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

[Ingonator]

Hi Ingo, C4 is middle C (midi #60, 260 hz). 8' foot pitch is fundamental (no octave switch).
If you have a standard 5 oct midi keyboard, the MOOG keyboard would start at the leftmost F.
Playing that note, pressing the Keytrack switches should have no effect, but on LEGEND they have.

I just changed my post above. The procedure mentioned in that article and that i just quoted is different from yours. It uses self-oscillation so is independent from the Oscs which makes sense.

[Ingonator]

http://www.soundonsound.com/reviews/moo ... og-model-d

Finally found the procedure there that is used for the test:

So how did the two Minimoogs compare? I switched off the oscillators on both synths and caused the filter to oscillate by maximising the emphasis, switched on both Keyboard Control switches so that the filter tracked the keyboard 1:1, and then tuned the top A on the keyboard to A=440. Then I played. Initially, the results were disappointing; I could play downward on my Minimoog for fewer than 20 semitones (to around 150Hz) before the emphasis diminished to the point that the self-oscillation was gone but, on the new model, I could play down to the lowest note on the keyboard and the self-oscillation was still going strong at 50Hz and below.

So i guess with "tuning" he means adjusting the Cutoff to receive 440 Hz in a tuning plugin? At pure self-oscillation tuning is only possible with the Cutoff.

Will try to check this with The Legend soon.

I tried to use the procedure mentioned above (in the SOS article) with The Legend and the help of a tuning plugin (the one included in Live 9). i tried to adjust the pitch with the Cutoff which is usual at self-oscillation and 100% filter key tracking.

With teh A4 note at my MIDI keyboard i was not able to set the Cutuff as low to tune taht key to 440 Hz but with the the A3 note pressed at the controller i was able to tune this to 442 hz which corresponds to 48 Hz at the Cutoff.
With this the first key that no longer works with proper self-oscillation is F1 at my MIDI controller. When holding a note (Amp Sustain level in The Legend is at maximum while all Env times are at 0) the sound stops after a short time while with higher notes the self-oscillation tone continues when holding a note.


UPDATE:
A screenshot of the test patch and the tuning plugin used (Cutoff at 48 Hz ; the frequency of the A3 note used at my MIDI controller shown in the tuning plugin):
https://dl.dropboxusercontent.com/u/532 ... st%201.png


UPDATE:
In Monark i get a simiiar result with the A3 key at my MIDI controller tuned to 442 Hz using the Cutoff and F1 at my MIDI controller is the key where the self-ocillation stops working which is the same key as in The Legend.



A quote of the second part of the original test procedure from teh SOS article:

Then I played. Initially, the results were disappointing; I could play downward on my Minimoog for fewer than 20 semitones (to around 150Hz) before the emphasis diminished to the point that the self-oscillation was gone but, on the new model, I could play down to the lowest note on the keyboard and the self-oscillation was still going strong at 50Hz and below.

In The Legend and Monark you could go down until around 108 Hz or 27 semitones (from A3 in my MIDI keyboard that was tuned to 442 Hz) until the self-oscillation "vanishes".

[ENV1]

I switched off the oscillators on both synths and caused the filter to oscillate by maximising the emphasis, switched on both Keyboard Control switches so that the filter tracked the keyboard 1:1, and then tuned the top A on the keyboard to A=440. Then I played. Initially, the results were disappointing; I could play downward on my Minimoog for fewer than 20 semitones (to around 150Hz) before the emphasis diminished to the point that the self-oscillation was gone...

Just for the record: Thats precisely how Arturias minimoog behaves. Above 150Hz the VCF oscillates relatively strong and stable, then it gets increasingly weaker and unstable, and finally the oscillation dies off to the point of inaudibility. (100Hz = ~ -72dB.)

Cutoff = -4.46 / 43.84Hz


Note: minimoog 1.6 and 2.5.2 are practically identical, so no apparent difference here between V1 and V2.

V3 not checked, but might be different since it has a new VCF.

[waltercruz]

Just for the record: Thats precisely how Arturias minimoog behaves. Above 150Hz the VCF oscillates relatively strong and stable, then it gets increasingly weaker and unstable, and finally the oscillation dies off to the point of inaudibility. (100Hz = ~ -72dB.)

Cutoff = -4.46 / 43.84Hz


Note: minimoog 1.6 and 2.5.2 are practically identical, so no apparent difference here between V1 and V2.

V3 not checked, but might be different since it has a new VCF.

If their page about the new filter is right, this behaviour was 'fixed' in the latest version. At least is what this text implies:

The previous implementation tried to address both the linear and nonlinear aspects of the filter at the same time, but it did so in a suboptimal way so no two were 100% correct. This was very evident when the filter was set in self-oscillation mode: if you kept the emphasis high enough and tried to sweep the cutoff, you could see that the amplitude of the sine wave generated by self-oscillation was very far from constant, and at the lowest cutoff settings it would just die out.

https://www.arturia.com/mini-v/behindthemini-v

There's some digression about this on the topic "Arturia synth development" here on kvr.

[Chris-S]

The procedure mentioned in that article and that i just quoted is different from yours. It uses self-oscillation so is independent from the Oscs which makes sense.

Yes, I was not talking about self-oscillation. Another thing.

[ENV1]

If their page about the new filter is right, this behaviour was 'fixed' in the latest version. At least is what this text implies:

Youre probably right.

Farther down theres this:

As well as giving the filter an unparalleled fidelity, it now self-oscillates linearly throughout the audio spectrum. Musicians can now take full advantage of the self-oscillation for sound-design, for instance when using it as a supplementary oscillator to beef-up a vivid lead or a nasty bass.

So in essence theyve 'fixed' something which was perfectly correct the way it was.

Thats great.

Wonder what Xavier Oudin thinks about this. Since he had nailed that aspect more or less perfectly, (if you believe Gordon Reids account), im not sure if hes laughing or crying.

[Ingonator]

The procedure mentioned in that article and that i just quoted is different from yours. It uses self-oscillation so is independent from the Oscs which makes sense.

Yes, I was not talking about self-oscillation. Another thing.

But your method does not seem to work opposing to that in the SOS article which is based on self-oscillation.
With that method both The Legend and Monark seem to behave properly.

Using only the filter with self-oscillation instead of having the oscillators involved too seems to make sense.

Could you mention a source/reference for the method that you have used like e.g. that SOS article that i used for the filter self-oscillation method?

[waltercruz]

So in essence theyve 'fixed' something which was perfectly correct the way it was.

Thats great.

Wonder what Xavier Oudin thinks about this. Since he had nailed that aspect more or less perfectly, (if you believe Gordon Reids account), im not sure if hes laughing or crying.

But now the filter behaves really better doing a filter sweep. Try to do a sweep on Mini-V2, is horribly stepped. Now it behaves more like a proper filter (except by this part above 200 Hz).

[Ingonator]

http://www.soundonsound.com/reviews/moo ... og-model-d

Finally found the procedure there that is used for the test:

So how did the two Minimoogs compare? I switched off the oscillators on both synths and caused the filter to oscillate by maximising the emphasis, switched on both Keyboard Control switches so that the filter tracked the keyboard 1:1, and then tuned the top A on the keyboard to A=440. Then I played. Initially, the results were disappointing; I could play downward on my Minimoog for fewer than 20 semitones (to around 150Hz) before the emphasis diminished to the point that the self-oscillation was gone but, on the new model, I could play down to the lowest note on the keyboard and the self-oscillation was still going strong at 50Hz and below.

So i guess with "tuning" he means adjusting the Cutoff to receive 440 Hz in a tuning plugin? At pure self-oscillation tuning is only possible with the Cutoff.

Will try to check this with The Legend soon.

I tried to use the procedure mentioned above (in the SOS article) with The Legend and the help of a tuning plugin (the one included in Live 9). i tried to adjust the pitch with the Cutoff which is usual at self-oscillation and 100% filter key tracking.

With teh A4 note at my MIDI keyboard i was not able to set the Cutuff as low to tune taht key to 440 Hz but with the the A3 note pressed at the controller i was able to tune this to 442 hz which corresponds to 48 Hz at the Cutoff.
With this the first key that no longer works with proper self-oscillation is F1 at my MIDI controller. When holding a note (Amp Sustain level in The Legend is at maximum while all Env times are at 0) the sound stops after a short time while with higher notes the self-oscillation tone continues when holding a note.


UPDATE:
A screenshot of the test patch and the tuning plugin used (Cutoff at 48 Hz ; the frequency of the A3 note used at my MIDI controller shown in the tuning plugin):
https://dl.dropboxusercontent.com/u/532 ... st%201.png


UPDATE:
In Monark i get a simiiar result with the A3 key at my MIDI controller tuned to 442 Hz using the Cutoff and F1 at my MIDI controller is the key where the self-ocillation stops working which is the same key as in The Legend.



A quote of the second part of the original test procedure from teh SOS article:

Then I played. Initially, the results were disappointing; I could play downward on my Minimoog for fewer than 20 semitones (to around 150Hz) before the emphasis diminished to the point that the self-oscillation was gone but, on the new model, I could play down to the lowest note on the keyboard and the self-oscillation was still going strong at 50Hz and below.

In The Legend and Monark you could go down until around 108 Hz or 27 semitones (from A3 in my MIDI keyboard that was tuned to 442 Hz) until the self-oscillation "vanishes".

I just did the same test with Arturia Mini V3 which i got included in V-Collection 5. Opposing to The Legend and Monark that work like the original Minimoog when doing this with Mini V3 even at playing a note around 30 Hz or less the self-oscillation still works.

Another thing is that even with the improvd filter in Mini V3 at self-oscillation with higher frequencies you still get noise which does not happen with The Legend and Monark. The Legend at pure filter self-oscillation does not produce additional noise in the whole audible frequency range.
Even without self-oscillation Mini V3 seems to produce aliasing at higher octaves and/or frequencies. With each octave up the aliasing gets worse.
The problem with the self-oscillation seems to be based on a general problem with aliasing at higher frequencies. The only way to improve this in Mini V3 seems to be using the host at 96 kHz sample rate.

I have not directly compared Osc FM and Filter FM in Mini V3 yet but it is possible that this does not work correctly too.
Self-oscillation and Filter FM are features that usually do benefit from zero-delay feedback filters like in The Legend, Monark and also Diva.

Anyway as a reference for how a real Minimoog should sound The Legend and als Monark seem to be much better choices than Mini V3.

While the other 2 synths arte better at emulating a Minimoog properly Mini V3 has it's own strengths in the advanced features like e.g. polyphony, Unison, Osc Sync, dedicated LFO, mod matrix, FXs and the vocal filter.

[ENV1]

But now the filter behaves really better doing a filter sweep. Try to do a sweep on Mini-V2, is horribly stepped. Now it behaves more like a proper filter (except by this part above 200 Hz).

Yes, it can be slightly 'steppy' when you drag the knob slowly with the mouse because there is no smoothing. (Most noticable at high resonance.) But thats not really the fault of the plugin, its due to the little 'micro-pauses' you make when you drag slowly, in other words it wont step if you drag fast/continuously, nor will it ever step with the EG or the LFO or host automation.


Anyrate, to me 1.6 is still the best version. From a functional perspective V2 has only 2 advantages, namely less aggressive voice-detuning in poly mode and faster pitch changes from one note to the next. (V 1.6 has a slight lag here, like a minimal amount of glide, which can get in the way when you want very fast sounds like pitched clicks.) Other than that V2 has (in my opinion) no real advantages, only disadvantages and lots of bugs that werent there before. And the new V3 isnt really my cup of tea either, so yeah, for me its definitely 1.6

[waltercruz]

Yes, it can be slightly 'steppy' when you drag the knob slowly with the mouse because there is no smoothing. (Most noticable at high resonance.)

Just detailing, I did the test with a controller. The stepping is great when doing a sweep with the filter in self-oscillation. I know that this is a somewhat extreme situation, but these situations kinda display the limits of the emulation.

[Mutant]

Yes, it can be slightly 'steppy' when you drag the knob slowly with the mouse because there is no smoothing. (Most noticable at high resonance.)

Just detailing, I did the test with a controller. The stepping is great when doing a sweep with the filter in self-oscillation. I know that this is a somewhat extreme situation, but these situations kinda display the limits of the emulation.

If the filter knob is hardware controlled via CC, then it will be only 127 values and no amount of smoothing will cure that problem.
In some VST synths (Xhip for example), you can hold CTRL and/or RMB to make mouse control more precise - this makes it as smooth as an analog knob, way better than knobs on current DSI synths.

[yul]

The real Minimoog had a decreasing resonance at lower cutoff frequencies afaik.

This gave the sounds a distinctive flavor as well as safeguarding against awful unwanted resonance spikes I believe.

What has exactly changed in Legend? I am not sure I fully understand? To have a fixed resonance is not natural to the MM unless I am missing something?

Thanks!!

[Ingonator]

The real Minimoog had a decreasing resonance at lower cutoff frequencies afaik.

This gave the sounds a distinctive flavor as well as safeguarding against awful unwanted resonance spikes I believe.

What has exactly changed in Legend? I am not sure I fully understand? To have a fixed resonance is not natural to the MM unless I am missing something?

Thanks!!

Some of the recent discussions here also were about Arturia Mini V3 and V2.

Not sure why you think that The Legend has a "fixed resonance" behavior. Did you check the demo of The Legend yourself yet?
At low Cutoff the resonance influence is indeed decreasing. Of course the result also depends on the amount of filter key tracking and the effect is biggest without key tracking.
The advanced features at the back panel of The Legend also allow adjusting the Cutoff and Resonance ranges beyond or below the original synth. Same is possible for the range of modulation for the Oscs and/or filter. The default values there are like in the original Mini.

The Legend is the closest you could get to a proper Minimoog odel D emulation and that without killing your CPU and without audible aliasing even at 44.1 kHz sample rate in the host.

While Richard at Synpase Audio tried to emulate even small details of the original synth the advanced features like e.g. polyphony, Unison, efects, 12dB LPF and 2 Bandpass filter modes add an option to go beyond the original features.

[yul]

[quote

[ Some of the recent discussions here also were about Arturia Mini V3 and V2.

Ah thanks for the clarification I must have misread then! I could not imagine that the Legend would go over that detail.

[
The Legend is the closest you could get to a proper Minimoog odel D emulation and that without killing your CPU and without audible aliasing even at 44.1 kHz sample rate in the host.

Of course I have tried the demo (it's stellar) about a month ago but was thinking something had changed in an update (which it hasn't other than new added filters)

I will pull the trigger quite imminently actually it's nearly impossible to pass as a software creation of this caliber.

Cheers!!