelope, that can add expression to your playing style.

      Regards,

      Nick

       
    • Andrew C

      Andrew C - 2008-12-16

      Just out of interest, how does the filter envelope affect the filter?

      I don't really understand how it can affect attack, decay, sustain or release on a filter?

      Maybe I need to get a 'synthesis for dummies' book. :-P

       
    • Nick Copeland

      Nick Copeland - 2008-12-17

      The filter has a frequency at which it 'cuts off' the sound, anything above this frequency is filtered out. Try moving the cutoff frequency from a low value to a high value and back with the mouse. You should hear a kind of 'wah-wah' effect. If you send the LFO to the filter and give it a high level you should get a sound similar to an automatic wah pedal. You can emphasise the sound by using extra Res on the filter.

      If you do this with an envelope generator you can have the filter open up as you hit the key the drop back to the sustain level for each note - the envelope contour generated here affects the filter cutoff the same way as the LFO. The thing about the envelope is that it is touch sensitive which means the harder you play the stronger the envelope and hence the more affect it can have on the filter and amplifier.

      Try lowering the filter cutoff to near zero, put the res at 50 %, Filter Env at 75% and pressing keys. To hear the effect put the Env attack at 50% (watch out here, there are two envelopes and only one of them goes to the filter).

      Regards, nick.

       
    • Andrew C

      Andrew C - 2008-12-17

      Well, here you go, my wannabe Jens Johannson Lead for the Bristol polysix.

      http://moose.linux-coders.org/poly22.mem

      Maybe I should've used the PWM waveform instead of PW. And taken off a lot of the modulation of the PW wave. One thing that bugs me though, is why there is a sort of 'glide' when you hold a key down. :-(

       
    • Nick Copeland

      Nick Copeland - 2008-12-17

      Hi Andrew,

      I have downloaded the file and will have a look into the setup. Glide can be caused by a couple of things, one of them is literally the Glide setting - each new key will glide up at this rate from the prevous key, any kind of fretless stringed instrument can do with moving your finger along a string whilst bowing, trombones do, etc. You can get rid of this by moving the glide dial to zero.

      Its not the only possibility - if you send the envelope to the filter (ie, configure some ENV level in the filter section) and give the attack value of about a quarter turn it can also sound like glide as the filter opens.

      You might also get it with LFO to the VCO: the LFO should be synchronised to a keypress (although I would have to check that in the code) and so each key will start the LFO from a fixed position giving a constant impression of glide.

      Kind regards,

      Nick.

       
    • Nick Copeland

      Nick Copeland - 2008-12-17

      Hi Again,

      Nice patch. The 'glide' is a bug - the emulator will mix to the VCO either if you have Depth configured via the VCO or if you have the modwheel above zero. That is wrong, the modwheel should control the depth of the LFO wherever it goes. The next release will have a fix, for now you will have to put the modwheel to zero but you really need that fix.

      Regarding the use of PW vs PWM you obviously followed your ear. The GUI shows a pulse and then a PWM wave however the engine has them in the other order. That is also a bug I suppose and is now fixed. Either way you have actually chosen PWM and it gives the patch a nice sound.

      Can I distribute this will the next release of Bristol?

      If you want some patched software let me know, I can build and upload a version pretty quickly.

      Regards,

      Nick.

       
    • Andrew C

      Andrew C - 2008-12-17

      Oh right, so you're saying that the LFO depth being controlled from the mod-wheel was always being assigned to the VCO and not following the LFO around to, say the VCF or the VCA?

      I picked the 3rd waveform (now truely PW i suppose) because it gave the sound a more 'fuller' sound which was precisely what I was looking for.

      Also, Of course you can distribute the patch with the next release of Bristol. I hope you'll also remember to edit my patch to go to the 4th waveform, now that it's PWM. :-D

      <Also, if you could perhaps upload maybe a diff against 0.30.1 or just a 0.30.1.1 with those changes I wouldn't mind. Still, no rush for you to do it.>

      EDIT: I just noticed you have 0.30.2 and I have 0.30.1. No harm done, I'll just update to .2.

      :-D

      Thanks a lot for the help and I really appreciate that you're this responsive to the questions and comments of your userbase.

      Andrew Coughlan.

       
    • Nick Copeland

      Nick Copeland - 2008-12-17

      Hi Andrew,

      Not sure if I can make you a diff because I am reworking the PWM code (it should not affect your patch, I just want to make PWM a float rather than int operation) and it affects a few files. Also, I am probably going to release 0.30.3 shortly - before the 25th either way, get it out before christmas and all that - and it will have the fixes plus your memory with the waveform selection changed. Will add your name to the contribution page also, I really like having people donate things back into the release.

      Also, there was another bug that came to light thanks to your work - if you configure a sound to be 'Mono' it will play all voices on the same note monophonicaly. It makes a very rich sound and your patch actually improves (in my opinion) with this option, with a bit of -detune the sound expands very nicely. Unfortunately it is not reconfigured when a memory is reloaded so the option look like it is set but it isn't, it just inherits the previous setting. Fixed now.

      So that sounds like three fixes for one cool patch. That is quite a lot of fixes but from my side I am happy since the code is now cleaner.

      Very best regards,

      Nick.

       
    • Andrew C

      Andrew C - 2008-12-18

      Yes, I originally created the patch using the mono switch on.

      Any chance of the mono bug being resolved before the .3 release (Though i wouldn't be holding my breath for this one)/or maybe even the .4 release? Just wondering. :)

       
    • Andrew C

      Andrew C - 2008-12-18

      Also, do you have any recommendation on what -detune value to use to expand the sound even more?

       
    • Nick Copeland

      Nick Copeland - 2008-12-19

      Hi Andrew,

      I don't think using a value bigger than 100 will work very well, the detune starts to sound occasionally false - a lot of these old synths did suffer from this which is why the parameter has been introduced however there were reasons why the old beast usually had a 'Tune' button to straighten it all out. You might want to try with 150, the Mono mode may mask the differences in frequency a liittle.

      There are quite a few changes in the code for the Polysix now and I have one last issue to clear up before the code is uploaded - the oscillator sync code had to change since it was broken for square waves. There is one last oscillator that is giving problems but I think code will go out on Monday either way, sooner if I wrap this issue up. It will contain all these fixes.

      Regards,

      Nick.

       
    • Nick Copeland

      Nick Copeland - 2008-12-19

      Release 0.30.3 has been uploaded, I cleaned up the last oscillator for PWM and Sync and uploaded it. All the fixes you were after should be there.

      Regards, Nick.

       
    • Andrew C

      Andrew C - 2009-03-14

      Sorry to revive a dead thread, but as you say the 'The GUI shows a pulse and then a PWM wave however the engine has them in the other order', so am I to take it that the order of the four waveforms is now: Triangle, Saw, Pulse, Pulse Width?

      Thanks.

      Andrew.

       

Log in to post a comment.

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:





No, thanks