KVR Audio

martin_l · Post by **martin_l** » Tue Feb 22, 2011 1:45 pm

LXNDR1 wrote:
martin_l wrote:Since it is not trivially possible to do both, I will go for CC and possibly write a SYSEX -> CC converter module.
won't this cut it as a trivial way?

LXNDR1 wrote:PatchMems you set to sysex receipt but for MIDI you use Kelly Lynch's MIDI modules and let select CC messages from lists, which will mean that there merely won't be MIDI learn

I have to experiment with those modules.

I think, if I stick to fixed CC's for the parameters, then having a Sysex->CC translator, indeed, is quite trivial. I have to see how it works with MIDI learning. I am planning to use Dave Haupts Controlmeister.

I will probably give it a try tonight.

Cheers,
Martin

Schiffbauer · Post by **Schiffbauer** » Wed Feb 23, 2011 12:50 pm

Hi,

some people asked me to share soundbanks for the JX - but the final vsti is
not ready and before it makes no sense. The other thing is, all my banks
are not sysex, so it is difficult to share. I wrote all my patches in paper datasheets and it looks like this:

I must wait for the final vsti, to recreate all (no - not all), some
banks for the masses.

But i've made a new sounddemo with some patches, I've played from my
JX-Collection

http://www.divshare.com/download/14142458-b62

They are named now (as played) for the upcoming soundbanks:
Wave-Gull (don't know it it was in factory yet); VangStringPad1;
Synth Bells 1; Synth Bells 2; Synth Bass 9; Synth Bass 10; Synth Drum 3;
Synth Organ 1; Pulse 4; Digital 16; Elements 3; Elements 7; Elements 5; Elements 2; Axel F, Prophet 8; Synth Piano 7; Synth Organ 3; Synth Organ 6; Synth Guitar 4,Aua; ElectroTouch; Atem (Breath); Synth Strings 7;
Synth Strings 5.

ciao
armin (aka Kujashi)

LXNDR1 · Post by **LXNDR1** » Wed Feb 23, 2011 2:37 pm

Schiffbauer wrote: ...

i can only imagine how much time and work went into programming these patches so i'd say this is really precious and invaluable

a VST, and a free one, with such a massive collection of authentic hardware patch banks is a unique thing, a combination you can only wish for

what you offer Armin is greatly commendable

Kriminal · Post by **Kriminal** » Wed Feb 23, 2011 5:27 pm

you havent heard the patches yet

im just kidding, im sure they are very good

martin_l · Post by **martin_l** » Wed Feb 23, 2011 6:03 pm

Schiffbauer wrote:Hi,

some people asked me to share soundbanks for the JX - but the final vsti is
not ready and before it makes no sense. ...

Thanks again for the generous offer of making some soundbanks!! Maybe, once I have the final version, you could post the scanned images somewhere, and we could distribute the 'coding' of the presets. I would be happy to take a share in the work.

Cheers,
Martin

martin_l · Post by **martin_l** » Wed Feb 23, 2011 6:14 pm

Just a quick update on the plug...

I fine tuned the envelopes and filter settings. Should fix that the synth sounded a bit too harsh and release sounded too long. I am nearly done with the key-assignment, i.e. normal poly, round-robin, unison1, unison2 (octave shift), mono, and mono with all voices in unison. Furthermore, I automatically determine the oversampling depending on the hosts sample rate. The auto-oversampling still caused a crash, which I hope to find soon.

I aim to release another beta with these updates by the weekend, and then want to concentrate on MIDI controllers (possibly with Midi learn) and some improvements to the GUI.

Stay tuned...

Martin

Schiffbauer · Post by **Schiffbauer** » Wed Feb 23, 2011 7:23 pm

Martin wrote:

Thanks again for the generous offer of making some soundbanks!! Maybe, once I have the final version, you could post the scanned images somewhere, and we could distribute the 'coding' of the presets. I would be happy to take a share in the work.

No, it's too much work for scanning and posting every paper datasheet. Instead when the final is ready, I will make about 10 soundbanks from these at first and releasing furthermore banks step by step every once in a while.

ciao
armin (aka Kujashi)

Sound Mechanics · Post by **Sound Mechanics** » Wed Feb 23, 2011 7:28 pm

Looking forward to hearing the developments. Midi learn always makes life easier with setting up controllers

luckie · Post by **luckie** » Thu Feb 24, 2011 4:00 am

First KVR post!
Martin: I think you have done an outstanding job with this. I actually registered on KVR so I could give proper cudos! It has a very nice sound (oscillators are completely alias free) and it isn't buggy! The slider inputs are nicely smoothed and it doesn't make nasty clicks even when reassigning voices.

I hope you make it freeware! ... It will be one of the best freeware VSTi's out. especially with a nice premade preset library!

martin_l wrote: About being a 100% accurate emulation: I think this will never be possible. For instance, the DCO's in the JX-8P run on a 8MHz clock, which is diveded down, and gives a certain set of frequencies and leads to fine detunings. In order to properly reproduce that, I would have to run the engine at ridiculous 8 MHz, which is not possible.

Hmmm. I've never coded a VSTi, but I thought that would be pretty easy. if the JX-8P is counting an 8MHz clock to generate the notes, that just means that the frequencies are 8MHz divided by an integer.
For a normal equitemper scale the frequency for a midi note number MN is
FreqET = 440Hz * 2^((MN-69)/12);
The 8MHz count to get that frequency is 8000000/FreqET
so the "8MHz count tempered" scale would be
Freq8MT = 8000000/round(8000000/FreqET);

The only question is if they did it that way... The worst case deviation would be about a quarter of a cent which is getting hard to hear.

Anyway, I really love your VSTi martin. looking forward to updates!
-Luckie

dalor · Post by **dalor** » Thu Feb 24, 2011 4:51 am

luckie wrote:Hmmm. I've never coded a VSTi, but I thought that would be pretty easy. if the JX-8P is counting an 8MHz clock to generate the notes, that just means that the frequencies are 8MHz divided by an integer.

The 8MHz refers to the CPU (used for IO such as Midi input) but not to generate the sound. The CPU controls the analog sound circuits but doesn't generate the sound. It's an analog synth.

luckie · Post by **luckie** » Thu Feb 24, 2011 6:44 am

dalor wrote:The 8MHz refers to the CPU (used for IO such as Midi input) but not to generate the sound. The CPU controls the analog sound circuits but doesn't generate the sound. It's an analog synth.

I don't think that is right. I thought the DCO's generated the waveform by creating a trigger sequence off the fast clock, and using that to reset the ramp. So it is still analog, but stable (unlike most VCO based synths).
I've never had a JX-8P, but I did have a Juno106. That was also a DCO synth.

wwjd · Post by **wwjd** » Thu Feb 24, 2011 4:07 pm

again, martin, thanks for your work bringing this one too us.

I THEEENK you guys were talking about this: when playing longer notes likes in the SOUNDTRACK patch, switching notes quickly kinda drops you in the middle of the sweep which makes the notes jump out louder and higher with no attack time. maybe the real one did that too? Maybe this was part of what you were discussing about triggers and unisons etc

When the next beta comes out, I'm happy to test the IDIOT-PROOFness of assigning controller knobs, at least in REAPER, on my Oxygen 49

dalor · Post by **dalor** » Fri Feb 25, 2011 5:04 am

luckie wrote:I've never had a JX-8P, but I did have a Juno106. That was also a DCO synth.

Sorry, of course you are right, JX-8P is a DCO synth! I could have sworn the Juno106 is also a VCO synth.. I have a waveform collection of the JX-8P and Juno106 and they look very 'analog' to me - that's why I thought they are VCO synths. I guess you can't judge by looking at a waveform recorded from an analog output.

martin_l · Post by **martin_l** » Fri Feb 25, 2011 9:06 am

wwjd wrote:again, martin, thanks for your work bringing this one too us.

I THEEENK you guys were talking about this: when playing longer notes likes in the SOUNDTRACK patch, switching notes quickly kinda drops you in the middle of the sweep which makes the notes jump out louder and higher with no attack time. maybe the real one did that too? Maybe this was part of what you were discussing about triggers and unisons etc

It was not exactly what was discussed, but it is related to that.
Indeed, the behaviour you describe, is the same in the hardware. The reason for it is, that the envelop generators are not reset, when a key is pressed, but they pick up someway down the release. Where they are depends on how long ago the voice has been released, and how long the release time is.

What was discussed about the different key-assignment modes is, which voice will be allocated next, when a new key is pressed. In the 'standard' mode there is a note-memory, i.e. a voice remembers which note it played last, and if that key is played again, the same voice will be re-used.

In contrast to that, the so-called 'round-robin' mode, always uses a new voice, even if another voice is still playing the release phase of the same note.

I hope that this was at least half-way understandable...

When the next beta comes out, I'm happy to test the IDIOT-PROFness of assigning controller knobs, at least in REAPER, on my Oxygen 49

New beta with MIDI learn should come soon... I hope I manage this weekend. If you could test it, it would be great. Also, I would like some feedback whether the automatic oversampling works in all hosts. I now determine the oversampling dependent on the samplerate, and the secondary oversampling in the DCO's according to the pitch of each note, so that high oversampling is only done, when necessary.

Cheers,
Martin

wwjd · Post by **wwjd** » Fri Feb 25, 2011 4:32 pm

Understood. Making it like the original is fine by me, just wanted to point it out in case it was a glitch. Looking forward to your next beta. Will you be recreating the factory presets from the original hardware possibly?

PG8X (inspired by the JX8P): new beta version uploaded