Cutoff and Resonance feedback

erikwelander · May 9, 2003, 08:43 PM

Great work on the driver which is shaping up quite nicely.

I would only give some feedback on the realtime controls I tested. The cc74 for cutoff works great in the 3532 release, but the Resonance Q -- cc71 control still needs some work in order to work realtime...

**Eugene Gavrilov** · May 10, 2003, 12:53 AM

unfortunately, this seems to be a hardware limitation - the resonance filter cannot be changed dynamically without some audio artifacts
anyway, 3533 will include some updated code and a new MIDI implementation chart
(for reference

)

/Eugene

erikwelander · May 10, 2003, 06:26 PM

Yes, Soundfonts still sound great in the latest release so you have most definitely made progress.

Regarding realtime support for resonance may be found in some VST Soundfont player, but hardware is still hardware and may yet prove to produce superior sound quality! (I tried some player that did not reproduce the sounds correctly).

Would this be true (no realtime resonance) even in the latest Audigy cards?

**Eugene Gavrilov** · May 11, 2003, 03:29 AM

well, realtime resonance functionality should be identical for all the 10k1/10k2-based cards

/Eugene

erikwelander · May 19, 2003, 09:54 PM

Hi Again; regarding the resonance we discussed I have done some testing with the Live! and the latest Creative drivers for the Live! (5.12.2.252) and it seems as if resonance is possible to use in realtime:

I was using the NPRN from this page (Soundfont 2.0 spec):

http://grassomusic.de/english/nrpns.htm

And the results can be listened to in the mp3 file:

http://medlem.spray.se/erikwelander/resolive.mp3

Best Regards from Erik

**Eugene Gavrilov** · May 20, 2003, 01:52 AM

ok, I'll check that

/E

Licho · Jun 7, 2003, 02:14 AM

Hi Eugene!. Some time ago, when I had to use APS drivers, I use to play a lot with resonance changes in realtime, and there were no sound artifacts, so I think it's not a hardware limitation. By the way, your work ROCKS!

Max M. · Jun 7, 2003, 09:15 PM

>when I had to use APS drivers, I use to play a lot with resonance changes in realtime - don't you mix up anything? As i remember Cuttoff Freq can be modulated in real-time, while Resonance cannot (e.g. when you change resonance amount it applies to notes playing after but not to currently played notes - this is how it was APS drivers) - this is really hardware limitation...

Licho · Jun 11, 2003, 05:01 PM

Sorry Max, you must be right. It's been more than one year since I'm not using APS drivers, and my memory must confused me.

By the way, in my PC with v3533b, the vibratos are much faster than in v3531. Have you noticed the same?

Thanks again:
Lisandro Puzzolo

**Eugene Gavrilov** · Jun 15, 2003, 01:40 AM

concerning modulation issue

--

according to Creative's SoundFont 2.01 specs (article 8.8.4), the modulation controller affects 'Vibrato LFO to pitch' in the following way: when CC=127, VibLFOFreq is added 50 cents

in order to check that kX implements this behaviour correctly, simply test the following:
set CC1=0, VibLFOFreq=50 --> check the sound
set CC1=127, VibLFOFreq=0 --> check the sound
(it should be identical)

so, either our formula for converting VibLFOFreq into hardware-dependent value is incorrect (there's no 'official' specification for the hardware registers), or Creative drivers don't follow their own SoundFont 2.01 specification...

btw, pre-3533 release uses a value of '18 cents' instead (based on the linux AWE driver)

/Eugene

Licho · Jun 21, 2003, 07:08 AM

Hi Eugene. I really apreciate all of your work. But really, the modulation frequency is very very fast. Right now, I can't use it because the instruments sound funny.

I believe you are doing what creative says, but we know the kind of people creative has on it's team. ;-)

So, they've surely documented the parameters in the wrong way. I have lots of soundfonts and all of them become unusable on their modulation frequency.

I hope you can fix it.

Thanks again:
Lisandro Puzzolo

**Eugene Gavrilov** · Jun 23, 2003, 02:41 AM

I suggest you help me implementing the correct modulation parameter by finding out the correct modulation coefficient

in the Vienna editor, tweak the 'ModLFO to Pitch' and 'VibLFO to Pitch' parameters
find out the ones that to your opinion should correspond to 'Modulation CC' = 127 (maximum)
post your results

btw, both LFOs can be used for simulating 'modulation' -- we should choose which one (or both) and choose the particular multiplier

/Eugene

May 9, 2003, 08:43 PM	#1
erikwelander DriverHeaven Newbie Join Date: May 2003 Posts: 5 Rep Power: 0	Cutoff and Resonance feedback Great work on the driver which is shaping up quite nicely. I would only give some feedback on the realtime controls I tested. The cc74 for cutoff works great in the 3532 release, but the Resonance Q -- cc71 control still needs some work in order to work realtime...

Jun 7, 2003, 02:14 AM	#7
Licho DriverHeaven Junior Member Join Date: Dec 2002 Location: Rosario - Argentina Posts: 44 Rep Power: 0	APS drivers Hi Eugene!. Some time ago, when I had to use APS drivers, I use to play a lot with resonance changes in realtime, and there were no sound artifacts, so I think it's not a hardware limitation. By the way, your work ROCKS!

Jun 7, 2003, 09:15 PM	#8
Max M. h/h member-shmember Join Date: Dec 2002 Location: Evil Empire Posts: 2,606 Rep Power: 66	>when I had to use APS drivers, I use to play a lot with resonance changes in realtime - don't you mix up anything? As i remember Cuttoff Freq can be modulated in real-time, while Resonance cannot (e.g. when you change resonance amount it applies to notes playing after but not to currently played notes - this is how it was APS drivers) - this is really hardware limitation... __________________ Last edited by Max M.; Jun 8, 2003 at 03:03 PM.

Jun 21, 2003, 07:08 AM	#11
Licho DriverHeaven Junior Member Join Date: Dec 2002 Location: Rosario - Argentina Posts: 44 Rep Power: 0	modulation frequency is not acurate. Hi Eugene. I really apreciate all of your work. But really, the modulation frequency is very very fast. Right now, I can't use it because the instruments sound funny. I believe you are doing what creative says, but we know the kind of people creative has on it's team. ;-) So, they've surely documented the parameters in the wrong way. I have lots of soundfonts and all of them become unusable on their modulation frequency. I hope you can fix it. Thanks again: Lisandro Puzzolo

May 10, 2003, 12:53 AM	#2
Eugene Gavrilov kX Project Lead Programmer and Coordinator Join Date: Dec 2002 Posts: 3,106 Rep Power: 73	unfortunately, this seems to be a hardware limitation - the resonance filter cannot be changed dynamically without some audio artifacts anyway, 3533 will include some updated code and a new MIDI implementation chart (for reference ) /Eugene

May 10, 2003, 06:26 PM	Thread Starter #3
erikwelander DriverHeaven Newbie Join Date: May 2003 Posts: 5 Rep Power: 0	Yes, Soundfonts still sound great in the latest release so you have most definitely made progress. Regarding realtime support for resonance may be found in some VST Soundfont player, but hardware is still hardware and may yet prove to produce superior sound quality! (I tried some player that did not reproduce the sounds correctly). Would this be true (no realtime resonance) even in the latest Audigy cards?

May 11, 2003, 03:29 AM	#4
Eugene Gavrilov kX Project Lead Programmer and Coordinator Join Date: Dec 2002 Posts: 3,106 Rep Power: 73	well, realtime resonance functionality should be identical for all the 10k1/10k2-based cards /Eugene

May 19, 2003, 09:54 PM	Thread Starter #5
erikwelander DriverHeaven Newbie Join Date: May 2003 Posts: 5 Rep Power: 0	Hi Again; regarding the resonance we discussed I have done some testing with the Live! and the latest Creative drivers for the Live! (5.12.2.252) and it seems as if resonance is possible to use in realtime: I was using the NPRN from this page (Soundfont 2.0 spec): http://grassomusic.de/english/nrpns.htm And the results can be listened to in the mp3 file: http://medlem.spray.se/erikwelander/resolive.mp3 Best Regards from Erik

May 20, 2003, 01:52 AM	#6
Eugene Gavrilov kX Project Lead Programmer and Coordinator Join Date: Dec 2002 Posts: 3,106 Rep Power: 73	ok, I'll check that /E

Jun 11, 2003, 05:01 PM	#9
Licho DriverHeaven Junior Member Join Date: Dec 2002 Location: Rosario - Argentina Posts: 44 Rep Power: 0	Sorry Max, you must be right. It's been more than one year since I'm not using APS drivers, and my memory must confused me. By the way, in my PC with v3533b, the vibratos are much faster than in v3531. Have you noticed the same? Thanks again: Lisandro Puzzolo

Jun 15, 2003, 01:40 AM	#10
Eugene Gavrilov kX Project Lead Programmer and Coordinator Join Date: Dec 2002 Posts: 3,106 Rep Power: 73	concerning modulation issue -- according to Creative's SoundFont 2.01 specs (article 8.8.4), the modulation controller affects 'Vibrato LFO to pitch' in the following way: when CC=127, VibLFOFreq is added 50 cents in order to check that kX implements this behaviour correctly, simply test the following: set CC1=0, VibLFOFreq=50 --> check the sound set CC1=127, VibLFOFreq=0 --> check the sound (it should be identical) so, either our formula for converting VibLFOFreq into hardware-dependent value is incorrect (there's no 'official' specification for the hardware registers), or Creative drivers don't follow their own SoundFont 2.01 specification... btw, pre-3533 release uses a value of '18 cents' instead (based on the linux AWE driver) /Eugene

Jun 23, 2003, 02:41 AM	#12
Eugene Gavrilov kX Project Lead Programmer and Coordinator Join Date: Dec 2002 Posts: 3,106 Rep Power: 73	I suggest you help me implementing the correct modulation parameter by finding out the correct modulation coefficient in the Vienna editor, tweak the 'ModLFO to Pitch' and 'VibLFO to Pitch' parameters find out the ones that to your opinion should correspond to 'Modulation CC' = 127 (maximum) post your results btw, both LFOs can be used for simulating 'modulation' -- we should choose which one (or both) and choose the particular multiplier /Eugene