Blur vs blocking

[Sagittaire]

http://www.on2.com/image.php?blob_id=539

What is your prefered picture: VP8 or H264?

Object of this pull is not to know if VP8 or H264 are better but just the picture and only the picture

[Dark Shikari]

Obvious troll thread?

If you're going to do a comparison, start with two fair images, then ask the question of blur vs. blocking; don't start with images that are from a test that is known to be biased (with old version of x264 with bad settings versus VP8 with best settings). :rolleyes:

I mean seriously, using a marketing image for real visual quality testing purposes? Is this a joke?

[wyti]

Claerly h264, if you're not too close to your screen.

Look at the background, in vp8 there isn't any details.

Even falsified test it's clear, i fear what juge differencies we would see in an honest comparaison

[Sagittaire]

don't start with images that are from a test that is known to be falsified

Prove that before ... and stop to watch XFiles with your falsified glasses ... ;)

[G_M_C]

Obvious troll thread?

If you're going to do a comparison, start with two fair images, then ask the question of blur vs. blocking; don't start with images that are from a test that is known to be falsified. :rolleyes:

I'll go with DS on this; Make 2 images from another source, not copy On2's on this. We dont know what settings they used to bias this comparison in favour of their own codec. These two images were used as marketing subject, and thus will by definition not be neutral.

So lets make an encode, based on a known source (like Elephands dream) and put those up for comparison.

[avivahl]

Obvious troll thread?Yep.
No point to even respond. :/

[Sagittaire]

I'll go with DS on this; Make 2 images from another source, not copy On2's on this. We dont know what settings they used to bias this comparison in favour of their own codec. These two images were used as marketing subject, and thus will by definition not be neutral.

So lets make an encode, based on a known source (like Elephands dream) and put those up for comparison.

Yes perhaps ... I don't know ... but it's not the thread subject. I want just to see the prefered picture. H264 picture or VP8 picture ... just for my curiosity ... lol

[shon3i]

I prefer one which closer to source. But i still prefer blured image over blocked. Aslo i think that best H264 encoder can beat VP8 easy, and this test are unfair completly, until we get VP8 for test

[microchip8]

h264 definitely looks better to me. If the vp8 was a bit more blurred, the background will look as if the guy is standing before a mixture of colors and vague distinguishable objects and not people and other things. Definitely vp8 is not good here

[Teddl]

Hi there!
At the first look, I would say blur because I do everything to avoid blocking of any kind if possible. But if I don't need to blur I don't do it if I encode something and less is usualy better.
Comparison between different codecs is not needed for me to make this statement at all. ;)
Teddl

P.S.: VP8=blur & H.264=blocking? What the..

[nakTT]

http://www.on2.com/image.php?blob_id=539

What is your prefered picture: VP8 or H264?

Object of this pull is not to know if VP8 or H264 are better but just the picture and only the picture
For still images (pictures) that you posted here, I'm more inclined to choose VP8 over H.264 (Blurring over Blocking).

However, from my (limited) experience, still image not necessarily represent the actual perceive quality of a video. Perhaps you could provide me with the video so that I can have a better look before I cast my vote.

[nakTT]

Obvious troll thread?

If you're going to do a comparison, start with two fair images, then ask the question of blur vs. blocking; don't start with images that are from a test that is known to be biased (with old version of x264 with bad settings versus VP8 with best settings). :rolleyes:

I mean seriously, using a marketing image for real visual quality testing purposes? Is this a joke?
IMHO, if the comparison was fair, I don't think we will get that kind of still image in the first place.

[Selur]

don't like both of them, first one is too blocky and second on is too blurry
but if the content should be watched from decent distance I would prefer the blocky one, if I have to watch it while typing at my monitor I would prefer the smooth one
so for a movie night or something like that the blocky one for me :)

Cu Selur

[Guest]

This is in the genre of "what is best?", its motivation is dubious, it spams a company's marketing material, and it's impossible to perform other related tests, etc. So I exercise my discretion to close this. Appeals to Doom9 please.

[benwaggoner]

We should also be watching these in motion.

One of the annoying things about blocking artifacts is that the basis pattern is static even when the content is moving, which can be quite distracting.

Also, this picture really isn't showing blocking per se; there isn't an obvious block pattern. We're seeing artifacts, certainly, but blocking is a specific artifact type which we're not seeing here (and in general H.264 isn't prone to blocking much).