So I have a bunch of clips from which I want to make a compilation. Some are in .mov format and some in .mp4 format, some are high crisp quality and some are lower quality. I am making tests with 6 clips (i have over 60 clips) to see which settings give me the best quality. By best quality I mean NOT downgrading the high quality clips, and keeping the lower quality clips as they are. My 6 clips make 176mb but when I export, the quality of the HD clips is gone and the file size often goes higher than 176mb which doesn’t make sense.
As export settings I am using the youtube preset (.mp4, h264, aac) and I have tested many different settings from changing quality % and GOP, interpolation and deinterlace etc…
I can’t find the right settings that will keep the quality of the clips as they are while also keeping approximately the same file size.
Bonjour, Voici un raisonnement tout simple. J’ai 2 clips de 30 secondes, un de haute qualité qui fait 100 Mo et un de mauvaise qualité qui fait 20 Mo. Total 120 Mo pour 1 minute de vidéo. Je veux faire une compilation des 2.
Au lancement de Shotcut je choisis une haute qualité (30 secondes donnent 100 Mo), j’obtiens donc un fichier vidéo d’une minute qui pèse 200 Mo. Le fichier de haute qualité conserve sa qualité, mais celui de mauvaise qualité n’a pas été amélioré par Shotcut.
Au lancement de Shotcut je choisis une mauvaise qualité (30 secondes donnent 20 Mo), j’obtiens donc un fichier vidéo d’une minute qui pèse 40 Mo. Le fichier de mauvaise qualité conserve sa qualité, mais celui de bonne qualité a forcément perdu en qualité.
En conclusion, au lancement de Shotcut, choisissez la qualité que vous souhaitez et laissez le faire en choisissant le mode automatique.
Here is a very simple reasoning.
I have 2 clips of 30 seconds, one high quality which is 100MB and one low quality which is 20MB. Total 120MB for 1 minute of video. I want to make a compilation of the 2.
When I launch Shotcut I choose high quality (30 seconds is 100 MB), so I get a one minute video file that weighs 200 MB. The high quality file retains its quality, but the low quality file has not been enhanced by Shotcut, it is still of poor quality.
When I launch Shotcut I choose a bad quality (30 seconds is 20 MB), so I get a one-minute video file that weighs 40 MB. The bad quality file retains its quality, but the good quality one has necessarily lost as.
In conclusion, when launching Shotcut, choose the quality you want and let it be by choosing the automatic mode.
Are these .mov files encoded as H.265/HEVC from a phone or camera? If so, that’s a major reason for the file size difference. The input files are using a higher compression (efficiency) codec than the export codec (H.264). So the output would have to be a larger file size to maintain equal quality because it is not equally good at compressing video.
Well that but also lossy + lossy = less quality for the same size. One needs to also carefully compare resolution of source with Video Mode in Shotcut (and not by increasing Export > Video > Resolution).
I have no idea, what the YouTube-Preset does.
Choose a video mode which fits the highest quality clips you have (resolution and frame rate)
Create your project
Export with those settings:
Deinterlacer & Interpolation: Best quality
GOP: Half of frame rate (so 15 for 29.97/30 fps or 30 for 59.94/60)
Check the box that says “fixed”
Set the CRF/quality setting as high as possible. I suggest at least 16 for good quality, although this depends on many factors. (Attention: lower number=higher quality!)
Audio: Whatever you prefer but keep in mind that the AAC encoder used here isn’t the best out there. For this reason I prefer Opus.
On the last tab, change the preset from whatever it’s set to (usually “fast” or “medium” but depends on other settings) to “slower” to get the best motion detection. (this will increase export times a lot, but it’s usually worth it, especially on videos with a lot of movement)
That’s it. There should be no visible difference in quality if you export with those settings (and choose at least 16 for the CRF value).
edit I just saw that you’re trying to keep the file size also. You have to play with the CRF value then, because with 16 or lower you will end up with a much higher bitrate (and therefore also filesize).
More Infos here: Encode/H.264 – FFmpeg
AFAIK Shotcut can’t enhance the quality of lower quality clips. All you can do is interpolate to a higher resolution. If you want to enhance the quality artificially you would have to use some other software.