Why burn subtitles into the video, when you can use the vtt or srt file that whisper can output?

Ya i agree, no need for that step why burn in subtitles, just use a player that supports .srt

> ... "just use a player that supports .srt"

Which is pretty near every video player in existence these days. Hardcoding the subtitles into the video is just plain rude innit?

To have control over how they appear, and also to appear in social media videos, especially shorts.

Nice work

I couldn't agree more. Having more control is crucial for me when writing this article. Thank you

I was pleasantly surprised to read the walkthrough as a quick start instead of install it.

One step that you explain that I rarely see explained for beginners is how to instantiate a python venv. Hard to unsee.

I’m installing this now do you have any other feature plans for this?

Thank you for your feedback, I appreciate it. Do you have any feature in mind that you would like to be implemented?

I had written a similar post using whisper.cpp and ffmpeg some months ago https://www.spapas.net/2023/05/22/ai-auto-subtitling/

I might be crazy - but in this article about adding subtitles to a video, did it feature (twice) a video with no sound or subtitles?

Sorry, I just updated the video example

in the website, you forgot to activate the virtual env.

I updated it, Thank you

Wouldn’t it be far better to pass a list of options into the subprocess call instead of joining it all into a string and then using shell=True?

Nice. Does this support speaker diarization and output of subtitles to json format?

It uses this, which does support diarization: https://github.com/m-bain/whisperX

Is it just me or does the first example actually nail the first 2 lines but then doesnt have the subtitle for "we hope you enjoy it"

I just updated it. Thank you