Reducing CPU load by lowering audio quality

Cubase
1

Hi all,
I’ve been searching the Forums but haven’t found what I was looking for (If I searched not well enough and if that topic has been discussed before, then a thousand times sorry about that, I would be pleased if you linked me to the topic so I can read the replies there instead, thank you :slight_smile: ).

Well, I’m a Cubase 6.5 user and thinking to upgrade to C9. My system is years old (Intel Core I7 first generation, 16GB Ram, Win7 64bit, no SSDs). Good enough for most plugs for me. :slight_smile: But not for all I guess (EWQL Hollywood or u-he Diva is out of the question, I guess?)

So my question is: Is there a possibility to just reduce the sample rate, bit depth and set stereo to mono, to free up lots of system/CPU/RAM resources for Cubase, without having to upgrade the complete system just for CPU-RAM-hungry plugins? That makes sense to me, but I am not sure if any Cubase version after 6.5 has such a feature implemented already?

For example, I’ve been using Fasttracker 2 (a music tracker software for DOS) back in the days with a 486 computer. There you could just reduce your Soundblaster settings to, say, 22khz, 8bit, Mono, to free up a LOT of resources to mix 32 (max) software channels in realtime, without buying a new Pentium system or whatever. That was absolutely fantastic back then! Because you could playback and edit your Module in a lower audio quality, and then render/outbounce the finished track in 44khz,16bit, Stereo - so yes it worked just perfectly even without a Pentium computer.

Nowadays, soundcards or audio interfaces do not even offer anything lower than 44/48khz, 16bit, Stereo, as this is of course just nowadays audio standard. It’s like a verbal middlefinger like: “Oh, you got a slow system there?! Ah-boo-hoohoo. If you want more resources, go buy a completely new computer, sucker”! :stuck_out_tongue:

In older EWQL Play versions, you could indeed set the audio quality to I think low, medium, high, to free up the CPU load. Worked great! In newer Play versions, no such possibilty anymore. :confused: Why don’t Plugin vendors at least release a mono-version of their plugins or something?

I’m no programmer, but I’m trying to answer myself: I don’t know if reducing the audio quality of your soundcard makes any sense on modern systems. I mean, it could be the case that the CPU has to mix all midi, audio and instrument channels anyway in Cubase beforehand, before giving the signal to the soundcard, so setting your modern soundcard to a lower quality, would NOT free up any CPU resources whatsoever? Could that be? But why did it work in DOS times with Soundblaster and trackersoftware then? That old EWQL-Play-Version proofed that it works - that is because it was done before the channel-mixing. So it would have to be an internal plugin or even Cubase-setting, and has nothing to do with the soundcard unlike in DOS-times?

It would be fantastic if there’s a way to tell Cubase that it just has to mix in mono, or 22khz, or 8bit, so you can use your extremely CPU-hungry plugins in realtime, and then outbounce your tune in high quality without the need to freeze tracks, or to buy a new computer (Yeah, I’m repeating myself, sorry. :wink: )

Thanks for your replies and feedback!

Cheers,
Rapture

2

In short: Do newer Cubase versions (after 6.5) have the option(s) to lower the sample rate (lower than 44khz, say, 22khz), set Stereo to Mono, set bit depth from 16bit to 8bit in your Cubase project (like this was possible in older EWQL Play versions, and also in Tracker software like Fasttracker 2 in DOS times) , before the CPU has to mix all Midi- Instrument- and Audio-Channels etc, thus reducing CPU and RAM load for older computer systems, without the need to freeze tracks, or to buy a competely new PC just for massively CPU- and RAM-hungry plugins? (you can render/outbounce the finished track in high quality then, but for working on the song, reducing the quality would massively free up system resources. Not a new idea, but I want for this in Cubase).

Anyone?

Thanks

3

I don’t know, quality is what makes me want to play.

If you’re having problems running cubase (bloatware) you can do different versions easy enough.

Take what you’re working on and export a mixdown then open a new project and put that into it and just have one track with plugs on it playing along with 1 stereo file. Then repeat.

That’s what I do if my project gets too big to track with low latency. Resource management.

4

Tacman7: Yes, that’s one way to do it of course - but not very flexible, because you can’t change the mixdown (audiotracks) then anymore. Quality is still maintained when you could at least switch your project from stereo to mono (50% CPU resources saved?) Is this possible? I could try this - if it’s also possible to mixdown your finished track in stereo again then. :wink: Not the best solution though as it is not possible to mix your track in stereo, yeah. Sample rate reduction would be better, 32khz if 22khz is too bad to play/listen…

Btw, your signature states you have 8mb ram - that’s not very much haha ;D;D;D ^^

5

Definitely not.
And apart from you, there is probably no one who can judge a stereo mix played in mono, even more when played in 22 kHz samplerate…

6

Sven: Maybe not 50% … but with a combination of Say, 22khz/32khz, Stereo and 8/16bit you could mix in stereo in with an ok quality and save up lots of resources. It’s a compromise, but makes much more sense than to force users to upgrade their hardware/computer.

Also it’s better than freezing tracks or bouncing everything into an unflexible mixdown and continue working from there. And I’ve added about Mono: “Not the best solution though as it is not possible to mix your track” so no need to become sarcastic about that. :stuck_out_tongue:

As said, I worked in trackers and reducing the overall audio quality helped massively to mix 32 channels in real time on 486 computer. Of course you can mix in 22khz/32khz, but I see you’re not even interested to include something useful to this thread.

Maybe other DAWs more advanced than Cubase in this point. :stuck_out_tongue: