Export does not use all the available PC resources

[DocTesla]

Hi.

First of all I must mention that this tool is very nice if you like to do some live visuals for a.e. a DJ set in a live stream. It's pretty intuitve and the fact that it runs on Mac and Windows is exactly what i need. Even as a beginner I was able to create some pretty nice scenes within a few hours of learning by doing.

My issue is atm that I've tried to export some visuals together with my existing audio-podcast-files to publish these videos on social media plattforms. For this I've played around a lot with the different codecs and (HW-)rendering options that are available. Finally I figured out exactly how to use the export feature to get the ideal mix between video-quality and compression. To get all the filters and effects clearly displayed similar to the live-window it's necessary to render the videos with at least 200 fps. 60 or even 120 is not enough. (havent found out why yet, but ok). So far so good.

Now I realized that the export of these videos takes like 4-5 times the time that the duration of the final video has. A.e. i have a 1 hour audio track to visualize and the export takes 4-5 hours. The point is that the overall CPU (8-core AMD) load in my computer is in average under 15% and for the GPU (RX 7800 XT) under 20% during the rendering process. My 32Gb of RAM as well is just loaded around 40% the whole time.
I assume that if i could use 100% of my avalaible ressources the export would run much faster.

Can I somehow tweak the SW to use the full potential of my system? If not, what is the reason for this behavior?

ATM it makes no senso to use the export feature for my purpose because it's much faster (and easier) to record a video in a.e. OBS with the live visuals from Magic with Spout2. 60 Fps recording looks exactly the same as the live-window.

Am I doing something wrong? What settings have I missed in Magic?

[Sadler]

15% sounds about right. Magic isn't multi-threaded on the CPU so the export can only use one core. Even if you've got 32GB, nvme and great GPU the single thread will be the limiting factor.

[DocTesla]

I see. Thank you.

[JBlongz]

Thanks for the question and the clarification. I was wondering the same. It seems there may be some benefit in modestly overlocking CPU for render tasks. Would running multiple instances (rendering two or more projects on separate cores) be a viable feature?