Hello Community,
are there any ideas about serial vs parallel filters?
Applying several filters at the same time will take longer to process, but makes it also difficult to determine the effect of a single filter.
Using only one filter at a time, it will need several runs and processes an already processed file again (not necessesarily negative).
Is there any recommendation?
Maybe it will solve itself over time when I gain more experience…
Thanks for your input 
I don’t know about recommendation, but enabling multiple filters in one process, the filters will run in a certain fixed order.
there used to be a post document in the past that showed the order that filters would be applied, but it seems the post is gone now.
this was the link.
https://support.topazlabs.com/article/129-which-filter-is-being-applied-first-in-topaz-video-ai
Most of the models work better if you can clean the noise out some from the video first.
Parallel filters doesn’t always put the best noise cleaning models first. The second enhancement option is better for that. I’ve been getting some great results by running Nyx1 manual for the first pass, then Proteus for the second pass.
Interpolation models have a trade off. If you want better motion with less artifacts, you need to run the interpolation after upscaling. It takes much more time, but might be worth it. Again, running in parallel will try to do the interpolation models first if it can.
Damn, it is gone indeed.
Alright, so I better try to optimize quality and keep the resolution and then reload the processed video and upscale it separately?
