Different vst CPU consumption in different DAW ?

[cowby]

Hi,

If this kind of question has been asked.... Sorry about that. I am using Studio one 3 and just wondering if I use the same set of VST in Cubase 9 or bitwig latest, will the CPU usage bit less ?

Cowby

[LawrenceF]

You'd have to try it and see.

Studio One isn't exactly (as reported across the board, people seem to get different results) the most efficient host out there and generally speaking, every workstation is a unique thing in itself architecturally and there's really no standard level of efficiency that every product will always meet as relates to that. If it was that simple every host would be as efficient as Reaper in any and all situations, and that doesn't appear to be the case.

In my testing, nothing is really as efficient as Reaper in so many different situations. Daw X may match it with VI's, DAW Y with audio plugins, but for everything combined, live sound, video, whatever else, it seems to beat everything in that regard when taken in it's totality. I think it's lack of being... shiny?... kinda plays into that. It's more of a bare bones audio daw (architecturally) than some others. There's very little shiny fluff on top of it eating cpu cycles. It's audio engine first, everything else secondary.

It's kinda like with cars. Every group of people designing them have (generally speaking) the same education in engineering, but they don't all get the same gas mileage. The difference with DAW's is trying to agree on what is the level of performance (on the low side) where it becomes no longer workable. Opinions on that vary.

[Autobot]

[...] It's kinda like with cars. Every group of people designing them have (generally speaking) the same education in engineering, but they don't all get the same gas mileage. The difference with DAW's is trying to agree on what is the level of performance (on the low side) where it becomes no longer workable. Opinions on that vary.

Nice comparison. I also think it highly depends on the DAW architecture. Reaper seems actually the most effective followed by Ableton Live then Cubase
. Thats my experience.

[.jon]

My test project maxed out with Reaktors, MMultibandConvolutions and Beatzilles, measurements taken from Windows 10 Task Manager, project running for 4 minutes:

Reaper 5.29 hovers most of the time @ CPU 43-45% (low spike 41.4 high 47.6)
Live 9.7 hovers most of the time @ CPU 41-43% (low spike 39.9 high 45.6)

Now you, dear reader, may suffer a horrible aneurism, sudden prolapsis, brainspasm and ask through tears in helpless confusion, how possibel?! Must lies! Riper sooo efizent, verysmall only 10 mega bits, most best 11!!!??? Reads on internet, riper greatest OF ALL, is made by winamp!!

Well, Reaper's own CPU meter at default settings showed anything from 30 to 50%, most of the time 10% less than actual. How could you have known?

Reaper uses 300Mb less RAM though, so if you have only like 1Gb memory, this might be very important.

And finally, adding one more note of polyphony in either DAW results in crackles, so the small % difference means nothing in practice. They both can run the same project and nothing more. I tested Bitwig as well back when I had the demo installed, same results. It's not the DAW that does the heavy lifting, it's the plugins, so equal performance is not a surprise.

ymmv of course.

[LawrenceF]

That's the thing though. Another person will run that same test on their system and get a different result.

None of them are "lying" about that, it's more that no two computers (especially Windows systems) are identical, especially for the general masses who use them for literally everything, not just music, and have all kinds of different things installed on them. There is certainly a "range" of performance where that comparison is more arguably legitimate, but to expect them all to behave exactly the same all the time is maybe not so realistic.

One guy sees X, another sees Y, with the same software... then they angrily debate the truth of all that for 3 days when the reality is that they're both telling the truth.

But... that's the Internet. "Whatever I see on my system is the only possible truth."

[.jon]

That's the thing though. Another person will run that same test on their system and get a different result.

Citation needed.

Seriously though, I'm not saying there couldn't be different results, but usually "Reaper is highly efficient" claim isn't based on such a test.

[LawrenceF]

Take note of how I phrased that comment.... words are important.

"In my tests" (i.e, which doesn't include anyone else's tests?) Reaper performs better overall in totality (like, measures well doing multiple different things, not necessarily beating everything at everything all the time). "Your testing" (on your system) may end up differently?

Again, the arguments often come from people viewing what they see as being the only possible outcome or truth, and then repeatedly arguing that. I'm not disputing what you see, I'm only telling you what I see.

Reaper being efficient (even if just a bit more in some tests) than some others is no big deal, and kinda directly logically supported by it's design. It's a really "light" (small) application in general. SAW Studio probably beats them all for similar reasons, it's (as a practical matter) a rather tiny application with very few dependencies. Probably smaller than Reaper, the size of it.

[Autobot]

That's the thing though. Another person will run that same test on their system and get a different result.

Citation needed.

Seriously though, I'm not saying there couldn't be different results, but usually "Reaper is highly efficient" claim isn't based on such a test.

Whoa ... why so pissed off by Reaper? At least I have no reason to hype Reaper and on my old system it have has a lower CPU consumption. Do not need to get upset dude ...

[.jon]

This topic discusses CPU performance running the same VSTs in different hosts. "Overall", "multiple different things", "small file size" are all really interesting I guess, but this is a rather simple and measurable thing in the end, I did the test and listed my results.

Maybe you want to provide yours?

[Autobot]

I have better things to do. An occasional post of my experience is enough.

[Zexila]

Well, in my test's, different DAW's had different idle usage of RAM and CPU and some of them acts differently close to overload, if getting 10 tracks more without glitches means nothing to one, than they all are same.

[LawrenceF]

There's always a "best" thing for any given singular literal technical measurement... even if it's a tiny gap. I think if there was a truly scientific test across multiple (like, 300) different systems Reaper would probably come out on top overall, for cpu efficiency, even if by a hair. So what?

Something will be more efficient than most, something else will be more powerful than most, something else will be more flexible, something else better for workflow, etc, etc, etc. Those gaps may be pretty small in many or most cases but they always exist in some way. No daw ever "wins" at everything, but they all win at something.

There's very few things like that which can't be subjectively or disputed by someone though. Very few. I mean, for example, saying Reaper has the best scripting of any DAW is one of those things that defy any argument, it's just (likely) a literal truth, but otherwise, most everything else can be disputed or argued in some way.

[.jon]

I have better things to do. An occasional post of my experience is enough.

I didn't expect you to, that was meant for Lawrence. I'm not "pissed off about Reaper", but this efficiency claim appears to false- at least according to the only actual test results posted so far.

I did my test expecting the often-repeated claim to be true, by the way.

[.jon]

There's always a "best" for any given technical measurement.

Something will be more efficient than most, something else will be more powerful than most, something else will be more flexible, something else better for workflow, etc, etc, etc. Those gaps may be pretty small in many or most cases but they always exist in some way. No daw ever "wins" at everything, but they all win at something.

There's very few things like that which can't be subjectively or disputed by someone though. Very few. I mean, for example, saying Reaper has the best scripting of any DAW is one of those things that defy any argument, it's just (likely) a literal truth, but otherwise, most everything else can be disputed or argued in some way.

Certainly. Reaper does approximately 10000 things Live never will, and is arguably better in many. I switched away from Reaper to Live for purely subjective, non-quantifiable reasons.

However, these aren't in any way relevant to the question presented by OP, or to the claim presented by you and Autobot.

[woodsdenis]

Wouldn't it be great if a developer made a plugin that just used up CPU in a measurable way and nothing else. Made it free, in that way a real measurement of different DAWs and CPU power would be possible.There are so many variables out there.