I don't like the auto-scrolling of the timeline when the red bar reaches the halfway point it's too jumpy.
Consider hosting some sample clips for people to play with.
I was able to wrest back some of my software by starting with
LIBGL_ALWAYS_SOFTWARE=1
This forces CPU rendering through bypassing the intel GPU shitstack.
Or that's at least what I think happened. Took a loooong time to figure out.
Edit: I should include that part of the issue might be hamfisted aspects of the updates that penalize variable frame rates, which phone cameras use, and also comprises 100% of my videos. Still not sure though.
It is worth noting that I did a lot of research in this space when I was working on my own video editor and there are a lot of these now (I have about 30 in a spreadsheet). Ever since there was a ffmpeg WASM port many people have made these web based editors. It is so cool we live in a time where that kind of powerful technology is freely usable by others to build these tools.
My angle was to add gen ai video as a first class feature. That might be something you want to investigate now that video gen apis are becoming more common.
No, it doesn't need that. It just needs to be a good web-based video editor.
If he just wants a good website to do the job, sure, genai is uselesse
But if he wants to market it, in this day and age, it's practically a need.
No, it isn't. AI content creation is a fundamentally different thing that a video editor. Maybe you can't tell the difference, but I can. AI isn't going to edit the video how I want, it isn't going to create the video I want from a prompt when I've already got the video clips I want to edit. I don't need to edit clips of people with 7 fingers, I need to edit my family vacation video, or whatever it is I need a video editor for. There's nothing about AI that would improve this as a video editor, at all.
If I could import this project into my existing React-based web app, I will use it to edit clips of content for my specific application, and for that it is useful. AI would just bloat it and be annoying where it isn't needed.
You act as if there aren't a bunch of non-AI video editing apps that try to do exactly that, nevermind AI apps that try and do that.
> I don't need to edit clips of people with 7 fingers
There's more to AI than image generation (and 7 fingers isn't the same problem it was 6 months ago). Specifically, picking out the important part of the clip, adding appropriate captions and music from the AI processing the video.
> I need to edit my family vacation video
Yes. What do you think an AI enhanced video editor would do? Be Photoshop? Editing the family video from hours down to a digestible video for sharing is a time intensive process, and while I'm sure you enjoy it, not everyone does. A "Make vacation video" button for those people for whom that's a chore and not fun, and don't remotely care to be semi-professional video editors is the target market for a program with a feature to find decent photos and snippets of video from a giant pile of stuff, and arrange them into something cromulent.
As far as state of the art for generated content, a cousin of mine used AI to take a still image of the extended family at dinner with their hands under the table, and turn it into a convincing video where everyone raises their hands and waves (with the correct number of fingers, in case that needed to be stated.)
Anyway, to answer your question, the react library backing this for building your own to drop in is called Remotion. If you'd rather a different video editor, https://omniclip.app/ was mentioned as being an open source no-upload video editor.
No, I do not want AI automagically editing my home videos for me. It does not know the moments that are important to me as well as I do. To suggest that AI will somehow do a wonderful job at that is ignoring how bad current AI really is. It's become a running joke.
I'm glad someone else downvoted your comment because I can't.
Aborted() The progress bar is experimental in FFmpeg WASM, so it might appear slow or unresponsive even though the actual processing is not.
Redacting an area of the screen is easy with a bit of practice, even though the location of the password on screen may move, you just have to trace a path for the black box to move along. However, thattakes a bit of time unlike a hypothetical hide password with AI button.
I will happily pay for this tool. AI to show me the part where there are no humans, I will pay more :) by ofcourse all locally and no cloud shenanigans. Of course people saying why record so much, why do we do anything ? :) try to be optimistic nihilist
Move them out as early as you can and do regular backups. Videos of your child are much more precious than a whole backup solution.
I tried exporting, but it's quite slow and seems to be using CPU only. Shouldn't it use GPU for rendering?
jony1266•8mo ago
some feedback so far:
- allow me to drag + drop video clips to add them to the media section
- allow me to trim the beginning of a clip - right now i can only trim the end of a video clip
- allow me to drag the playhead (right now i can only click to move the playhead somewhere)
- make most of the text on the page be select-none - there were many times when i tried to drag something and it selected a bunch of text on the page
mohyware•8mo ago
jacktheturtle•8mo ago
purplecats•8mo ago
mohyware•8mo ago