Does Variaudio use alot of CPU Power ?

Yes I am being lazy here because I could just do my own test , but I thought I would just throw this question out there wondering if someone else has already done the dirty work of testing this.

I have noticed sometimes my songs seem like they are struggling thru and I don’t know why becaus the inset plugins arn’t that heavy to me ( sometimes during the chorus ) and I just saw a post here regarding flattening Variaudio . Then it made me wonder if Variaudio is the culprit ?

In my mind I have always just kinda assumed that pitch correction in Variaudio gets done and then it’s done and the main CPU hit was just being used at the time of the editing but not while the song is being played back (as a non flattened audio track)

I have never used Variaudio and then flattened that track afterwards . Maybe I am naive , maybe not . Please share any testing that someone has taken the time to do . Thanks

If you haven’t “flattened” the variaudio info in some way then it’s always calculated in realtime and will take more processing.

There are some very weird issues in Cubase with resources when using variaudio.

For example, I have a 3 min song with say 30 tracks and I want to write in some tempo changes. So I enable variaudio on all parts and play with the tempo - no problems.

However as your session length expands time wise, Cubase really struggles to handle variaudio. If you have a 2 hour mix (dj session, film score etc) with a few variaudio tracks and a few tempo changes in it Cubase starts to crawl. Copying audio or altering a tempo change node can take over 5 seconds to work.

So when you say “enable variaudio” are you meaning when you double click the event and have it analyzed, or is there some other way to have Variaudio enabled that I don’t know of ?

So what are all the ways to flatten audio ?
I know of the flatten button withing the audio editor
and soloing and exporting the track
and selecting events and then select “bounce audio”

are there any other ways to flatten ?

I’m sure some of you are thinging just read the manual rookie . I have, and I have benn recording using Cubase vst v1 believe it or not , but I admit It’s easy for me to miss the obvious sometimes

Thx for any help here

Hi. Yes, VA does use a lot of power and the data it generates also bloats the project file, taking up memory. I have found it preferable to be very selective about what you want to treat and I proceed as follows:

  • select the range to process and bounce out a new (interim) file, taking the replace events option
  • VA this interim file
  • bounce the VA’d interim file to a new file, again replacing events
  • crossfade if necessary

This leaves you with having to store VA data for only the interim file, which you can easily revert to and tweak if you need. (If the project still gets bloated you can delete the interim file from the Pool but leave it on disk.)

Hope that helps,
C

I would echo Crothety’s points too, exactly what I do. My projects are fairly complex and I’ve noted instability and lag in the final stages if I leave everything to real-time processing. So I bounce or flatten a lot of things. But I duplicate tracks beforehand, disable them and keep them hidden in a closed folder (or old project versions so I can retrieve them). I also bounce down tracks with a lot of edits (e.g. comping or time correcting) because these take resources too, like graphics resources and crossfade CPU. It all helps Cubase to run smoothly and that isn’t to be sniffed at when you’re in a hurry with things.

Mike.