Reversed CC 74 (cut-off) response

dialcee

Hi, I am a new SBLive user, I just bought one 5.1 card a few days ago. I think the possibilities opened by Kx drivers are amazing and I wish to give a big THANK YOU to developers because they deserve it.

Being new to the Creative and soundfonts world I'm not sure that what I'm pointing out here is a bug or not. (until now I played around with an yamaha XG card). I like to play a lot with the cut-off controller (AKA brightness) realtime and here is what I found: with creative drivers nothing hapens at all With Kx driver the cut-off value is modified realtime but the value of cut-off frequency is decreasing when the CC 74 value increases. This gives reversed results with my old files.

By the way: I am very disapointed by the quality of this effect, the brightness of the sound is varying very un-smooth, very coarse and gives some realy ugly artifacts, amplified by the reverb. I expected much higher quality from such a well rated card. (I fact with my old yamaha ymf 724 I got similar results, the only midi synthesiser that worked corectly for me - regarding this aspect - was yamaha's softsynth SXG 100).

chromal

I believe that the Creative implementation only allows realtime filter changes with NRPN messages. Needless to say, they are a huge pain to use, and using the continuous controller offset approach is a huge improvement. I beleive that CC74 is also used by XG for "brightness", maybe others MIDI standards as well?

A larger CC value for a lower cutoff frequency is a little counter-intuitive, although the rationale may have been that when zeroing controllers, voices wouldn't be LPF'd into oblivion? A toggle option to reverse this would be welcome, yes, mostly because the current implementation has been causing me huge problems when collaborating MIDIs with my XG-using friends. They send something to me; I convert the CC messages, send it back to them. They revert it to something that works on their hardware, send it back to me, etc.


chromal

Eugene Gavrilov

I'll review this issue

/E

dialcee

Thanks for looking into it.

I tried to obtain realtime cut-off variation with creative drivers using NPRN's. It worked, I mean the cut-off frequency increased with the controller value. Other than that the result obtained was horrible and useless for any purpose, much worse than with KX drivers. It's a bad joke. The variation should be continue (or perceived that way), instead is made jumping from one value to another
I just hope that KX drivers will get better and better.