Extreme voice stealing problem (3.0.2)

[AntonySharmman]

Yes, that's exactly as expected. (Polyphony is consuming 8 *voices,* not 8 notes.) In this case, 1 note = 8 voices.
If it worked as you're suggesting it should, then the Program wouldn't sound correctly!

Got it !

I believe that your concerns are about voice-stealing in general - not related to Timbre Priority or Max # of notes.

Indeed , and I just try to figure out and understand why in OS3 an additional 20% voice stealing occurs !

Thanks

[arne v]

Hi,
another "voice stealer" is the piano string resonance. If you have a combi with piano using string resonance you will see the difference when turning on/off

[GregC]

Hi,
another "voice stealer" is the piano string resonance. If you have a combi with piano using string resonance you will see the difference when turning on/off

Thats likely if Karma is playing many parts/parts on that piano with the string resonance.

When I played several Berlin Programs, string resonance on 127, I could not even get the Voice stealing meter to budge. Forced it with heavy sustain, big /many chords. No voice stealing. On zero/0.

feel free to name the Combi

[arne v]

That too, but i can also see it when karma is not active.
btw im having the 1st version of Kronos 61

Combi is something i made

[GregC]

That too, but i can also see it when karma is not active.
btw im having the 1st version of Kronos 61

Combi is something i made

ok, likely your piano based ( ?) combi is demanding of voices.

If you have polyphony on all programs, I would not be concerned.

[rob_tky]

I repeat once again that I wrote before some posts !
Voice stealing is a matter of polyphony limitation regardless of CPU & OS

Yes Antony, though isn't the polyphony limit there in the first place because of power of CPU at hand?

I won't go into discussion why rather low power CPU was used in this equipment. I believe Korg did necessary analysis, considering average processing needs, power consumption (=>heat + fans), cost (though I personally wouldn't mind paying extra in this case), etc.

2 suggestions though.
- starting from Kronos 3, maybe offer a choice of hardware for Kronos, eg regular MB and power MB for polyphony warriors
- provide an optional 'link' of Kronos with PC/Mac to allow borrowing processing power when needed.

[StephenKay]

As I wrote, Timbre Priority does not apply to sounds being played directly from the keyboard.

Are you playing the sounds directly from the keyboard?
Are the Timbres on the Global Channel?
If so, Timbre Priority does not matter. All such sounds will have high priority.

Hi Dan,
Sorry to wake up an old thread, but back when you wrote this, I scratched my head because I thought I could swear that, in the past, I had needed to use Timbre Priority to keep sounds being played on the keyboard from being stolen.

Today, while working on finalizing the new Catalyst Vol. 2 sound set, I did in fact find that this checkbox does affect timbres being played on the keyboard and on the Global Channel.

For example, in a particular combi, I have two pad sounds in the Left Hand area on GCh. You can latch these with the sustain pedal.

So, I noticed that when I latched a left hand chord with the sustain pedal, and released the keys so that the pad notes were sustaining, the KARMA groove would eventually steal the notes.

When I turned on Timbre Priority for these two timbres, the voice stealing went away. So it does indeed seem to function on the voices played from the keyboard. To be accurate, the left hand voices would not be stolen as long as I continued pressing the keys; once I released them, however, and the notes are being latched by the damper, they can be stolen and using the Timbre Priority checkbox in this case solves the problem.

[esch]

why can't we completely turn off several midi tracks in a combi that we're not ising, would this capability npt solve the problem or mitigate it? especially when those midi tracks are demanding for example using karma or sustained notes based on what we're playing?-arent those channels using voices unnecessarily?

is it possible to do this? in any discussion arpund vpoice stealing i'v never seen that as a possibility.


what about programming keyboard zones to nothing for those tracks? would the unused tracks still be using resources?

you guys know a lot more than i, sorry if my ideas are simplistic -i'm a simpleton as you can see from my terrible tablet grammar and spelling...

[esch]

tried these things and seem to have solved it with a combi i was having trouble with but still would like your comments and will try more tomorrow if my efforts arent considered futile .
d

[AntonySharmman]

When I turned on Timbre Priority for these two timbres, the voice stealing went away. So it does indeed seem to function on the voices played from the keyboard. To be accurate, the left hand voices would not be stolen as long as I continued pressing the keys; once I released them, however, and the notes are being latched by the damper, they can be stolen and using the Timbre Priority checkbox in this case solves the problem.

Just saw that , Stephen I already knew and wrote that in my previous post , thanks you can confirm it !
This is the only way at present moment to limit OS 3 voice stealing and has been applied to my commercial Kronos External Sound Libraries !

[JPROBERTLA]

As I move further along with my Kronos2-88, I have had to create work-arounds for the note stealing associated with 3.0.2. I don't experience this when playing live, but in sequencer mode, and playing along with the sequence, the note stealing is not as good as it was in previous OS's. Organs and acoustic pianos are the usual culprits, so I have had to dumb-down some programs for theses voices for use in the larger sequences where these parts are not as dominant and/or I have to play that part live.

I don't understand why this has become a problem. I have used the same programs as with the original Kronos and did not have voice stealing issues. I did have polyphony problems, but not voice stealing hickups. I hope that this is corrected or made better in a future OS update.

[NormC]

If one was lucky enough to own 2 Kronosses you could double the polyphony like you could with The Oasys. But then again you have to have 2 of them.

This is why a rack module version would be great.

[chini]

If one was lucky enough to own 2 Kronosses you could double the polyphony like you could with The Oasys. But then again you have to have 2 of them.

This is why a rack module version would be great.

Considering how long it's taking Korg to sort this problem out I would not be surprised if a software update may not be enough to resolve the issues OS3.02 unfortunately bestowed on both original Kronos owners like myself and new buyers of Kronos 2. Even though it is possible to switch off the features added to the Kronos in OS 3.02 to theoretically release more power available for the chipset I suspect that it is OS3s re-written architecture that may well be too convoluted to unravel into a better performance re voice steeling.

However, on a more positive note, the possible trade off from this may well be a welcome release of a rack version that many have been hankering after from day one!

The OP has unfortunately discovered this anomaly just as I did back in december when I upgraded my original version Kronos (OS ver 2.06). I still have no news from the guy I spoke to at length at Korg in the UK on this important issue. The glaring evidence of this voice steeling for me after having installed v3 was when patch changing from one particular custom combie of mine to another that has a drum track triggered by note on. All notes and effects tails of the first patch get cut off and the combie patch I change to even abruptly halts playing its drum track just after having been triggered to start!

We can only wait and see what Korg do but I hope they both find a solution to this in an upgrade AND finally release a KRONOS RACK to boot!