Tal U-NO-60
-
- KVRist
- 194 posts since 2 May, 2007 from Swiss
Modeler wrote:Crikey, that was a quick turn-around on the chorus. It was like "I don't think I can do a chorus... Oh wait, hang on... How's this?"
It was the eternal fight against aliasing and the stereo phase scope
I have thought about it ...i think next year i will do that.Modeler wrote:Togu - have you thought about entering the developer challenge?
I thank you all for your awesome comments. There are still some tiny bugs to fix
TAL Software GmbH
https://tal-software.com/
https://tal-software.com/
-
- KVRAF
- 2249 posts since 6 May, 2003 from rat city au
~20% CPU usage here - on a p4 2.4, FLStudio 7, one voice.
rises to anywhere btwn 50 and 70% with multiple voices and chorus... Needs some optimisation, perhaps?
cheers, sk
rises to anywhere btwn 50 and 70% with multiple voices and chorus... Needs some optimisation, perhaps?
cheers, sk
-
- KVRist
- 194 posts since 2 May, 2007 from Swiss
Hi sk,stk wrote:~20% CPU usage here - on a p4 2.4, FLStudio 7, one voice.
rises to anywhere btwn 50 and 70% with multiple voices and chorus... Needs some optimisation, perhaps?
its already optimized with assembler and approximations. But maybe i will find a way for other optimations. The filter uses(2xoversampled) 32 tanh approximations to get his characteristic sound.
On my Intel Duo Core 2.666 GHz it looks a bit better(Cubase process in the Task Manager):
Standalone Chorus: 0-2%
One Voice with Chorus: 2-5%
Two Voices: 4-7%
There is a CPU issue with long release times. This will be fixed in the next update...
TAL Software GmbH
https://tal-software.com/
https://tal-software.com/
-
- Banned
- 1842 posts since 4 Aug, 2004 from just right here
Thanks, I can use this one....yes 
-
- KVRist
- 194 posts since 2 May, 2007 from Swiss
In the current version, this delay works only for the filter LFO. I will fix that... i hope you mean the pitch LFO delay?cleverr1 wrote:Just one small thing - the LFO delay setting doesn't appear to work in Cubase SX 3.1.
Thanks for this input.
TAL Software GmbH
https://tal-software.com/
https://tal-software.com/
-
- KVRian
- 597 posts since 29 Nov, 2005
Yes, it's just the DCO pitch that's affected.ToguAudioLine wrote:In the current version, this delay works only for the filter LFO. I will fix that... i hope you mean the pitch LFO delay?cleverr1 wrote:Just one small thing - the LFO delay setting doesn't appear to work in Cubase SX 3.1.
Thanks for this input.
-
- KVRAF
- 2249 posts since 6 May, 2003 from rat city au
hmm okay, maybe denormalling, or a FLStudio problem? i dunno. Crazy cpu use was on the presets with little / no release time too.ToguAudioLine wrote:Hi sk,stk wrote:~20% CPU usage here - on a p4 2.4, FLStudio 7, one voice.
rises to anywhere btwn 50 and 70% with multiple voices and chorus... Needs some optimisation, perhaps?
its already optimized with assembler and approximations. But maybe i will find a way for other optimations. The filter uses(2xoversampled) 32 tanh approximations to get his characteristic sound.
On my Intel Duo Core 2.666 GHz it looks a bit better(Cubase process in the Task Manager):
Standalone Chorus: 0-2%
One Voice with Chorus: 2-5%
Two Voices: 4-7%
There is a CPU issue with long release times. This will be fixed in the next update...
cheers, sk
-
- KVRist
- 194 posts since 2 May, 2007 from Swiss
I cant find any problem with FL Studio. I try to optimize the synth more...stk wrote:hmm okay, maybe denormalling, or a FLStudio problem? i dunno. Crazy cpu use was on the presets with little / no release time too.
U-NO-60 uses a seperate filter for every voice, like the original. This makes that the CPU rises nearly proportional to the playing voices.
TAL Software GmbH
https://tal-software.com/
https://tal-software.com/
-
- KVRAF
- 8519 posts since 7 Apr, 2003
are there denormal checks/fixes/workarounds? us p4 owners (i have a p4/1.9ghz at the moment) have probs with those.ToguAudioLine wrote:Hi sk,stk wrote:~20% CPU usage here - on a p4 2.4, FLStudio 7, one voice.
rises to anywhere btwn 50 and 70% with multiple voices and chorus... Needs some optimisation, perhaps?
its already optimized with assembler and approximations. But maybe i will find a way for other optimations. The filter uses(2xoversampled) 32 tanh approximations to get his characteristic sound.
On my Intel Duo Core 2.666 GHz it looks a bit better(Cubase process in the Task Manager):
Standalone Chorus: 0-2%
One Voice with Chorus: 2-5%
Two Voices: 4-7%
There is a CPU issue with long release times. This will be fixed in the next update...
-
- KVRAF
- 2042 posts since 22 Nov, 2003 from Mars, Solar System
-
- KVRist
- 194 posts since 2 May, 2007 from Swiss
There is a workaround for denormalisation problems in the filters. I'm trying to improve the performance. May i can do some more assembler optimationsgrymmjack wrote:are there denormal checks/fixes/workarounds? us p4 owners (i have a p4/1.9ghz at the moment) have probs with those.
TAL Software GmbH
https://tal-software.com/
https://tal-software.com/
-
- KVRist
- 194 posts since 2 May, 2007 from Swiss
Version 1.0.3 is out now, with important improvements:
Changes in Version 1.0.3
* CPU optimations (17%)
* Fixed denormalisation at long release times
* LFO delay works now also on the pitch
Changes in Version 1.0.3
* CPU optimations (17%)
* Fixed denormalisation at long release times
* LFO delay works now also on the pitch
TAL Software GmbH
https://tal-software.com/
https://tal-software.com/

