petzku / aegisub-scripts Goto Github PK

Output from util.io.run_cmd includes the command prompt on Windows, e.g.:

C:\Users\dad\AppData\Roaming\Aegisub\log>ffmpeg -i G:\e.mp4 -f rawvideo -pix_fmt rgb24 - 
[actual output starts here]

This makes it impossible (or at least unreasonably difficult) to use things from stdout, as there's no way of telling where the actual data starts.

Test script that demonstrates the problem, and its outputs. ffmpeg_out is what it should output (and what ffmpeg does if writing to a file), and petzku_out is what it actually outputs. You'll probably want to use the small test video so you don't fill your drive if you run it yourself (it outputs raw video).

Would prefer to have it fixed here, and not have to do if windows this else that, but if it's not feasible I'll manage.

cc @Zahuczky

Add an option for the script to add a new time (beginning and end of the line) to the move tag. This is convenient for some signs.

When there are multiaudio (english-japanese) and I select japanese in aegisub, the script takes the first one (in that case english)

It is likely that I did not understand how this script works but when in doubt I write the problem.
Technically it should create sequences based on the selected times using MPV (which I put in the feedDump folder) latest version for windows.
Whether you use RAW or subtitle mode he exports the clip with subs overlayed.

Failed to open D:\Mux[Ñ] 027.mkv.

[encode] no data written to target file
Exiting... (Interrupted by error)

PD. FIX: https://discord.com/channels/131816223523602432/710562732973490238/872225912387690527

Happens on 23.976 fps videos, not on 30 fps. Probably need to test other framerates.

Files will contain the last frame, but some players (especially web-based ones) seem to not show it. MPC-HC does (even when looping to start), mpv does not mpv skips to the next file just before the last frame if a playlist is loaded, but shows the frame if there is no next video and keep-open is not no. Browser video players seem to pause before the last frame. Probably an issue with file timestamps.

Sample file and last two frames: