Originally Posted by Sebo Synths ➡️

I did the tests for Feedback level.

First of all I just set some patches that produces high frequency noise at Feedback 7 and tried to lower Feedback with Shape control. At 90% aprox. noise is gone.

Then I set some patches I know well and set Feedback at 85.7%, in general they seems to sound good and close to Dexed. I feel some needs a bit more, and some a bit less, but I think 85.7% is OK.

Sorry if I made you to change it in the past, but it was different (keep reading).

I made a more "precise" comparison with Dexed like I did in the past.
I used this file:
https://drive.google.com/file/d/1l0g...ew?usp=sharing
If you want to check in PLG, there are 8 voices, all using algorithm 32, which is the only one that have a Feedback Op as carrier.
I tried to make some voices to test with different carrier amounts, but having the feeedback in non-carrier operators makes things very noisy, very fast, so I gived up....

The thing is, I can't make FM64 produce a waveform like Dexed.
Dexed wave with Feedback at 7 is like a square with noise bursts, but FM64 is like a sawtooth with noise burst, if I reduce the Feedback using Shape the noise bursts fade but the wave never get squarish....

Weird thing is I have a recording of FM64 1.2-0 (I keep all the recordings in a project) with Feedback set to 100% from Shape control and the waveform is almost exactly as Dexed (square with noise bursts).
I don't know what has changed...

Anyway with current version 85.7% works well.

Originally Posted by vlz ➡️

The feedback FM wave is supposed to be sawtooth -like, that's what the equation gives us. The noise bursts happen when the index of modulation is beyond 1, whatever that translates into the % control range.

Originally Posted by dukesrg ➡️

Have you tried to reduce velocity?
1.2-0 was wit fixed compensation but I suppose that should not affect alg32 that way.

OK I'll reduce the feedback =7 and try to check with PLG.

Originally Posted by SkyWriter ➡️

Hey duke, does the NTS-1 have any state retention? Or 'presets'? Or do I have to load it up every time I turn it on? Although I got it for more debug and hardware investigation - but it's looking like a good eurorack multi effects kit. I could get a pi to do file loading from a shared development share to serve up programming to a couple of NTS-1's. I just hate having a computer 'part of what i do' for music. Development is fine, but when I'm noodling around or seriously playing, I don't want to see a mouse, trackpad, GUI, etc... if I can program them like the rest of the models, that's ideal.

Is it a full single User/VPM voice + three effects layers?
------
I should put one inside the Poly D. It cold easily fit in back though, like this H9.
------
Ok, I see it's at least the above; do I get the same number loadable slots:
16 osc
16 mod
8 del
8 rev
Except I get del and rev concurrent? And it's stored in non-volatile store? That would rock racked up in my rig. Yum!! Where's @ psionic11 !?? :0)

Originally Posted by Sebo Synths ➡️

For the test I used this bank with velocity sensitivity set to 0 on all operators, so velocity should not affect anything... I think.
I could try to do it again with velocity set to 0...

Originally Posted by Sebo Synths ➡️

I think I saw a project using Arduino to save presets for the NTS-1, but I can't remember where... could be on Github....
I will look for it.

Originally Posted by dukesrg ➡️

Yes, velocity itself should also be set to moderate level

upd: checked non-percent parameter - for both NTS and XD it have a maximum span of 0 to100 which is displayed 1...101. It costs NOTHING to make it a least 7-bit ti comply with MIDI standard, but KORG coult not even align the controller numbers betwee the logue series....

Originally Posted by Sebo Synths ➡️

The MIDI implementation of the Prologue is a mess, I had 80s synths with better MIDI implementation.

Originally Posted by dukesrg ➡️

Yes, velocity itself should also be set to moderate level

upd: checked non-percent parameter - for both NTS and XD it have a maximum span of 0 to100 which is displayed 1...101. It costs NOTHING to make it a least 7-bit ti comply with MIDI standard, but KORG coult not even align the controller numbers betwee the logue series....

Originally Posted by Sebo Synths ➡️

I tried the Feedback in ver 1.2-0 and 1.3-2 with velocity set to 50% and 0%, and I don't have any difference with previous test.

But I managed to get the same waveshape of Dexed with FM64 1.3-2, the difference is I played a lower note. I don't know if the difference is because Feedback changes with note number on FM64 or if have to do with aliasing/filtering.
Anyway I think it doesn't matter...

Originally Posted by dukesrg ➡️

Shape = Velocity
Alt = ?

Originally Posted by AndyHornBlower ➡️

Can FMxx respond to MOD?

I seem to remember there's no LFO, but I prefer to use MOD to control EG Bias, on the TX7 / TX802, rather than controlling the LFO.

Originally Posted by dukesrg ➡️

Still can't find the best option for custom params.
Bipolar params are not interchangeable with SHAPE/ALT, since as soon as them bound to either param or SHAPE/ALT, the default value can't be controlled.

In case of FMxx it means that for most of the parameters, only params can be used. The good thing, it is possible to add both 0x-2x scale and +/- 100 % (or whatever) parameter options so one could bound the more suitable one. That will use some more memory and code, but will be more versatile.
From the other hand, Shape will be fixed to Velocity and I can't find the good option for Alt for now.

So for now, I can propose
Shape = Velocity
Alt = ?
Param1 = Voice
Param2 = Bank (or can be customizable when banks <= 3)
Param4-6 customizable

Params is selectable from (not per-op inf not stated explicitly):

1. Algorithm offest +/- 39%
2. Level offset +/-100% (+/-50)
3. Level scale +/-100% (0x-2x, does it make sence - it is an exponent?)
4. KVS offset +/-100% (+/-7, clipped)
5. KVS scale +/-100% (0x-2x, non-clipped)
6. KLS offset +/-100% (does it make sence to ad offset to the level scaling?)
7. KLS scale +/-100% (0x-2x, non-clipped)
8. Feedback offset +/-100% (+/-7, clipped)
9. Feedback scale +/-100% (0x-2x, non-clipped)
10. Rate offset +/-100% (+/-50(?), clipped)
11. Rate scale +/-100% (0x-2x, non-clipped)
12. KRS offset +/-100% (does it make sence to ad offset to the rate scaling?)
13. KRS scale +/-100% (0x-2x, non-clipped)
14. Detune offset +/-100% (+/-50)
15. Detune scale +/-100% (0x-2x)
16. Waveform 6 - 1-8
17. Waveform 5 - 1-8
18. Waveform 4 - 1-8
19. Waveform 3 - 1-8
20. Waveform 2 - 1-8
21. Waveform 1 - 1-8
22. Waveform 6/5 - 11-88
23. Waveform 4/3 - 11-88
24. Waveform 2-1 - 11-88

Any more ideas?

Originally Posted by Sebo Synths ➡️

Hi, I'm a bit busy lately because I'm rehearsing a lot. We have a gig (kind of) after a year of no live playing... We will play in a bar without people, it will be for streaming only, at least it is something... and some welcome money

I tried ver 1.3-3, I didn't know what controls I have to assign to Shape, but I think 4 is rate shift, and 5 is rate scale.

Rate Shift (4) doesn't work well, it almost kill the sound. If I turn the knob to the left I have a very short sound at 0 to 10%, anything higher than 10% it is just silence...


Rate Scale (5) works well and IT IS FANTASTIC!!! Thanks!

About the Parameter list, I think it is OK, right now I can't think of anything else, execept may be you forgotten to include LFO routing.
[*]Level scale +/-100% (0x-2x, does it make sence - it is an exponent?)

To me it makes sense, in general I preffer mult to offset...
[*]KLS offset +/-100% (does it make sence to ad offset to the level scaling?)

I dont't think it is much useful...
[*]KRS offset +/-100% (does it make sence to ad offset to the rate scaling?)

In this case I don't know, I have to try it...


Also, do you have time to check Feedback level on PLG?

Originally Posted by dukesrg ➡️

I'm also way too busy - 2 weeks till the end of Q4. Even can't find the time to lay XD on the table and test my own updates
Yep, Rate Offset on 1.3-3 was +50+150 Fixed in FM64 1.3-4. Please chech which one of those to methods is more usable. Offset might require more precision because it can vary rate in much wider range. But I suspect it should keep voice character more predictable.

Originally Posted by Sebo Synths ➡️

I tried ver 1.3-4

Rate Shift (4) works well now.

Rate Scale (5) is not working in this version, it does nothing.

I noted that Rate Scaling (envelope timing changing with note number) is not working in none of the 2 last versions (1.3-3 and 1.3-4).

It is hard to say which method is better. I can't compare as they didn't worked at the same time. I think both are useful in different ways. May be after I play a bit more with both I can pick a preffered one. I agree with you about offset have a wider range, and with small values it gives a more predictable result, with big values it mess the character. I didn't noted a big change of character with Scale, but the range is more limited.
I think that any of them will work...

Originally Posted by pauperrasta ➡️

Hi, another manual and simple config for Eurorack system.

Originally Posted by dukesrg ➡️

Pinged Etienne over a week ago and then asked Sinevibes founder if he was able to contact Etienne. No idea which option worked, but today Etienne is finally awoke, Unit Index updated!

Originally Posted by Deleted e199339 ➡️

Thank you Oleg! Glad to know we're not forgotten.

Originally Posted by dukesrg ➡️

Also did not miss a chance and noted to Ettiene about several minor firmware issues at github that needs attention.

BTW, Artemiy (Sinevibes) also responded, no he did not reach Korg yet, so was also happy that index was updated

Originally Posted by Sebo Synths ➡️

OK, it seems that I was wrong and Korg is still developing Logues:

https://www.musicradar.com/news/korg...ur-own-samples

An updated SDK?

Originally Posted by SkyWriter ➡️

'Revamped' SDK is what I heard. Apparently what duke saw Etienne working on github.

It would be nice if they at least fixed the SDK to meet the spec; Q and CF. I would love velocity - though not necessary.

Not interested in drum machines. Probably in the XD form factor. Oh well, the wait continues...

Originally Posted by SkyWriter ➡️

'Revamped' SDK is what I heard. Apparently what duke saw Etienne working on github.

It would be nice if they at least fixed the SDK to meet the spec; Q and CF. I would love velocity - though not necessary.

Not interested in drum machines. Probably in the XD form factor. Oh well, the wait continues...

Originally Posted by Sebo Synths ➡️

I like the DrumLogue idea.
I'm waiting to buy a drumachine for long time, but my ultra moded TR-707 is hard to beat.
I'm a big fan of the LinnDrum, but it is out of my reach (even I don't like Behringer, I will buy one if they clone it).
The DrumLogue initial specs are apealing to me.