I’m using SimpleScreenRecorder utility to capture my desktop and voice. The resulting file is around 66 Mb and 22 minutes long. It is mp4 file, resolution is 1280x720, whereas video is encoded with H.264 (constant rate factor of 23 and preset is “superfast”), while audio is Vorbis, 128 kpbs, 48 kHz.
I then edit this file in Shotcut, mostly doing some cuts, but also applying volume gain filter.
The resulting file after encoding (now around 18 minutes instead of 22) is around 220 Mb when using H.264 Baseline Profile (other H.264 profiles give similar or worse results).
So, my original file has trippled in size!
I played around with various export settings, and settled down to Youtube profile, where i additionaly reduced audio bitrate. The resulting file is now around 180 Mb.
Playing around with video bitrate doesn’t help. Too low of bitrate leads to significantly worse picture quality. I could lower the video bitrate so that the file size matches original file size, but with such bitrate the quality is not acceptable. How is it even possible?
Are there some hints to further reduce the file size? Ideally, it should be less than original screen+voice capture, because what i actually did is i cut out parts of the video.
Please let me know if any additional information from my side can help.
i have the same problem. Sometimes i can export a 10 min HD video no problems. Ohter time it will only export ilike 33% and say out of memory on a 5 min Hd video. Does including text or audio take up that much more space?
The problem is that the percentage is tricking you since the standard setting is decent quality and ~70 % is high quality; everything above is not really useful. In the FFmpeg documentation the sane CRF values are between 18 and 28, 18 corresponds to 64 % in Shotcut’s UI and is stated as “crf” next to the percentage (https://trac.ffmpeg.org/wiki/Encode/H.264 ).
So your 100 % is bound to reach lunatic values and was never intended for actual use.
It is impossible because your source is not 100%. 100% is basically uncompressed or lossless compressed. Your request demonstrates a basic misunderstanding of media compression because typically compression is < 100%, and also most video sources (camera, game capture, files, downloads, etc.) are compressed to < 100%.
It is not exactly lunatic; it is lossless compressed, some people specifically want or need it, and it is usable even if some media players are incompatible with it (see complaints about black video after export here).
True, I exaggerated too much. I apologise for this inaccurate statement. I should have clarified for this case in which lossless compression is not reasonable since YouTube transcodes uploads anyway.
maybe, after an update, the videos after conversion were impossible to visualize (black video).
the only solution that worked and allowed me to upload videos to yuotube that were being viewed was: .wmv
About 100%, 100% quality means 100% quality.
maybe, if there was a scale next to it that specified that to have “the same” recording quality you had to have 70 as a value, it was certainly more understandable. or remove% or put an information note in the box, but a visiile conversion scale I think is the best solution.
However I have had these problems for about a year and I was forced to use the .wmv (custom) extension as the only solution that worked on all codecs.
Today I’m testing h264 and youtube setting with 60-70 quality and it works.
videos in .mp4 format are seen again (maybe there was an update)
The size of the edited video is about 30-40% larger than the original but is acceptable (not 20X).
In one year you were the only ones who were able to give me answers and solved some nice problems. thank you