Hello,

I would like to be able to decode videos using my gpu rather than using my cpu for it. As far as I can tell, there isn't any way to decode a video using gpu. If the concern is that users might misuse the feature and put a bunch of videos on hardware decoding, maybe it could be included as an "advanced" setting or something.

Currently, running a broadcast with multiple videos (3 or 4 .mp4s that are mostly under 100mb in size) pins my R9 5950x at 95% - 100% usage and makes broadcasting (even with hardware encoding) pretty much impossible due to the lag.

^ I had about 85 inputs, with a few virtual inputs based on those videos

My solution was to use my laptop to run multiple SRT feeds containing the videos I'd like to use and then to use hardware decoding to decode the SRT feeds and it worked flawlessely and drastically reduced the load on my cpu, but of course I would not like to run an SRT feed every time I want to use more than a couple videos in my project.

Thanks

Off-topic, but you wouldn't happen to be the Hazza from Rocket League, would you?

+1
I would like to be able to have a tick box like OBS to enable/disable the hw decode of any media source