Linked by Thom Holwerda on Mon 25th Mar 2013 21:09 UTC
Legal Late last week, Nokia dropped what many consider to be a bomb on the WebM project: a list of patents that VP8 supposedly infringes in the form of an IETF IPR declaration. The list has made the rounds around the web, often reported as proof that VP8 infringes upon Nokia's patents. All this stuff rang a bell. Haven't we been here before? Yup, we have, with another open source codec called Opus. Qualcomm and Huawei made the same claims as Nokia did, but they turned out to be complete bogus. As it turns out, this is standard practice in the dirty business of the patent licensing industry.
Thread beginning with comment 556795
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[6]: Big picture...
by lemur2 on Wed 27th Mar 2013 07:35 UTC in reply to "RE[5]: Big picture..."
lemur2
Member since:
2007-02-17

"
Your recollection is utterly inaccurate. The conclusion was that, for video on the web resolution and bitrates, if you are prepared to put up with a slower encoding speed, then for a given number of bits you absolutely can make at least as good if not better encoded video with webm as you can with h.264.

You have to go to extreme high resolution/bitrate/quality profiles before there is any noticeable advantage (other than encoding time) for h.264 that webm cannot match reasonably closely. Such videos simply aren't used over the web.


Then see for example on page 27 (the same page I pointed out in the old thread) how x264 constantly out performs VP8 at ANY BITRATE. How can you still say that VP8 is better than x264? Do you have evidence?
"

All one has to do is go to a higher profile (at the cost of encoding time). You would argue that one can go to a higher profile also for h.264, and that is true, but one can go to a higher profile again for VP8, and so the two chase one another until we reach such a high profile that it is beyond what is used on the web.

As I said, for any h.264 video such as might be actually used on the web, for a given number of bits, it is possible to get a VP8 video to match the h.264 video as long as one is prepared to take a longer time to encode the VP8 file (normally done by having to use a higher profile).

This does not say that VP8 is better than h.264, it merely says that one can get the equivalent quality per bit (as far as any blind test can tell, within the range of quality as is used over the web), provided one is prepared to take the hit in encoding time.

Since videos over the web are commonly decoded & rendered thousands, if not millions, of times more often than they are encoded, if there is one area where a compromise has to be made, the encoding time is the best area to choose.

Edited 2013-03-27 07:36 UTC

Reply Parent Score: 1

RE[7]: Big picture...
by saynte on Wed 27th Mar 2013 07:49 in reply to "RE[6]: Big picture..."
saynte Member since:
2007-12-10


All one has to do is go to a higher profile (at the cost of encoding time). You would argue that one can go to a higher profile also for h.264, and that is true, but one can go to a higher profile again for VP8, and so the two chase one another until we reach such a high profile that it is beyond what is used on the web.


False, the high-quality presets were used in both cases. That test represents the upper limit, that's why I used it. There is not an infinite ladder of profiles, there are just a handful.

Seriously, can you stop now? You haven't presented any hard evidence, so there's nothing to discuss until you do.

Reply Parent Score: 3

RE[8]: Big picture...
by Radio on Wed 27th Mar 2013 08:34 in reply to "RE[7]: Big picture..."
Radio Member since:
2009-06-20
RE[8]: Big picture...
by lemur2 on Wed 27th Mar 2013 09:35 in reply to "RE[7]: Big picture..."
lemur2 Member since:
2007-02-17

"
All one has to do is go to a higher profile (at the cost of encoding time). You would argue that one can go to a higher profile also for h.264, and that is true, but one can go to a higher profile again for VP8, and so the two chase one another until we reach such a high profile that it is beyond what is used on the web.


False, the high-quality presets were used in both cases. That test represents the upper limit, that's why I used it. There is not an infinite ladder of profiles, there are just a handful.
"

Sigh! Most video on the web doesn't use high quality presets, but rather a low quality or mid quality preset. Because they aren't the same codec, the presets are not directly comparable. Sometimes to get comparable quality for VP8 you have to use a higher profile. Sometimes VP8 is better than h.264 at the "same" profile. The curves don't follow the same slope, they cross over.

As poster Radio pointed out, in the recent tests done by ITEF, the quality curves for VP8 and h.264 cross over each other at the extremes, and are all-but-identical in the middle::

http://downloads.webmproject.org/ietf_tests/vp8_vs_h264_quality.htm...

If you want to get the same quality for webm as you get for h.264, then in some cases only you have to go to a higher preset (sometimes you don't), but it will cost you in encoding time.

http://downloads.webmproject.org/ietf_tests/vp8_vs_h264_speed.html

Seriously, can you stop now? You haven't presented any hard evidence, so there's nothing to discuss until you do.


Seriously, can you stop now? Why are you trying to pretend h.264 is better than VP8 when there is effectively nothing in it (apart from VP8 longer encoding time)?

Who is paying you?

Edited 2013-03-27 09:42 UTC

Reply Parent Score: 1

RE[7]: Big picture...
by zima on Mon 1st Apr 2013 22:42 in reply to "RE[6]: Big picture..."
zima Member since:
2005-07-06

Since videos over the web are commonly decoded & rendered thousands, if not millions, of times more often than they are encoded, if there is one area where a compromise has to be made, the encoding time is the best area to choose.

I don't know, a strong majority of YT videos is mostly ignored... only very few reach "thousands, if not millions" views.

Reply Parent Score: 2