do you want to disagree with what i said or just shrug? i can shrug too.

So on one hand you think giving your opinion based on experience about which host you feel has been most efficient would be a generalization and dishonest, but you conclude that Reaper users are the least educated and most arrogant? That itself is dishonest and arrogant.

you didn't read my post yet. do that.

DELETED

Yeah, I'm done with this, all I have left is a

well the only three options you have are:

- they're not being dishonest because the generalization is valid
- the statement wouldn't be a generalization
- give up

regardless, Logic PC beat the crap out of Cubase PC in terms of track count as well.

The question then becomes whether Reaper Mac is less efficient than Reaper PC. From what I know though, Cubase PC uses it's own AISO driver to access Core Audio, a layer of abstraction that Reaper Mac wouldn't have.

For what it's worth I'm not a Logic man anymore, hence the v8 state of Logic on my machine, but I am curious, so I'll run this in the next couple days. I know Logic 5-6-7 were all a little more CPU saving than 8.

@hibidy yeah that could be a drag, but at least for me in 7.2.3 you could stop and start the transport and it would even out the CPUs.

Before I bought Reaper years ago, I made an efficiency comparison between Reaper, Cubase [SL], EnergyXT 1.41 and Podium, as I was deciding which way to go after Cubase. I decided to ditch Cubase in favour of Reaper. it turned out to be the most efficient, and I didn't have a particularly powerful computer at the time, so I needed a efficient VST host. Second came EnergyXT 1.41 that I still use, too. Excellent VST host, and it does work with multi-core CPU in a bit peculiar way... I don't think EnergyXT 2.6 can do that? I also tried v2.6 and I found it unstable, so I stopped thinking about it and I didn't like the way the development of EnergyXT went after 1.41. Podium was close by, I love that program, too. Cubase was the least efficient of the bunch. The test just showed me how software made by big companies is just pure cra* made for fanboys. I wish I could test Live and Studio One at the time, but Live has a weird workflow for me, and Studio One didn't exist at the time, but I now know it's not as efficient as others.

Now that I have a monster machine, I don't care so much about the efficiency, but workflow. However, I stayed with Reaper and EnergyXT 1.41, as I don't find changing VST hosts particularly fun.

Cheers!

haha, I'm going to keep with the thread but stop paying the troll toll with a certain poster. Personal insults coupled with arrogant snottiness is the end of that.

Unfortunately, I don't have a way to test this. But I agree that different hosts are different on different platforms. (I still had my mac when I was using cubase and cubase ran better on my allegedly antiquated PC. Not by a ton, but better)

The one thing to consider also is the ASIO driver you use. There can be quite a difference in CPU usage. I used more than a few of those throughout the years and RME drivers turned out to be the most effiecient in my experience, and I like checking things out when I buy somethign new. I love RME drivers, but I kinda hate their unjustifiably overpriced hardware. So I buy RME only second hand... currently I have HDSP 9652 in my main machine... I gave 300 euro for it. M-audio Audiophile is collecting dust now... I never liked M-audio drivers even though I must admit they are stable in my experience, and rather efficient. The big problem for me is that M-audio ASIO is NOT multiclient. I hated that. Even the Creative Soundblaster ASIO driver is multiclient! But the driver itself is the least efficient of the three. My XtremeMusic card is also collecting dust...

"SoundPlaster"

Cheers!
Last edited by DuX on Tue Apr 10, 2012 10:37 pm; edited 2 times in total

Well, my experience with live back then was that it crapped out once the cpu hit a certain threshold (not a problem below that threshold) studio one has always seemed a bit intensive but I own version 2 and I can't use it. It's not buggy at all, it's just not able to run certain plug ins. I mean cripes, it IDOLS at 5+% with nothing loaded (and I have an i7)

And since you can't sell sonar, I still have that. It's one of the weirdest. It has a tendency to favor the the first core, but it never really reaches that and causes trouble. Once that core is being pushed, it starts spreading. This has been explained to me a number of times but I don't understand it. The reap is the only host that is more or less inline with the task manager. What it reports ram/cpu is almost exactly what reaper reports. I take that as a good thing. Ymmv.

could just mean that reaper uses the functions provided by windows for that. they all have their own set of advantages/disadvantages. see any topic about timing on windows.

asio drivers - that would probably be due to i'm guessing spending a lot of time in the driver. if you write your own code that's easy to test. there are a lot of different sync and buffer handling methods that can be used by drivers. many drivers even have options. some modes will give you lower latency, other modes will spend less over-all cpu time allowing you to get that extra 5% at a higher buffer length setting. not all methods will be supported on all platforms and operating systems so it starts to involve your main-board, cpu, bus chip-set and others.

a large amount of time spent while idling could be a lot of things. updating vu meters takes quite a bit of time. who knows what insane kind of crap some hosts do.

think about this though: if you wanted the lowest possible latency you'd want to generate the buffer as quickly as possible, then sit there asking the card "ready for this buffer yet?" over and over. that would mean idle would require 100% cpu. you'd also be able to use up very close to 100% of your cpu time with plugins before any buffer-overrun happens.

JMFC !

Why is it so hard to understand what is really important for a normal DAW user (musican and/or engineer) when we are talking about CPU efficiency ?

It is not how it was coded, it is not how it works inside.

It is simply the number of tracks + signal processing plugins it can run without forcing its user to use the freeze feature to avoid crackling at low (comfortable) enough buffer/latency settings.

= the end result.

And it is extremely easy to check which DAW can do it better:

track1: midi sequence, synth, eq, compressor
track2: midi sequence, synth, eq, compressor
track3: midi sequence, synth, eq, compressor
track4: midi sequence, synth, eq, compressor
track5: midi seque... etc etc...
or
track1: audio file, eq, compressor
track2: audio file, eq, compressor
track3: audio file, eq, compressor
track4: audio file, eq, compressor
track5: audio fi... etc etc...

= a test that is fairly close to the normal real world project.

So why don't you all stop arguing about stuff that doesn't matter at all and do the test already ?

BFP !
.