H.264 vs. flv -- which is better?
I currently encode our videos to flv files for web and post them to our websites in that format. However, with the rise of mobile and tablets that don't support flv and HTML5 that has chosen H.264 for video tags support on the way, I think I need to take a look at how we are approaching video on our company websites.
Could anyone please give me their thoughts on FLV versus H.264? Is one any better, slower, faster, larger, smaller?
I don't really see much difference on the quality end, but I know that there is pros/cons to both. Any input is much appreciated. Thank you very much!
Video Production Manager
Shaw Suburban Media
Your question hints at the tricky underlying problem - better in what sense? http://en.wikipedia.org/wiki/Flash_Video
H.264, a subset of mpeg4, is clearly a decent codec offering good quality at low files sizes. Its support on a range of platforms is good, and some see it as the future of web video.
H264 appears to be broadly equivalent to flv / VP6 in the quality versus filesize tradeoff, with Apple users often seeing more of a slight advantage to H264 than other users report. For most non-technical viewers the quality advantage will be unnoticeable, and is likely to be overwhelmed by differing choices in data rate and picture size, alongside original programme quality.
Flash video support for transparency (and I think cue points) currently requires flv / VP6.
H264 is a complex codec making significant hardware demands. While H264 is good on newer and faster machines doing one thing at a time, on the "long tail" of systems out in the wild it seems likely that more users will experience problems from the greater demands of h.264. Older test systems here that play flv / vp6 smoothly sometimes struggle to decode h.264
If high definition video becomes more prominent on the web, this issue may become more pressing. On2 are already offering (in their proprietary encoders) a simplified version of vp6, to make for easier playback of high definition video (at the price of a little quality loss.)
Apple does seem to have made a surprising choice, with no Flash support in the iPad (like the iPhone), given Flash's near-omnipresence on the web just now.
Will Apple sustain this in the face of user reaction to a Flash-less web experience?
While people are perhaps willing to accept that the web experience on phones is quite limited (with a tiny screen, what could we expect?), how will they react to a Flash-free web in an iPad-sized device?
It will be interesting to see how that plays out.
Today, Flash is the dominant force in web video. The Flash player supports H.264. But making video available for the iPad will mean a page that does not call Flash to load the video, but presumably relies on a system player (QT Player, Windows Media Player) either embedded in the web page or called via a file link - either way, not choices that many web designers have been making much for a good while now. The Flash player has advantages.
Supporting the iPad may be putting developers into a position where they need device-specific pages, or they have to change their choices on their existing pages.
It's not really clear to me what Apple's strategy here is. My guess is that they are hoping to shift as many as possible onto H.264.
HTML 5 may be a long time coming. It has already been a long time in the negotiation. There have reportedly been some issues over choices of video codec and format support. When and in what form HTML5 will be finalised, and then supported by browser writers, remains to be determined.
The pluses of a non-proprietary set of formats and codecs for online video could be offset if we were to revert to needing to supply multiple versions in different encodings. The joys of offering QT, Windows Media, Flash and Real versions of each video were limited. But it seems we may be already back there, or heading part of the way back there, for those wanting to support both mainstream web and the iPad experience.