wassup
Recently I used an easycap (640x480 30fps) to record an old vhs tape in VLC…
As a result, the final video pooped out a fat 70 giggybyte hour and a half MP4 file.
70 GIGS
thats fatter than a blu-ray
Y’all can help me out? I tried youtube only to become overwhelmed
what magic must be done to encode this.
I’ve used handbrake and would rather just use/learn FFMPEG
Can less than 1 gigabyte be acheived on h265 Standard Def without looking like mud sloshing around the screen???
also i got a fat graphics card, will that help?
Not sure if it’s going to be under 1 GB, but the documentation is here:
https://trac.ffmpeg.org/wiki/Encode/H.265
Also 70 GB is fairly normal because you probably recorded uncompressed.
2 Likes
I have been doing this kind of thing a lot the last few years. I however use handbrake with saved custom presets that took me a bit of work to get just right for my use cases.
I would suggest you do something like I do. Create a short clip of your video, about 60 seconds or so, that includes something like a camera pan or people or objects that move across the scene. Then use this clip to experiment with your ffmpeg settings to find what your looking for. This won’t let you know how large the final video will be, but it will let you determine how low you can get the settings without making a mess of the video quality.
1 Like
I will try to read this without falling into a wormhole of impatience.
Seems like Adobe has the better implementation than handbrake IMO
https://trac.ffmpeg.org/wiki/Encode/PremierePro
That requires Premiere Pro though which is not exactly cheap (well, legally anyway). For an alternative you can look into Blender.
edit: also if you read the article, Premiere isn’t actually encoding using ffmpeg. They show ways how you can use ffmpeg after exporting from Premiere.
But, quite honestly, the most flexibility you are going to get with commandline ffmpeg anyway. And for testing you don’t need to encode the entire video, a sample of a couple minutes should be plenty (or various samples depending on the source material). There’s also the -ss
, -to
, and -t
flags you can use to only encode parts of the file from a specified starting point, up until a specified ending point, or for a specified amount of time respectively.
1 Like
thanks for the response
so many options, I’d rather ask what settings would you recommend for 480p lossy compressed 90 mins (no audio) just to get in the ballpark.
It’s VHS tape so it’s already potato quality, just don’t wanna tread into mashed potato territory.
and what is this I read about 1st and second pass?
There’s no universal setting one can recommend, but I think x265 should have presets or quality-options instead of bitrate? I only used x264 before so I’m not entirely sure.
2-pass encoding is a method of getting better compression for the same visual quality (or better visual quality for the same compression, however you see it) by analysing the footage in the first pass, then using that data in the second pass. It’s basically looking at the footage and how it can optimally compress it without actually doing so in the first pass, then using that data in the second pass. It takes a bit longer but it can be worth it depending on the footage and usage of it.
Thanks
I will definitely look into ffmpeg
I think VLC has a convert option if I might remember
Download ffmpeg.
Extract the zip.
Open a system terminal. (cmd)
Change directory to where ffmpeg is extracted to.
Then run (as one line) (change the file paths)
ffmpeg -i “C:\Users\owner\input.mp4” -c:v libx265 -preset medium -crf 18 -map 0:v? -c:a libopus -vbr on -b:a 160k -compression_level 10 -map 0:a? -map_chapters 0 -map_metadata 0 -f matroska “C:\Users\owner\output.mkv”
2 Likes