You're being modest in describing it as an FFmpeg "wrapper". Although technically true, I can see a lot of thought and effort went into the UX, and I think that's where the value lies. Good job!
Yes that's exactly right, the extremely complex and clever code in ffmpeg that's understood by about 12 people in the world and is the basis for the entire media consumption in the world isn't where the value is, some gui frontend is.
I agree with the GP here. Yes, you can do all this with just FFMPEG, but now you have to wade through all the docs, flags, values to craft your CLI args with a bit of trial and error. The value add here is a coherent front-end that abstracts away all the complexity and lets you cut right to what you want. Further, if it's something you do infrequently, or your workflow changes, you have to rerun the gaunlet of figuring out what flags/values you need. FFMPEG can do a ton, but most people just need to crunch the file size, transcode, and occasionally convert a snippet into a gif
> …you can do all this with just FFMPEG, but now you have to wade through all the docs, flags, values to craft your CLI args with a bit of trial and error. The value add here is a coherent front-end that abstracts away all the complexity and lets you cut right to what you want.
Very true, but just want to note that ChatGPT is also great at creating CLI input for ffmpeg, yt-dlp, etc. And of course, Handbrake remains the O.G. for open source, cross-platform video encoding.
Right, the value add should obviously remain locked in the hands of a few elite professionals and the rest of the world should have to beg and grovel at their feet to get anything done. Any attempts to simplify workflows should be looked at with scorn and an aura of arrogance.
There are things that are simple to use and user-friendly, but Handbrake is not one of those. Personally I use the command line, but different people have different needs, and it looks like there is a market for some video conversion software that does not scare a normal person.
I don't think ffmpeg is actually _extremely_ complex. A lot of the complexity (all the flags) you feel belonging to ffmpeg actually belongs to encoders (libx264, libx265, aomenc, libvpx, all the hardware encoders from each vendor). They each have various options meaning different things. ffmpeg is just an interface for them.
And it's a genuinely complex problem space. ffmpeg supports some pretty wild audio/video transformations, especially once you start getting into filter graphs (e.g. combining multiple audio/video streams into a single output); its command-line interface reflects that complexity.