VP8 vs. h.264

[Prescience500]

The problem with x264 is that it's using the h264 codec, which is patented.

[popper]

The problem with x264 is that it's using the h264 codec, which is patented.

if your not producing AVC commercially or getting 10.000+ users/viewers then why do you care,its not costing you anything it's not like anything but AVC (and OC old mpeg2)is Already Everywhere in the world right now...

....from your basic STB free to air terestial,SAT,Cable,hand held DVB broadcast's, your IP stat-muxed Pro web content providers, your generic HD BR and even your basic home LAN 'PS3 media server' , and its only going get ever larger coverage as time passes, and OC the worlds SOC (System On a Chip)vendors provide ever more powerful AVC chipsets to play this latest IP AVC SD/HD content every day.

perhaps you should have a Good read of this Dark Shikari tech blog http://x264dev.multimedia.cx/ and pay special attention to his practical VP8 analysis.

after all it IS his and the other x264/ffmpeg dev's code that HAS actually being written and used Today.

other dev,s dont even seem interested in even trying to write anything independent of the existing ffmpeg VP8 code, never mind write something as comprehensive and FAST as ffmpeg code is today, hence why goggle themselves recommends that you read the ffmpeg code as relates to the so called VP8 spec today

[Mr James]

Perhaps by the time flash becomes obsolete we'll get a working FOSS player.

[evolution]

Personally, I prefer to use H264 over VP8. Altough it's commercial for corporate users, a common (personal) user like me, can use it without worrying about patents and other things ATM.

Furthermore, I think the video quality of VP8 videos is much lower (similar to H264 baseline, at best) than most H264-encoded videos, and H264 encoder (x264) performance is higher than the libvpx one.

That's my personal opinion,
Cheers

[Mr James]

Normally, FOSS versions of proprietary apps take forever to come out. But when they do, they kick the proprietary version's ass.
See AMD FOSS vs proprietary drivers for example. When my son's son sees feature complete and perfect FOSS support for my card it will so kick the proprietary version's ass.

Therefore, H264 is what we have to live with until a proper matching FOSS version comes along.

[BlackStar]

Normally, FOSS versions of proprietary apps take forever to come out. But when they do, they kick the proprietary version's ass.
See AMD FOSS vs proprietary drivers for example. When my son's son sees feature complete and perfect FOSS support for my card it will so kick the proprietary version's ass.

Therefore, H264 is what we have to live with until a proper matching FOSS version comes along.

FOSS doesn't mean what you think it means.

Let me explain. We have a perfectly fine FOSS H264 codec, called x264. It is FOSS and it is superior to any closed-source offering at this time.

The issue is that the H264 spec is patent-encumbered. This means that as soon as you touch the FOSS x264 codec, you leave yourself at the mercy of the MPEGLA consortium and their licensing terms.

Right now they are offering you a big, nice carrot: you don't need to pay for fewer than 10K views. You feel safe because you don't believe anything you encode will ever reach 10K views and the H264 codec is pretty good - quite tasty a carrot, isn't it?

However, you forget that the MPEGLA (Microsoft, Apple, ...) can change their licensing terms at any point. It is they, not you, who are in control. Today it's free but tomorrow it won't be - you can bet your ass on it.

We've seen this happen with GIF. We've seen this happen with MP3. And the sheep are flocking to H264 now.

But this time, we actually have a chance to fight back with VP8.

[Mr James]

I meant "matching in functionality", in that they both are the same thing basically - compression algorithms.
I was referring to VP8.

[Remco]

Yeah, if the x264 devs get on VP8 (and they do, I think), it will be pretty much H.264 quality. Maybe it doesn't get that last 1% in quality, but who cares? It's not like the money people care about quality. (YouTube...)

[Prescience500]

Libvpx and I'm sure ffmpeg's VP8 encoder/decoder are improving in performance and quality with every new release. Does anyone know if it's gaining on x264? One thing I've found interesting is that Google is also making a lossy image format called webp which is based off of VP8 video. Maybe we will EVENTUALLY start seeing VP8 and webp used being the formats of choice in some of the video recorders/cameras out there.

Popper, I care about patent free media for the same reason I cared enough not to accept government money/aid when I fell on hard times at one point. I think it's wrong. Maybe we'll ultimately have to wait for VP9 or something before we get a better specification that can truly be superior, but in the meantime, isn't VP8 a start?

[Prescience500]

P.S. It should be noted that webp is still in the alpha phase so it currently has some gaping issues right now. From what I've read, it seems clear that x264 developer is fairly biased with is analysis. I'd much prefer a newer analysis, using the latest encoders/decoders and made by a neutral party.