Closed WoofinaS closed 1 year ago
Woofina goated
Might be a good use for FFMPEG, directing the frames like Waifu2x-ext-gui does, making chunks by time (60 seconds or user choice) to see the progress.
Also, something from the waifu gui, they have the option of a RAM DIsk. Spinning rust seems to be the most likely second bottleneck, and using an SSD for the frames will thrash the poor thing.
I haven't looked over the repo just yet to see what is being used, so my apologies on plugging things like that.
Might be a good use for FFMPEG, directing the frames like Waifu2x-ext-gui does, making chunks by time (60 seconds or user choice) to see the progress.
Also, something from the waifu gui, they have the option of a RAM DIsk. Spinning rust seems to be the most likely second bottleneck, and using an SSD for the frames will thrash the poor thing.
I haven't looked over the repo just yet to see what is being used, so my apologies on plugging things like that.
I'm using ffmpeg already, together with vspipe. Effectively reading the frames directly from the file through a Pipe.
Just came across the ffmpeg sections, javascript is not something I am familiar, much different project layout
But, for the devices, it should be as simple as switching output codec to the hardware expected format.
Maybe a box that lets you select hardware and then grey's out anything they don't work with in the codec section
Device: CPU, QuickSync, AMF, NVENC
Also, you made a very beautiful interface for this stuff.
Just came across the ffmpeg sections, javascript is not something I am familiar, much different project layout
But, for the devices, it should be as simple as switching output codec to the hardware expected format.
- h265_amf
- h264_nvenc
- h264_qsv
- hevc_amf
- hevc_nvenc
- hevc_qsv
Maybe a box that lets you select hardware and then grey's out anything they don't work with in the codec section
Device: CPU, QuickSync, AMF, NVENC
I will keep that in mind for a future update. Thanks for your input. Definitely seems viable.
Also, you made a very beautiful interface for this stuff.
Thanks!
catiosnik
@78Alpha @WoofinaS Implemented in #e28cd4c.
Due to the massive speedups from Tensor RT it is possible for both CAIN and RIFE to both be bottle necked by transcoding speed instead of the interpolation itself. To Avoid this there is already support for NVENC. However enhancr also supports AMD gpus and soon apple silicon which both don't have NVENC. It would be nice for people on AMD and apple silicon to have hardware encode support to avoid transcoding and cpu bottlenecks.
Currently enhancr only supports HEVC with NVENC as well. It would be nice to support more formats like the older h.264 for compatibility reasons as well as newer formats like AV1 who are file size concerned.