On2 VP8

[Manao]

Bond : golden frames are indeed like references in h264, but limited to 2 references max : the previous reference, and another one, that can be updated whenever they want. It seems it can be updated piecewise too, which makes it a bit more flexible than 2 references with h264. As for the usefulness of that, I don't know. It must help in some contrive cases, but beyond that... In any case, it will help less than bframes & bpyramids would, but they can't use those because they are afraid of patents (though bframes should be patentless by now, I think).

Their description of their loop filter and its adaptivity to motion doesn't impress me. h264 loop-filter also adapts to motion (though the way it adapts to motion seems different from VP8's).

Their visual results for h264 seems a bit strange, since I seem to see a good old IDCT mismatch drift during the GOP, which can't happen with AVC. As stated by nm, it more looks like ASP than AVC.

As for the PSNR curves, they must always be taken with a (huge) grain of salt, wherever they come from. Since they don't say which encoder and which settings are used, however, the curve most probably are correct. They just don't mean anything and can be ignored.

Dark Shikari : sagittaire is right, when it came out, VP7 was lighter than existing AVC decoders, and more than hold its ground quality-wise. Our encoder, at that time, only managed to consistently beat it when we included High profile.

[CruNcher]

Yeah about the PSNR results they dont say alot tough for H.264 basicly with inloop deblocking it means higher PSNR and most times looks awfull tough with VP8 it doesn't have to be the same because they use a complete different deblocking aproach which seems really interesting , though 50% is a heavy claim thats like they wanted to say: "hey look we drove ahead of MPEG before they even get started with H.265"

[IgorC]

Wonder if they will go to VFW?

[Sharktooth]

i wonder if they will release a free-for-personal-use encoder.
then speculation will end...

[LoRd_MuldeR]

Originally Posted by Sharktooth

i wonder if they will release a free-for-personal-use encoder.
then speculation will end...

Plus a free-for-personal-use decoder library, of course

[CruNcher]

Maybe this also implies they gonna Open Source VP7 now

[slavickas]

Originally Posted by CruNcher

Maybe this also implies they gonna Open Source VP7 now

well they promised to release VP4 as open source and it never happened ( big IINM )

[Inventive Software]

VP4's decodable in ffmpeg, yes no? As is VP5, yes no?

[smok3]

i think they missed all the trains, at least i can't see what could happen, even if they would develop the world-best-codec.

[LoRd_MuldeR]

Originally Posted by Inventive Software

VP4's decodable in ffmpeg, yes no? As is VP5, yes no?

Most likely this is thanks to the reverse engineering capabilities of the ffmpeg developers, not thanks to On2...

[Sharktooth]

i wonder if on2tech is still around.
his last forum "activity" was October 2007.

[HKisd]

Some new info about On2 VP8 available.

http://www.dspdesignline.com/howto/214303691

[IgorC]

Originally Posted by HKisd

Some new info about On2 VP8 available.

http://www.dspdesignline.com/howto/214303691

Old news. The same hype about magic adaptive inloop filter and golden frames etc.

I think it isn't correct thing to wait really new technology from On2.
VP6 was H.263++(-like) codec. As inverse engineering shows VP6 uses reference frames, deblock as well as H.263++ or H.26L

VP7 is mostly H.264-like codec.

Right now there is some work around H.265.
http://yufeng1684.bokee.com/6776683.html
http://iphome.hhi.de/suehring/tml/download/KTA/

on2 may just buy licenses for some compression tools from big boys.

There is some information that VP8 may be released in the end of April (after more than 6 months of waporwaring (farting) and paper release and rumors from 2007)

[Dark Shikari]

Originally Posted by IgorC

VP7 is mostly H.264-like codec.

Doubtful, considering it doesn't even have B-frames, last I recall... and given that in many cases it came out worse than MPEG-2, I don't think we can expect much from them.

[*.mp4 guy]

The first one is better. You can't measure quality solely based on the amount of background noise present (I say noise, as the mpeg2 is a sea of artifacts). All this picture demonstrates is that vp7 could benefit from aq, which is incredibly ironic considering how long x264 needed aq before it got it, and considering that people who said x264 was killing details were often shouted down, until aq is released, then its the reverse bias to a similarly absurd extent.

[Dark Shikari]

Originally Posted by *.mp4 guy

The first one is better. You can't measure quality solely based on the amount of background noise present (I say noise, as the mpeg2 is a sea of artifacts). All this picture demonstrates is that vp7 could benefit from aq

If you look at all the other images from other encoders--including other H.264 encoders--even those without AQ, the correct conclusion seems to be that VP7's deblocking filter is far too strong. Mainconcept, for example, doesn't have near the blurring problem that VP7 does despite not having AQ.

AQ helps solve the problem, but it's not a scapegoat.

And yes, you can measure quality based on the amount of background noise. It's the job of an encoder to make the result look like the source to the human eye, and if background noise is what is necessary to do that, then so be it.

[IgorC]

@DS

Both screenshots you post look badly. VP7 is way smooth but MPEG2 is a soap of macroblocks. And please, can you post original.png too?
I marked the few areas of you screenshot where MPEG-2 had simply unacceptable quality.http://img11.imageshack.us/my.php?image=mpeg2marked.png

One encoded video isn't representative of whole codec efficiency.
Maybe VP7 is worse than MPEG-2 in 2% but is better in other 98%.

Here is other example (2 pass mode, 1500 kbit/s)
Source: lossless
ftp://ftp.tnt.uni-hannover.de/pub/sv...rig_02_yuv.zip

MPEG-2 (HC encoder, best quality)
http://img19.imageshack.us/my.php?image=hc1500.png
VP7 (vp7.0.10.0, best quality)
http://img19.imageshack.us/my.php?image=vp71500.png
Original
http://img19.imageshack.us/my.php?image=originalu.png

Encoded videos:
http://www.mediafire.com/download.php?u2onmmztgui

Actually VP7 is good as x264 was in 2005 (mostly first middle of the year). Then on2 has stopped to improve VP7.

[*.mp4 guy]

I'm not sure what the fuss is about, marketing departments always exagerate beyond all semblence of reality, and its always nice to have more competition, and more options, though I am quite certain i will be sticking with x264 for the forseable future.

I think you have missed my point, or are avoiding it. vp7 is clearly better here, look at the text, look at the pillars, look at the wing, look at the main character, and the foreground effects. vp7 does look like its filter is too strong, but you are aware that it is adjustable on playback, right?

I'm not saying that x264 isn't most likely better then vp8, or that vp7 is still competetive with x264 (though it must be said that x264 is not representative of the state of avc as a standard). However your posts in this thread show clear bias. Vp7 was as good as they claimed it was when it was released, or rather it was close enough, and much more capable of living up to the marketing then most codecs.

I don't used vp(insert number), but I have tested most of them, 3 is about on par with mpeg2, 4 and 5 don't apear much different then 3 to me, 6 was better then asp, but had poor rate control, 7 was competetive with the best available codecs upon release, but was extremely slow to encode with good settings, and was eventually eclipsed by avc, in particular ateme and then x264. I expect vp8 to be a good codec, of similar quality to x264, and if it is released for personal use, it will be interesting to test.

Surprisingly, I agree almost completely with Sagittaire.

[Dark Shikari]

Originally Posted by IgorC

@DS

Both screenshots you post look badly. VP7 is way smooth but MPEG2 is a soap of macroblocks. And please, can you post original.png too?
I marked the few areas of you screenshot where MPEG-2 had simply unacceptable quality.http://img11.imageshack.us/my.php?image=mpeg2marked.png

One encoded video isn't representative of whole codec efficiency.
Maybe VP7 is worse than MPEG-2 in 2% but is better in other 98%.

Here is other example (2 pass mode, 1500 kbit/s)
Source: lossless
ftp://ftp.tnt.uni-hannover.de/pub/sv...rig_02_yuv.zip

MPEG-2 (HC encoder, best quality)
http://img19.imageshack.us/my.php?image=hc1500.png
VP7 (vp7.0.10.0, best quality)
http://img19.imageshack.us/my.php?image=vp71500.png
Original
http://img19.imageshack.us/my.php?image=originalu.png

Encoded videos:
http://www.mediafire.com/download.php?u2onmmztgui

Actually VP7 is good as x264 was in 2005 (mostly first middle of the year). Then on2 has stopped to improve VP7.

Here's the full comparison with ~12 encoders. mapping.txt stores the mapping of numbers to encoders.

If after looking through this you can still say with a straight face that VP7 is competitive with AVC--let alone ASP!--you're clearly just shilling for On2; there's no other reasonable explanation.

I myself thought VP7 wasn't too bad until I ran this test. The results shocked me as well. Maybe it's just this video, but it's rather biased to go and write off all situations where an encoder fails as a "pathological case" and only accept those that it does well on.

IMO VP3 is roughly competitive with MPEG-2, VP6 with ASP, and VP7 is roughly competitive with VC-1 on average (but obviously, in some cases, it can be much worse). VP6 doesn't seem to have the kind of pathological cases VP7 has.

[IgorC]

Originally Posted by Dark Shikari

Here's the full comparison with ~12 encoders. mapping.txt stores the mapping of numbers to encoders.

???
Who asked you to post them?

Originally Posted by Dark Shikari

If after looking through this you can still say with a straight face that VP7 is competitive with AVC--let alone ASP!--you're clearly just shilling for On2; there's no other reasonable explanation.

Don't put in my mouth the words I don't say. I didn't say that VP7 has the quality of todays H.264 encoders.
I do say: VP7 is H.264 like codecs. (speaking of generation, not quality). Actually VP7 is good as x264 was in 2005 (mostly first middle of the year). Then on2 has stopped to improve VP7.

Originally Posted by Dark Shikari

I myself thought VP7 wasn't too bad until I ran this test. The results shocked me as well. Maybe it's just this video, but it's rather biased to go and write off all situations where an encoder fails as a "pathological case" and only accept those that it does well on.

Do you have another video on you hdd that isn't your PC game video horror.
What part of the statement you didnt understand?
One encoded video isn't representative of whole codec efficiency.

You get on people nerves easily.
Manao, Sagi, me and mp4 guy had told you:
VP7 was comparable to x264 in 2005 and it's way better than MPEG2, ASP.