Well, yeah... but I'm sure I'm using the multithreaded ones here... will check... (currently working on the Mac...)Mike Janney wrote:I just don't use any external libraries
Not even any standard C/C++ library?
Later,
UrsZebra2 Latest Version: 2.1 - Jan 19 2007
-
- u-he
- Topic Starter
- 30209 posts since 8 Aug, 2002 from Berlin
Later,
Urs-
- u-he
- Topic Starter
- 30209 posts since 8 Aug, 2002 from Berlin
Yeah, got the \MT switch... that can't be it...Urs wrote:Well, yeah... but I'm sure I'm using the multithreaded ones here... will check... (currently working on the Mac...)Mike Janney wrote:I just don't use any external libraries
Not even any standard C/C++ library?
Urs-
- KVRian
- 864 posts since 4 Apr, 2001 from Finland
This is actually very hard to pin down by testing. They just happen to go in a "state" somehow. Like, I now have this 9-instance loop that started going bad when choosing certain patches for the last Z2 instance. CPU jumped to max instantly after I selected some patches (for some others not).
After I closed and started eXT again with this same setup, I can choose and change those same patches without problem.
It simply goes into a tricky "state" at some point...
After I closed and started eXT again with this same setup, I can choose and change those same patches without problem.
It simply goes into a tricky "state" at some point...
-
- u-he
- Topic Starter
- 30209 posts since 8 Aug, 2002 from Berlin
Arpeggiator bug - fixed
Pooha... need a break now. Let's see if the concept of food/eating is still part of my knowledge...
Afterwards I might compile new versions for public testing to get approval for the bugfixes. Maybe fix the weird clicks-on-preset-change bug before that. Will see...
Urs
Pooha... need a break now. Let's see if the concept of food/eating is still part of my knowledge...
Afterwards I might compile new versions for public testing to get approval for the bugfixes. Maybe fix the weird clicks-on-preset-change bug before that. Will see...
Urs-
- KVRist
- 370 posts since 13 Jul, 2003 from Berlin
Not even any standard C/C++ library?
As a matter of fact: he's not even using your computer, Zebra is processed entirely through networked idle PCs worldwide...
Sorry... I'm in silly mode.
As a matter of fact: he's not even using your computer, Zebra is processed entirely through networked idle PCs worldwide...
Sorry... I'm in silly mode.
-
- KVRian
- 864 posts since 4 Apr, 2001 from Finland
Yeah, bon appetit!
I don't like telling these annoying things, but I guess it has to be done.
Don't be mad at us!
It would be nice to get a new version for some testing tonight still, if you have the energy. I'll go visit my parents for the weekend tomorrow, so I can't test anything for a while then.
I don't like telling these annoying things, but I guess it has to be done.
Don't be mad at us!It would be nice to get a new version for some testing tonight still, if you have the energy. I'll go visit my parents for the weekend tomorrow, so I can't test anything for a while then.
-
- KVRAF
- 2582 posts since 24 Apr, 2003 from Canada
Hi Urs,
I tried posting this to the secret forum but there are some problems over there. soooo:
But what I'm saying is if I load the Vintage2 filter in the fxgrid *with no modulations whatsoever* there is alot of clicking.
I tried posting this to the secret forum but there are some problems over there. soooo:
Ok. I understand.u-he wrote: Yes, like all modulations in the FX (i.e. by ModMatrix) the last voice takes over... which *can* cause clicks and pops at times... so use that feature with caution... -> global LFOs or single triggered MSEGs work well, for instance
But what I'm saying is if I load the Vintage2 filter in the fxgrid *with no modulations whatsoever* there is alot of clicking.
-
- u-he
- Topic Starter
- 30209 posts since 8 Aug, 2002 from Berlin
We'll see... I'm fully braindeaded right now, so I won't dare merging sources from Mac and PC. Phone kept me from eating, I'll try to do that now, while watching another episode of Criminal Intent and see how it goes when the sun has vanished in 45 minutes...
Later,
Urs
Later,
Urs-
- KVRian
- 864 posts since 4 Apr, 2001 from Finland
Hey, just an idea: could others do some heavy testing with multiple instances in other hosts than energyXT?
It's not a surprise if it crackles if the CPU is at 100% but it shouldn't start to behave erratically, me thinks. I think that I'll try some with Logic.
Oh and Urs, you seem to have fixed many things today already. Thanks for that!
It's not a surprise if it crackles if the CPU is at 100% but it shouldn't start to behave erratically, me thinks. I think that I'll try some with Logic.
Oh and Urs, you seem to have fixed many things today already. Thanks for that!
-
- KVRian
- 530 posts since 27 Oct, 2004
I own eXT and Podium. I can test multiple instances of Z2 in Podium, but later, probably tomorrow.spritex wrote:Hey, just an idea: could others do some heavy testing with multiple instances in other hosts than energyXT?
-
- KVRian
- 864 posts since 4 Apr, 2001 from Finland
Hmm. From my initial testing, as long as the process buffer setting in Logic is not "small", the playback simply stops if it can't handle it. All is well after pressing play again. I haven't managed to get it in the "sorry state".
I think that part of the problem is that Zebra2 tends to be pretty wild regarding CPU usage. Playing a simple loop with several sounds playing all the time can seemingly randomly vary the CPU usage from 40% to 100%. (Actually, I am not talking about real CPU usage, but the meters in energyXT or Logic which cause problems when at full. I wonder what they are really showing since the Windows CPU meter doesn't show nearly as high numbers.)
Maybe we should wait for the Shaper/Ring denormals and OSC fixes before drawing further conclusions? None of these are in the current release yet, right?
I understood that the OSC fix might take a while?
I think that part of the problem is that Zebra2 tends to be pretty wild regarding CPU usage. Playing a simple loop with several sounds playing all the time can seemingly randomly vary the CPU usage from 40% to 100%. (Actually, I am not talking about real CPU usage, but the meters in energyXT or Logic which cause problems when at full. I wonder what they are really showing since the Windows CPU meter doesn't show nearly as high numbers.)
Maybe we should wait for the Shaper/Ring denormals and OSC fixes before drawing further conclusions? None of these are in the current release yet, right?
I understood that the OSC fix might take a while?
Last edited by spritex on Thu Jul 27, 2006 6:19 pm, edited 1 time in total.
-
- KVRian
- 743 posts since 14 Apr, 2004
I will try this thing with FLStudio 6.0.8 (..and 6.1 beta to have some fun!)kangul wrote:I own eXT and Podium. I can test multiple instances of Z2 in Podium, but later, probably tomorrow.spritex wrote:Hey, just an idea: could others do some heavy testing with multiple instances in other hosts than energyXT?
-
- KVRAF
- 26963 posts since 3 Feb, 2005 from in the wilds
On the Mac, AU version I get crackles when playing notes fast on some presets... CPU use no more than 30% and audio output well in the green and it crackles... I have not noticed a particular pattern to account for all crackles, but the shapers seem to be problematic. That is not only it hough as presets without shapers will also crackle...
-
- KVRist
- 75 posts since 5 Mar, 2004 from London, UK
Don't know if this question is still outstanding or even still matters but the answer is no, I don't have this problem w/ ACID.raikard233 wrote:The same thing [sequencer loop mode problem] can be tested on Acid...it happens the same?
-
- KVRian
- 743 posts since 14 Apr, 2004
First loading of FLStudio6.0.8, 8 Zebra2 instances loaded, inserting notes on piano roll for Z2#1, Z2#2, Z2#3, Z2#4, Z2#5 (with loop play), changing a patch on Z2#5 caused this thing: some patches outputs sound and other patches don't emits any sound.raikard233 wrote:I will try this thing with FLStudio 6.0.8 (..and 6.1 beta to have some fun!)kangul wrote:I own eXT and Podium. I can test multiple instances of Z2 in Podium, but later, probably tomorrow.spritex wrote:Hey, just an idea: could others do some heavy testing with multiple instances in other hosts than energyXT?
All the remaining instances were working without problems.
I've saved the project file then I've opened it with FLStudio 6.1beta and I haven't found problems. All ok.
Closed and reopened FLStudio 6.1beta I have redo the loading of the 8 Zebra2 and the drawing of the notes on the pianoroll and I've tried to change the patches on all Zebras instances, no problems found.
All this with Zebra2 25_07_2006
With Zebra2 26_07_2006 and FLStudio in both versions i haven't found problems.
The thing is that I've found problems on patch loading only with the first loading of FLStudio 6.0.8 using Zebra2_25_07_2007, then closing and reopening FLStudio, redoing the procedure, I haven't found problems.
Never saw cpu spikes on patch changing.
AMD X2 3800+