Originally Posted by Dick Applebaum
I agree with most of what you say -- except I don't like how Google is trying to ditch h.264 -- proven to be the best available codec.
I followed your link on snow. Rather than doing a web alamende left, could you help define snow? What are the advantages of snow? How does snow compare to h.264 accuracy/performance/fps/bandwidth/ file size? When will snow be available for general use?
FINALLY, AN INTELLIGENT REPLY!
This forum barely has any, which is why I'm quite surprised.
Well, as long as you're not liking the idea of ditching H.264 because Apple's supporting it, I'm fine with your point of view there.
I just support because of the promotion of open, royalty-free standards. (Which is impossible because of WebM's nature, but it's one more browser.) It still doesn't make any sense for this suicidal move... but oh well.
Back to your reply - Snow is/was an experimental codec from FFMPEG, a well know swiss knife encoder.
Going into nitty-gritty details, it's a wavelet codec. The most famous (if you lower the standard of famous) is Dirac (http://en.wikipedia.org/wiki/Dirac_%...sion_format%29
), which is another wavelet codec developed by BBC, potentially for their own use. (Yes, the British BBC.)
Going back to the real world, some testing shows that it performs much better* than x.264** - good compression (more clarity for the size). FPS? If you're talking about encoding it, that I don't know. You'll have to pick up a copy of ffmpeg and encode a Snow video yourself.
Accuracy/bandwidth/file size is what I've just answered above - it does pretty well, if not better, than x.264.
Performance? That's also for you to figure out.
It certainly is NOT a massive hardware supported codec, due to the fact that no one has even heard of it.
Finally - it's available for "general" use, if you download FFMPEG and run it with the proper commands/args - but it's not popular, and it's very much hidden away, since the project kind of died. So it depends on what you mean "general". It's still beta, last time I checked.
I wish there was a massive corporate movement to gather all the smart mathematicians, as well as lawyers, and let them create a patent free efficient codec that we can all use. That way, there would be no need for silly debating, and HTML5 might actually become reality (believe me, as a web developer and video editor, the future looks dim for HTML5).
I'd prefer to continue this discussion over PM/email/IM, since this forum isn't exactly the best place to discuss.
*Do note that the tests were performed a long time ago, and that x.264 could have been improved substantially. You are on your own to benchmark them, both visually and quantitatively.
** I use x.264 as a reference instead of H.264. They essentially are the same codec, but x.264 is an encoder of H.264
, and has been known to be very good at compressing H.264. (Efficient, fast - you name it!)