Login

CPU usage depending on the host?

Tips & tricks, working results, technical support

CPU usage depending on the host?

Postby manuel.soruco » Fri Dec 11, 2015 6:11 pm

So I have been using Nebula for about 2 months now, mostly on Cubase.
However, since I have a secondary PC in the studio for Kontakt samples when I´m writing to picture, I thought I might split my processing power for my mixing sessions as well, since Nebula is making my main PC squeal.
So I started offloading instances to VE Pro I started noticing some weird things regarding CPU usage. The programs I was offloading from my I7 Quad core (main PC) were liberating considerably more CPU than what it was taking my i5 Quad (Slave PC) load those same instances.
So for example, 8 stereo compressors and 8 Reverbs were taking about 60% on Cubase on my i7. I loaded the same instances in VE Pro on a i5 and the needle barely hit 25%...

I had read that VE Pro is build to be more resource efficient than most DAWs, so it might make sense.

However, I noted something simmilar with Nebula Man.
I´m working on an album for a band that has 6 songs, and for each song we recored 3 or 4 takes. So that is a little over 1 hour of recorded music.
I started inserting a 1073 preamp simulator and the 3 tape stages from R2R on every channel and then I was going to freeze the tracks. However it was too slow of a process since I have to do it track by track, and I couldn´t even insert all the instances on the channels before my CPU started overloading, so that´s when I decided to buy nebulaman to make my life easier, and indeed it did!

So according to my calculations, for 1 hour of material that the CPU can´t fully process in real time, logic would dictate that the batch processing would take more than 1 hour.
So I start the batch conversion, start shutting down speakers, lights, cleaning a bit the studio and just about to go to bed to leave it converting overnight, when I get a message that it´s finished... o.O
It took less than 10 minutes...

So my question is... What the hell is going on?!?! :shock: :lol:

Are both NebulaMan and VE Pro coded to be way more efficient than Cubase?
Is Nebula coded in a way that is more efficient when it´s working in non-realtime?
Do I have some configurations on Cubase that might be making it work like crap? (1024 ASIO buffer, Activate Multi Processing, Asio Guard: High, Audio Priority: Boost, Steinberg Power Sceheme: Off)

Btw. I´m using the default Nebula Reverb preset for everything.
manuel.soruco
User Level II
User Level II
 
Posts: 23
Joined: Fri Oct 30, 2015 2:56 pm

Re: CPU usage depending on the host?

Postby namooz » Sun Dec 27, 2015 11:05 pm

I don't know if this will do you much good, But let me explain. You made yourself very clear, well written. When I moved from Pro Tools 9 (32 bit) to Reaper (32/64-using the latter), I doubled my Nebula count. The Acqua plugins draw a little more. I was getting glitches while adjusting a plugin while playing, too, in PT. I know that 11 and 12 are more efficient, but why use it if Reaper can do that for a fraction of the cost/daw? I am not associated with the company in any way, but took the time to get a handle on it and now I'm very glad I took the little plunge-around $60 unless you make over $15k/year using it, something like that. If you're not aware, you can change Reapers skins, colors etc. Several Nebula forum users and developers use it. I think it's a great daw with a great interface, even though I am still quite new to it.

I want to add, and excuse me if you're already aware, but you'll need around 6 gig ram and if you can, move your drives from external to internal. Good luck with your decision.

I'm using a 2008 Mac Pro and get around 20-24 instances which works well with all my outboard exchanging plugin/outboard or in many cases, both. Good luck with your decision.
namooz
User Level VII
User Level VII
 
Posts: 71
Joined: Mon Dec 30, 2013 3:07 am

Re: CPU usage depending on the host?

Postby Support » Mon Dec 28, 2015 10:50 am

manuel.soruco wrote:So I have been using Nebula for about 2 months now, mostly on Cubase.
However, since I have a secondary PC in the studio for Kontakt samples when I´m writing to picture, I thought I might split my processing power for my mixing sessions as well, since Nebula is making my main PC squeal.
So I started offloading instances to VE Pro I started noticing some weird things regarding CPU usage. The programs I was offloading from my I7 Quad core (main PC) were liberating considerably more CPU than what it was taking my i5 Quad (Slave PC) load those same instances.
So for example, 8 stereo compressors and 8 Reverbs were taking about 60% on Cubase on my i7. I loaded the same instances in VE Pro on a i5 and the needle barely hit 25%...

I had read that VE Pro is build to be more resource efficient than most DAWs, so it might make sense.

However, I noted something simmilar with Nebula Man.
I´m working on an album for a band that has 6 songs, and for each song we recored 3 or 4 takes. So that is a little over 1 hour of recorded music.
I started inserting a 1073 preamp simulator and the 3 tape stages from R2R on every channel and then I was going to freeze the tracks. However it was too slow of a process since I have to do it track by track, and I couldn´t even insert all the instances on the channels before my CPU started overloading, so that´s when I decided to buy nebulaman to make my life easier, and indeed it did!

So according to my calculations, for 1 hour of material that the CPU can´t fully process in real time, logic would dictate that the batch processing would take more than 1 hour.
So I start the batch conversion, start shutting down speakers, lights, cleaning a bit the studio and just about to go to bed to leave it converting overnight, when I get a message that it´s finished... o.O
It took less than 10 minutes...

So my question is... What the hell is going on?!?! :shock: :lol:

Are both NebulaMan and VE Pro coded to be way more efficient than Cubase?
Is Nebula coded in a way that is more efficient when it´s working in non-realtime?
Do I have some configurations on Cubase that might be making it work like crap? (1024 ASIO buffer, Activate Multi Processing, Asio Guard: High, Audio Priority: Boost, Steinberg Power Sceheme: Off)

Btw. I´m using the default Nebula Reverb preset for everything.


In order to get an accurate response, we need to know how you measure you computers resouces and CPU load and the complete version of each host tested.
Best regards,
Enrique Silveti.
Acustica Audio customer and technical support
http://www.acustica-audio.com
User avatar
Support
Expert
Expert
 
Posts: 2885
Joined: Mon Feb 24, 2014 11:33 am
Location: Lodi, Lombardia, Italy.


Return to Working with Nebula

Who is online

Users browsing this forum: No registered users and 2 guests