Synapse Audio Minimoog emulation "The Legend" for VST/AU and RE released!

VST, AU, AAX, CLAP, etc. Plugin Virtual Instruments Discussion
Post Reply New Topic
RELATED
PRODUCTS
The Legend

Post

Ingonator wrote:
Yorrrrrr wrote:
fluffy_little_something wrote:There is something about Legend that makes it sound as if it were slightly out of tune, a few cents too low, to be precise. I am not sure that is the case, but for some reason I get the impression. I compared an init patch (just one saw, full cutoff and sustain) to several other synths, which all (seem to) sound slightly higher in my view.
Anyone confirm that?
Yes, it's the oscillators keytracking. Change it to 0 (mid position) in the back panel: you will have perfect key tracking and the oscillators will be tuned perfectly across the entire keyboard. The init patch has a value of 1 by default (which shows a -4 cent detuning with the help of a tuning plugin, GTune).
This is what makes the oscillators different from precise digital Oscs like e.g. in DUNE 2.5 and this is the behavior of a proprly modeled Minimoog Model D.

To tune final patches as good as possible you will need a tuning plugin. Most of the factory patches were fine tuned as good as possible. Anyway if it is in tune at e.g. C3 it could be a few cents different at lower and higher octaves (around 1 cent per octave wit hte Init patch).

FWIW NI Monark has exactly the same kind of behavior.

Example: In the Init patch of The Legend you have to set the global fine tune to +4 cents to have it in tune at C3. If you add a second Osc at a fine tune of around +8 you could set the global fine tune back to 0.
With key tracking at 0 and a second Osc added at +8 cents you would have to set the global tune to -4 cents. So the global tune also depends on the Osc detuning used (which could happen also in other synths).

If you set Osc keytracking to 0 like Yorrrrrr already mentioned the global tuning is at 0 cents with one oscillator. Anyway this way it sounds less analog and/or less like a real Minimoog Model D (even worse with settings below 0).

Adding Osc drift (back panel) at the maximum amount adds a variation of around 2.5 cents. With that used you could"reduce the "correction" from +4 cents mentioned above to +2 cents (Init patch).

Can u summarize The Legend VS Minimoog ModeL D?
There is alot of pages here :)

Post

Here is someone who actually owns one today....
brok landers wrote:hi fellas,

i think it's time for me to chime in here to clear up some stuff.
as maybe some of you know richard from synapse audio and me developed legend.
he did the coding (which was a tremendeous and very tedious effort), i brought the mini (in i must say _exellent_ condition - thanks, till!) and measured it via ear, waterfall diagrams, oscilloscopes and analysers and whatnot and tweaked the "rough cut" to as close as imo one can get. i also made the dry/wet demo comparsion files, which are posted on the propellerheads forum.

the minimoog, although not a very mojo-filled synth as one who only knows it from the myth behind it might think, is very tricky in quite some aspects, that, if you do not put weight on these aspects, lead to a sound that can never reproduce the behave of the minimoog.

the envelopes are complex beasts in terms of behaviour, they have a quite noticable amplitude-spike within the first milliseconds, which flattens as soon as the attack and the decay are raised (other fully analog synths show that behaviour, too). as you can imagine, _this_ makes the sound _very_ snappy. then there's some kind of dc-overflow in the filter env amount, which has the tendency to rise and add up from a lower point of the amount originally being set up with every keystroke, up until it reaches the full amount that has been set up. to get this behavoir right over the full range of the pots, we did sweat blood. :)
then there's, obviously the cascaded ladder filter which the minimoog is so famous for, which interacts with several filter-internal gainstages and dc-offsetfactors, of which the typical, characteristic drive behaviour comes from. the vca stage does the rest of the silky and round sound. also, the minimoog ladder, due to some kind of highpass in the feedback of the filter, is flat response at max resonance, down to around 150-100hz (depending on how carefully the original was maintained and serviced), then the resonance rapidly drops off into nirvana. this also is the reason, why the mini handles basses so well, as the resonance, as soon it reaches the first few harmonics of the osc, doesn't overlay these, as it drops away in level at below this frequency.

also the osc shapes _mandatorily_ have to be done perfect, otherwise the filter doesn't behave like in the original. the square and the two pulses were really hard and crucial, but the hardest was the triangle - because in the late versions of the model d (83/84) were _far_ away from a real triangle shape - strange enough the older versions were way closer to a triangle.

then the pink noise, which, when used for frequency/filter frequency modulation, is filtered down in the control path only to some kind of red noise, was also very crucial. if you want fm that sounds close to a minimoog, the osc behaviour as well as the red noise has to be quite acurately modelled, otherwise it will sound _nothing_ like a minimoog.

there' s plenty of stuff i could go on about, but you might understand what i am trying to say.
i myself owned three minis, one from 76 and two from the last ones in the 80s.

now as i was deeply involved in the development of legend i can tell you that up to now there is _no_ emulation of the minimoog coming so close as legend does - in all aspects. richard did a einstein-like job on this, sitting here in my studio day after day, programming like a maniac. but of course you don't have to believe me (of course you might think i'm biassed) - just listen to the dry comparsion wav files i did once they're on the synapse audio site, ready to be downloaded or get them from the link of the propellerheads forum (i believe someone posted the link to these here in this thread), if they're still there, and judge for yourself. and get the demo, once it is out and fiddle around with it.
i can only say, that legend sounds and behaves nearly identical to the minimoog we had here, and as i stated before, it was in excellent condition and well serviced.

the video demo which kevin made is a very nice one, but i can see some users wanting something completely different, when they have the sound of a minimoog in mind, so for them a rather contemporary edm/bigroom demo might be a red flag. but believe me - legend delivers.

so, to sum it up, if you are after the minimoog-sound/behave for your daw, legend is the closest to get there. while i know all minimoog emulations there are, and some of them are great synths in their own right, they fail when it comes to some of the above stated aspects, which make a great part of the minimoog sound (even monark fails f.e. deeply in the noise fm modulation). and on top of it legend offers quite some tweaking possibilities, that an original minimoog only gives you, if you open it and tweak the screwpots. and it's polyphonic - so _that_ is unique, as there obviously never was a polyphonic minimoog.

now there you have it.

Post

michaelbrac wrote: Can u summarize The Legend VS Minimoog ModeL D?
There is alot of pages here :)
As mentioned earlier here The Legend is as close to a real Minimoog Model D as currently possible in software.

Besides that The Legend at the current version has those additional features compared to a vintage Minimoog Model D:
- Patches and the Modwheel position (= modulation amount) could be loaded/saved
- 4 voice polyphony mode with adjustable stereo width (+ Glide/Portamento works there too)
- 4 voice Unison mode (monophonic) with adkustable Detune and stereo width amounts
- For the 4 voice polypnony and the 4 voice Unison mode the "vectorized core" in The Legend makes it possible to use only around the double CPU amount (or less) of that used for a single voice in Mono mode
- full ADSR envelopes (oposing to ADS or ADSD) in the original Minimoog
- additional dedicated fine tune knobs for all Oscs (only a single tune knob for coarse ad fine tuning Oscs 2+3 in the real Mini) ; The "SEMI" tune knobs are still capable of fine tuning, especially in combination with the SHIFT button at the computer keyboard
- 12dB/oct LPF (based on the same topology as the Minimoog Model D ladder filter)
- 12dB/oct + 24dB/oct BPF (based on the same topology as the Minimoog Model D ladder filter)
- Delay + Reverb effects
- two Model revisions (Early and Late, at the back panel)
- aftertouch assignable to Cutoff or mod amount (at the back panel)
- velocity adjustable to Volume or Cutoff (at the back panel)
- adjustable pitch bend range (back panel)
- note priority could be set between Last (= default) and Low/RR (back panel)
- adjustable mod amount range, Cutoff range and Resonance range (back panel)
- adjustable Osc key tracking + Osc drift amount (back panel)

Anyway you could ignore all the advanced setting and use only the original Minimoog features if you like.
You would have to set the envelope Release to either 0 or teh same value as the Decay then.
The parameter ranges match to a Minimoog Model D so besides fine tuning of the patches you could reprogram Minimoog patch sheets mostly 1:1.

Besides that The Legend also includes an effects version ofthe plugin to use it with other synths or add an additional filter (or multiple ones) to The Legend. If you route a MIDI track to the FX plugin you could also use the built-in envelopes and the Oscs.
Ingo Weidner
Win 10 Home 64-bit / mobile i7-7700HQ 2.8 GHz / 16GB RAM //
Live 10 Suite / Cubase Pro 9.5 / Pro Tools Ultimate 2021 // NI Komplete Kontrol S61 Mk1

Post

I noticed that The Legend uses another base note for filter key tracking compared to Monark.

On Legend it's G1.
On Monark it's F2.

Which of the two is correct (like on the real minimoog)?

Also the global transpose/coarse works different.
Monark does "Midi transpose" (means you can compensate a octave switch on the synth by doing a note-shift),
Legend does "Oscillator transpose". The real minimoog didn't have a global octave/transpose knob as fas as I know.

Post

Chris-S wrote: Also the global transpose/coarse works different.
Monark does "Midi transpose" (means you can compensate a octave switch on the synth by doing a note-shift),
Legend does "Oscillator transpose". The real minimoog didn't have a global octave/transpose knob as fas as I know.
Sorry, i do not notice a difference in sound when using either the transpoe in my MIDI keyboard controller or the transpose in the The Legend GUI (TUNING section).

Of course a workaround for this is also setting the octaves in all oscillators up/down (the only way how it worked in a real Minimoog Model D).

FWIW I do not see a bug in the global transpose feature so far.

Chris-S wrote:I noticed that The Legend uses another base note for filter key tracking compared to Monark.

On Legend it's G1.
On Monark it's F2.

Which of the two is correct (like on the real minimoog)?
How exactly did you measure this?

Did you use the Init patch as a starting point for this or another one?
Ingo Weidner
Win 10 Home 64-bit / mobile i7-7700HQ 2.8 GHz / 16GB RAM //
Live 10 Suite / Cubase Pro 9.5 / Pro Tools Ultimate 2021 // NI Komplete Kontrol S61 Mk1

Post

Ingonator wrote:
Chris-S wrote: Also the global transpose/coarse works different.
FWIW I do not see a bug in the global transpose feature so far.
I didn't say it's a bug. Just noticed that Legend and Monark are different in this.
Ingonator wrote:
Chris-S wrote:I noticed that The Legend uses another base note for filter key tracking compared to Monark.

On Legend it's G1.
On Monark it's F2.

Which of the two is correct (like on the real minimoog)?
How exactly did you measure this?
Starting from init I activated filter key track and searched the note where key track on/off has no effect on the cutoff freq.

Post

Well, first Monark is not the ultimate refererence concerning what is right or wrong concerning a proper Minimoog emulation while personally i would trust The Legend more than any other plugin in that respect.
Richard has done really a ton of measurements and semed to take care even of small details.
Chris-S wrote:
Ingonator wrote:
Chris-S wrote:I noticed that The Legend uses another base note for filter key tracking compared to Monark.

On Legend it's G1.
On Monark it's F2.

Which of the two is correct (like on the real minimoog)?
How exactly did you measure this?
Starting from init I activated filter key track and searched the note where key track on/off has no effect on the cutoff freq.
I guesd with key track on/off you mean either 100% (= both switches on) or 0% (= both switches off)?

With both switches on (= 100%) and then off (= 0%) I could replicate your method for Monark but not with Legend (for none of the G notes in different octaves and also no other notes yet).

Anyway i have another method based on the filter self-oscillation where i receive a Note "G" from Legend:
The 24dB LPF is set to self-oscillation wit hall 3 Oscs set to "Off" (= pure filter self-oscillation, Resonance at 100% and 100% filter key tracking (= noth key tracking switches on). while holding C3 (woudld also work with otehr notes played) and using a Tuning FX plugin with those Cutoff settings the note is in tune (at "C" or other notes when those are played instead):
46 Hz, 96 Hz, 195 Hz, 395 Hz, 791 Hz, 1.6 kHz (out of audible range with 3.2 kHz)

If the Cutoff value corresponds to the pitch (which at self-oscillation is normally the case) those frequencies would correspond to notes F#1 (while close to G1), G2, G3, G5 and G5 then.

The settings used enable to play the resulting Sine tone received from the filter self-oscillation at a more or less proper pitch over multiple octaves (where the tuning is more off the more you move away from the C3 note that was used for the tuning here). The Cutoff frequencies mentioned above then correspond to transpose the played Sine to lower/higher octaves.

The values found above correspond to the Cutoff Range and Resonance Range at the back panel of The Legend set to default (= 0). If you set a different Cutoff range value at the back panel also the Cutoff values mentioned above will change. A cutoff range change from 0 to around +1.90 results in a pitch change at self-oscillation of around one semitone and a cutoff range of +4.05 around +2 semitones. A maxiimum Cutoff range value of +5 would correspond to around +2 full semitones and 37 cents (similar amounts with negative Cutoff range values).

Monark does not display the Cutoff values as a frequency so this method does not really work there.
The Cutoff frequency displayed in The Legend seems to fit quite well to the measurement in an Analyzer (with e.g. Osc 1 activated, Resonance below self-oscillation and key tracking off).


If you want to adjust The Legend closer to other Minimoog emulations like e.g. Monark there are multiple options/parameters at the back panel of The Legend including those in the OSCILLATORS and FILTER sections (Cutoff Range was already mentioned above).
Last edited by Ingonator on Fri Feb 17, 2017 3:47 pm, edited 2 times in total.
Ingo Weidner
Win 10 Home 64-bit / mobile i7-7700HQ 2.8 GHz / 16GB RAM //
Live 10 Suite / Cubase Pro 9.5 / Pro Tools Ultimate 2021 // NI Komplete Kontrol S61 Mk1

Post

Chris-S wrote:I noticed that The Legend uses another base note for filter key tracking compared to Monark.

On Legend it's G1.
On Monark it's F2.

Which of the two is correct (like on the real minimoog)?
Legend is correct to the hardware it used to model, Monark is prob the same.

Post

From the quote below:

"then there's some kind of dc-overflow in the filter env amount, which has the tendency to rise and add up from a lower point of the amount originally being set up with every keystroke, up until it reaches the full amount that has been set up. to get this behavoir right over the full range of the pots, we did sweat blood."

That seems to be one of the biggest obstacles when doing Minimoog sounds in other synths. I wonder how one could emulate that behavior...

Was that behavior even deliberate on the hardware or more of a design flaw?

Post

I created screenshots of teh 7 waveforms and 2 Noise shapes for both the Late and Early revison models of The Legend. Drive an dFeedback were set to 0 and also Amp saturation is at minimum (default value) to keep the waveforms as original as possible.

A) The Legend Late model revision (Triangle at upper left to White Noise at lower right):

https://dl.dropboxusercontent.com/u/532 ... eforms.png
Image


B) The Legend Early model revision (Triangle at upper left to White Noise at lower right):

https://dl.dropboxusercontent.com/u/532 ... eforms.png
Image
Ingo Weidner
Win 10 Home 64-bit / mobile i7-7700HQ 2.8 GHz / 16GB RAM //
Live 10 Suite / Cubase Pro 9.5 / Pro Tools Ultimate 2021 // NI Komplete Kontrol S61 Mk1

Post

Ingonator wrote:I created screenshots of teh 7 waveforms and 2 Noise shapes
Why?

Post

AnX wrote:
Ingonator wrote:I created screenshots of teh 7 waveforms and 2 Noise shapes
Why?
Because he can. Why not?

Post

And the hardware waveforms to compare them to? 8)

Post

fluffy_little_something wrote:And the hardware waveforms to compare them to? 8)
Yes, it would be interesting. And Monark's. And any other minimoog emulation.
Last edited by Yorrrrrr on Fri Feb 17, 2017 5:19 pm, edited 2 times in total.

Post


Post Reply

Return to “Instruments”