I use Photo AI for improving some 4k videos that the video house should be ashamed of. Basically for sharpening and face detection. That said, I don’t use Video AI as it seems to create an inferior output for what I like anyway. So I go thru the pain of converting let’s say 154k of frames with Photo AI. Since Photo AI doesn’t really use much of my system resources, it’s slow going. That said, with Topaz 2.0.0 out now, and half way thru converting the frames, any thoughts if I should stick with 1.5.x that I’m on or switch to 2.0.0 half way thru the process? Wasn’t sure if the change in version would be noticeable.
Before we even had the Video a. i. I made an experiment to upscale a 15 minute long video recorded in full HD to 4 k. It took 15 days to complete and my GPU died shortly after. I would not do that again and use Video AI instead. 
1 Like
For me my GPU and CPU as basically sleeping. Water cooled AMD 7950x and RTX 4090. I’m not upscaling just sharpening and Video AI doesn’t do a good job at that overall.
How about Sharpen AI instead? Photo AI is a bit slow still, that’s true.
1 Like
