Bitwig

xbitz · Post by **xbitz** » Mon Feb 06, 2017 6:17 pm

ThomasHelzle wrote:Guys, I have a question:
Some of you know that I always said that running plugins in the audio engine should be faster than running each one in it's own sandbox. (Plugin Preferences -> "only as bitbridge" vs. "every plugin"). It's a long time I did tests but recently I re-did an old test with multiple instances of Diva with the BWS 2 beta.
To my bafflement I found, that I can actually have twice as many tracks (10) with Diva playing 5 notes and a heavy preset with "every plugin" than with "only as bitbridge" (~5).

Could you guys maybe test your projects as well with the two different settings?
Make sure you close and restart Bitwig after changing the setting, you can't switch this on the fly.

Here is my testproject with Diva, that I can just about play without crackles with 48kHz/128 samples on an i7 6core @ 4.1 gHz:
http://www.screendream.de/stuff/Bitwig/ ... .bwproject

I asked the devs on the betalist about this massive difference and hope we can get to the bottom of this.

Cheers,

Tom

2.0b1 win10x64, 9 instances in both cases, seems same for me, I've used Diva 1.4.0x64 ones, so not the ones as were in your project (in another project 34% vs 20% to bitbridge)

ThomasHelzle · Post by **ThomasHelzle** » Mon Feb 06, 2017 6:19 pm

Thanks xbitz. And you did close and restart BWS after changing the setting?
I can recreate the difference in both 1.3.15 and 2.0 Beta 1 here.

Cheers,

Tom

xbitz · Post by **xbitz** » Mon Feb 06, 2017 6:22 pm

ThomasHelzle · Post by **ThomasHelzle** » Mon Feb 06, 2017 6:23 pm

Thanks!

analogo13 · Post by **analogo13** » Tue Feb 07, 2017 12:39 am

ThomasHelzle wrote:Here is my testproject with Diva, that I can just about play without crackles with 48kHz/128 samples on an i7 6core @ 4.1 gHz:
http://www.screendream.de/stuff/Bitwig/ ... .bwproject

A 24-core @2.5 GHz machine can play 24 of your tracks (each plugin) without crackles and 48kHz /128 samples, but I switched the Diva plugins to MultiCore.

With only as bitbridge only 1-2 tracks!

ThomasHelzle · Post by **ThomasHelzle** » Tue Feb 07, 2017 9:53 am

Thanks analogo13 for testing - that is even more crazy than my own result!

I tried setting Diva to multicore, but in my case performance is getting worse (as expected, u-he also warns that when using many instances of Diva in a project, one shouldn't use multicore but let the DAW do the threading).
How many tracks can you run when you disable multicore in Diva?

Let's hope the devs can get to the bottom of this. I linked here from the BWS 2.0 beta mailinglist (the beta behaves exactly the same for me).

Cheers,

Tom

analogo13 · Post by **analogo13** » Tue Feb 07, 2017 10:22 am

You´re welcome, Tom!

Without Diva+MultiCore the BW 2 beta sound is full of crackling, even with 1 track!

I will try your Diva tracks with Live 9.7.1.

Viele Grüße nach Berlin

ThomasHelzle · Post by **ThomasHelzle** » Tue Feb 07, 2017 10:32 am

Fascinating, but maybe it makes sense:
You have many threads but not a very high speed on each individual core (compared to my 4.1 GHz I mean), so if the individual voices are not spread out (using Diva multicore) you get performance problems.
I'm more baffled by the "each plugin" vs. "only as bitbridge" divergence...

Cheers und viele Grüße zurück,

Tom

analogo13 · Post by **analogo13** » Tue Feb 07, 2017 1:58 pm

Maybe, but the CPU clocks are jumping to 2.8 GHz.

With Live your tracks are still more disastrous. Crackling with only 1 track (MultiCore on or off)!

Perhaps it´s a problem with Diva?

Cheers

ThomasHelzle · Post by **ThomasHelzle** » Tue Feb 07, 2017 4:41 pm

Well, in Live you always have plugins running in the core...

I find it especially funny because a while ago, the programmer of the Melda plugins got quite vocal here on how bad it is to do plugin sandboxing and how inefficient it is. He claimed some very deep knowledge on processor architecture and just couldn't be convinced that having each plugin separate from all others is a major blessing for users, at least from the usability side.
Now it looks as if there is some factor in this that suggests it's rather the other way around with performance as well with many cores...

I'll test some other plugins too.

Cheers,

Tom

humanbeingbeing · Post by **humanbeingbeing** » Tue Feb 07, 2017 6:16 pm

I ran this diva test for myself on Bitwig 2 b1 and Ableton Live 9.7.1, again virtually no difference, Live slightly edges Bitwig.. used 10 divas, sample midi clip, 512 samples, Babyface Pro, Windows 10 i7 6700k. There are plenty of other issues that could be happening... I think if anyone here jumps to conclusions without thorough testing, they're making a mistake... every test I've run so far shows little to no difference between Ableton and Bitwig (and Bitwig 2) in terms of CPU efficiency...

as for sandboxing, I'm not a fan of the feature, it strikes me as a way of blaming the 3rd party plugin for a crash, plugins crash in every session and they don't crash nicely with the sandbox, Live can stay open for days without crashing the same plugins--- I have literally never had Bitwig restore a plugin without losing the sound and it de-stabilizes the host, Bitwig acts strange and then eventually crashes again, sometimes the plugin crashes in an infinite loop, it's never restored the original sound, I do a lot of sound design, the feature is useless to me if it loses all my settings...I haven't been able to use a single VST3 without crashing... its a creativity killer, I usually move quickly back to Live for work, frankly I would rather it behaved like live, crash the host and then restore my project stable when I reload the host... Live has a similar quirk in that if I haven't done a save, I can lose my preset but if I have saved, my preset it restored on load.

See screen shots for performance graphs.

divabit2.png

divable2.png

Regnas · Post by **Regnas** » Tue Feb 07, 2017 6:24 pm

@ humanbeingbeing

Your tests are great but you should look also in the individual processes, not just on the overall load of the cpu, because something else could be eating the cpu as well..

Cheers..

EDIT: Btw, that Live skin/theme sucks big time...

Regnas · Post by **Regnas** » Tue Feb 07, 2017 6:26 pm

humanbeingbeing · Post by **humanbeingbeing** » Tue Feb 07, 2017 6:33 pm

Regnas wrote:@ humanbeingbeing

Your tests are great but you should look also in the individual processes, not just on the overall load of the cpu, because something else could be eating the cpu as well..

Cheers..

EDIT: Btw, that Live skin/theme sucks big time...

No, the tests are running the same processes and nothing is eating up the CPU, not even sure what you're talking about, the result shows minimal CPU usage and neither is being eaten up.... you're not suggesting something is adding to the Bitwig number?... this isn't complicated for me. and BTW, I do not care at all about the theme... I only care about the sound, and the topic is CPU efficiency.

humanbeingbeing · Post by **humanbeingbeing** » Tue Feb 07, 2017 6:39 pm

Regnas wrote:@ humanbeingbeing

Your tests are great but you should look also in the individual processes, not just on the overall load of the cpu, because something else could be eating the cpu as well..

Cheers..

EDIT: Btw, that Live skin/theme sucks big time...

It would be helpful if I could see something from your setup, I recall you saying Live doesn't like you, maybe we could figure it out if you share your results.

Bitwig

KVR Audio

Finding Bitwig Studio 1.3.15 more CPU efficient than Live! :-)