Reputation: 9049
I've been working with the video/canvas capabilities of the new browsers I haven't been impressed.
I was thinking I could get better results making large sprites containing each frame of my videos. I want full screen video, so I guess I would have to scale the images, or canvas accordingly.
Of course, to limit the http requests I want a very limited number of sheets. But, that also means they will be extremely large. Let's say 1280x720.
Is it insane to use that large of sheets?
Are there any automated scripts that will make these sheets since they will be frames of rows and columns?
Lastly, I'm finding compressed video of better quality than compressed images of equivalent size. Why is that? Is it possible to yield similar results with images or is Photoshop's save for web not as good?
Thanks
Upvotes: 3
Views: 922
Reputation: 1348
At it's most basic, video compression works by taking keyframes every so often and then storing frames between keyframes as the change between it and the last frame. That means if things don't move between frames they are not stored again. This cuts down the file size significantly. There are other things done, but that's the simplest. This means that no matter how good the compression on a set of individual frames, as long as there is some similarity between the frames the video compression will always achieve a better result.
As a question, why can you not just use the HTML5 video tag?
Upvotes: 1