Mulab 8 has allergy to VPS Avenger

[mgiambro]

Why does Reaper handle demanding plugins so much better?

Reaper has some anticipative fx processing magic going on, which I believe has it's downsides - latency etc.. When it comes to CPU usage, I think it's widely known that you can throw more CPU hogging VSTs at Reaper than any other DAW.

I recently did a test with my most CPU-heavy vst, Poly-Ana. I compared MuLab, FL Studio and Zynewave Podium. From memory, playing a 3-note sustained chord, FL Studio crackled and choked at 11 instances with 68% cpu usage. Podium fared better. MuLab better still. I was getting underruns at 13 instances - 52% cpu, but only heard crackles at 14 instances.

I get that there are a lot of variables at play e.g. processor type, audio interface, driver etc.. For me MuLab performs well compared to other daws. I'd be interested to hear of peoples experience in comparing MuLab's cpu performance to DAWs other than Reaper..

I can't say anything about Poly-Ana, Podium and FL Studio, I only know those by name.
So far I haven't noticed latency issues in Reaper.
I don't even use any demanding plugins, yet, but I might like to.

Fair enough, but if you're benchmarking against Reaper in this regard, you'll be wasting your time. Reaper is (as far as I know) is the most CPU efficient DAW because it works differently to most. If I were recording bands and audio, Reaper would be high on my list. MuLab (IMHO) is a lot more fun and far better as a creative tool, so I'm prepared to take the additional cpu hit(which isn't significant in my case). Try comparing MuLab with other electronica-focused DAWs - FL Studio, Bitwig, Ableton etc or all-rounders like Cubase, Studio One .. I'd be interested to know how you get on.

[mgiambro]

An alternate workflow to play that triad with 13-voice unison is to sample Poly-Ana into the two MuSynth oscillators where each osc can have up to 20 layers with pitch, drift, phase, volume and pan, with next to no CPU. So if you can't run a high-CPU synth, sample the sounds you like into MuSynth. Just an example....

Thanks for the tip

[heks]

Not a very scientific test, but quick. Instrument used Lush 101 (it just kills hosts) preset was "Beautiful Pad", a triad play'd over four bars looped.
Mixcraft 8, five instances before sound breaks up.
MuLab 8 , three instances before sound breaks up
Mixcraft ( at this time) seems to be at a similar price to MuLab so more relevant competition than say the big names.

[mutools]

En that's surely with exactly the same audio setup (samplerate, blocksize, multicores) and without Mixcraft doing some automated pre-rendering like Reaper? There must be some reason that explains the difference because MuLab does nothing special with VSTs that could explain such difference.

[mutools]

I see that Lush has multi-core support. Did you disable Lush's multi-core support in both MuLab and Mixcraft before doing the test?

[e-crooner]

An alternate workflow to play that triad with 13-voice unison is to sample Poly-Ana into the two MuSynth oscillators where each osc can have up to 20 layers with pitch, drift, phase, volume and pan, with next to no CPU. So if you can't run a high-CPU synth, sample the sounds you like into MuSynth. Just an example....

Thanks for the tip

If you are bored, it would also be interesting to see the number of instances for Reaper

[mutools]

e-crooner as has been said multiple times already reaper uses pre-rendering. You cannot compare test results of pre-rendered processing with pure realtime processing.

[heks]

I see that Lush has multi-core support. Did you disable Lush's multi-core support in both MuLab and Mixcraft before doing the test?

I did nothing but load Lush straight into both hosts. Don't know if Mixcraft did anything automatically? I'm not that Knowledgeable about Mixcraft. While the wife's out I will load up Studio one and see how many that can take.

[mutools]

It's important to do the test with multi-core disabled in Lush!
Though i don't know if Lush has a user setting that can disable its multi-core processing.
If it would not have such setting then Lush is not suitable as a VST to properly compare cpu usage.

[heks]

I don't know about the multi-core stuff, but Lush crippled Studio one 4 with two instances. In defence of studio one I haven't loaded it up for ages and as it updated on loading it will be in default state. As you can imagine, I use lush only occasionally, but it is so "lush" sounding. I'll have a look at Lush's settings.

[e-crooner]

I think many plugins have MC support enabled these days, without mentioning it or allowing the user to disable it.

Anyway, I don't understand why you think it must be turned off in comparison tests. Either you turn it off on all DAW's, or you turn it on on all DAW's. It's fair either way.

Anyway, here is the Repro-5 load per core/thread with MC support off vs on:
https://app.box.com/s/ofplvbc0rfr31i2j6d83yllkiq1ov42y

(A single instance with a pad chord in HQ mode. The lower 8 boxes are the cores/threads.)

[heks]

Now that was weird, I turned multi core off in lush and Mixcraft could only handle two instances. I then did the same in Mulab, but just after it started in the first bar the cpu meter at the top went red and turned off the sound. I was expecting Mulab to do three instances again. Maybe Lush is not the best instrument to test DAW's with ( I chose it because it such a cpu taxing plugin). Is there a recognised standard "test" to see how hosts manage plugins?

[skarabee]

Why should DAWS have to be incriminated when plug-ins eat all the power?
When I encounter problem with a plug-in, the first thing I do is asking support from the plug-in manufacturer. And if, like I have read here and in other forum, big Compagny DAWS can hardly support 2 or 3 instances of such plug-in, I have a radical solution: >desinstall and forget
A good plug-in should not suck all the power for itself, be it in Mulab or other DAW.
my 2 cents.

[e-crooner]

Why contact the plugin maker?! If it works flawlessly, what are they supposed to do? Lush, Repro etc. are optimized for sound quality. There is nothing wrong with it and that is not the issue. The issue is that Mulab seems to be less capable than other DAW's of handling such plugins.

[e-crooner]

Now that was weird, I turned multi core off in lush and Mixcraft could only handle two instances. I then did the same in Mulab, but just after it started in the first bar the cpu meter at the top went red and turned off the sound. I was expecting Mulab to do three instances again. Maybe Lush is not the best instrument to test DAW's with ( I chose it because it such a cpu taxing plugin). Is there a recognised standard "test" to see how hosts manage plugins?

Obviously Lush is an excellent choice for testing CPU efficiency.