I mentioned in the previous post that I have been experimenting with the Lossless option for NVENC, which should give me a higher quality original version of the recordings to work from. After that I re-encode it for the archive, simply because I do not need lossless quality for the long term and it is a LOT of data (try 584 GB of these lossless archives for 12 videos, while re-encoded archives, the upload video, and a separate copy of the audio mix which is included in the re-encoded archive is just 69.7 GB).
Well, I couple weeks ago I spent the time putting together a file to use the blend filter in FFmpeg to show the differences between two videos. Decided to run that on the lossless version of one of the game videos and the re-encoded version, to see how different they might be. I can see some difference on my laptop display, but there is no way I would notice it outside of this difference video. You can look for yourself, but remember YouTube re-encodes everything for itself, so some detail might be lost for you:
The videos I've done this lossless testing with are of Killing Floor 2, which I used because recordings in it before seemed to demand a higher bitrate than I had chosen, so Lossless seemed a good fit. Just ran through every official map for a separate video and I intend on using these as on-offs to fill out weeks, so it may be awhile before they all come out. (There is no campaign component to the game, so this seems appropriate.)
No comments:
Post a Comment