Using Shotcut 22.11.25, my resulting file, for 42-minute HD footage, is about 2.3 GB. The same footage, with the same settings (except when using a graphics card instead of a processor), after conversion in Shotcut 26.4 is almost 6 GB! What’s going on!
- 1920x1080p, 16:9, 60 fps
- mkv (result file)
- H.264 High Profile
- libx264 (nvidia)
- 65%
- GOP 120
I have a strange feeling that Shotcut is unreliable, because with every version, strange things happen, either the graphics card won’t work, or the files suddenly become huge, etc.
In general, how is it that the converted file is always so much larger, and if we adjust the settings to keep it the same size, the quality is worse? Couldn’t Shotcut automatically suggest one of its export settings so that the resulting file has the same image quality as the input file? Then we could decide whether we wanted a smaller, lower-quality file, or whether we wanted to keep the original quality of the edited video.
Which input file? It is a multitrack with potentially many files. Basically, you do not seem to understand lossy video compression, and it is explained here many times already even in FAQ . So, basically the “same image qualtiy” would require lossless, and then the file will be much larger than what you are experiencing.
we could decide whether we wanted a smaller, lower-quality file, or whether we wanted to keep the original quality
You can already do that. BTW, you picked “High” profile, which is higher quality % than the default. And I already mentioned lossless.
Feel free to use something else because it is clear I am never going to be able to give you what you want.
The quality % for different encoders is not exactly the same, and libx264 and NVIDIA are different encoders. Also, between 2022 and 2026 the default quality % for the H.264 High Profile preset increased by 10%, which is NOT a direct linear increase in file size. You would need to compare the exact same settings. Try using the defaults or “Default” preset without hardware encoder; it uses 55%.
That’s not what I had in mind. I meant a simple project where, for example, we’re recording something from a TV in HD, and then we want to remove unnecessary parts, like commercials, and export the video file so we can watch it. Recording at specific settings, like 1920x1080 and 60 fps, gives us a certain level of video quality. However, even though the commercials are removed, the converted file is still much larger than the edited one. If we use a setting other than High Profile and 65% (B frame - 3), the image quality is worse than the original file.I compared the exported files and checked it. Besides, in that case (I mean one file, not multitrack), I never know, because the program doesn’t suggest it, what settings to use to achieve the same image quality after export (I’m ignoring the exported file size at this point). The program doesn’t suggest this. Everything has to be done through experimentation. And that’s my point.
Thank you for the tip, but for me it’s a bit tiring, constantly figuring out new settings in each new version of the program, I prefer to stick with the old version of the program and the settings I’ve already checked, even if the program doesn’t support the graphics card (I don’t know why) and the export takes a long time.