Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > Video Encoding > High Efficiency Video Coding (HEVC)

Reply
 
Thread Tools Search this Thread Display Modes
Old 22nd January 2017, 19:49   #1  |  Link
mandarinka
Registered User
 
mandarinka's Avatar
 
Join Date: Jan 2007
Posts: 729
Post-HEVC technology development: JVET "Next Generation Video Codec"

Today I accidentally found this: http://www.cs.brandeis.edu//~dcc/Pro...s-Sullivan.pdf
Pages of interest: 10-11 and 16-17.

I have not seen this reported anywhere yet (but maybe I just don't pay enough attention), but apparently there is work on post-HEVC next generation compression, "Next Generation Video Codec" in the mpeg & companions land, the so called JVET (joint video exploration team). Apparently there is already second version of the research/reference encoder "JEM".

The slides have some measured numbers of JEM 1 and JEM 2 comapred to HM and list the proposed techniques, for example more intra modes (67), "adaptive loop filter", cross-component prediction (chroma from luma, was proposed for HEVC already, so this is not just Daala idea , Overlapped block motion compensation, 64x64 transform, up to 256x256 CTUs, and others (look at the mentioned pages of the pdf).

mandarinka is offline   Reply With Quote
Old 23rd January 2017, 04:45   #2  |  Link
hajj_3
Registered User
 
Join Date: Mar 2004
Posts: 1,120
if they think that this is going to be widely used outside of broadcast then they are sadly mistaken. AV1 is going to take over the web. For their sake lets hope they create a single patent pool this time with sane licencing fees.
hajj_3 is offline   Reply With Quote
Old 23rd January 2017, 15:09   #3  |  Link
turab
Registered User
 
Join Date: Apr 2012
Posts: 38
I believe they've changed the internal name, but anyway, those figures where they've summed the percentages is quite misleading. They've assumed the techniques are completely independent of each other, but are they? Also, when there's a rate reduction of 0.1%, for example, how do we know it's significant? It could be a rate increase for a particular video source.
turab is offline   Reply With Quote
Old 23rd January 2017, 15:25   #4  |  Link
mandarinka
Registered User
 
mandarinka's Avatar
 
Join Date: Jan 2007
Posts: 729
Quote:
Originally Posted by hajj_3 View Post
if they think that this is going to be widely used outside of broadcast then they are sadly mistaken. AV1 is going to take over the web. For their sake lets hope they create a single patent pool this time with sane licencing fees.
Well maybe AV1 surprises a lot (currently I wouldn't count on more than it gaining some limited edge on HEVC at best, but in the wild that will be negated by immature encoder). In any case I expect more from this, it will probably be ahead of AV1.

The question is how long will it take to standardize. After all the introduction will probably lag AV1 bistream freeze significantly (more so given how aggressive Google is about that, evidenced by those bugs in VP9). Maybe by 2 years, if not more?

The important thing is that the development continues on and the compression performance (quality) will again be able to progress. I don't care about ideologies and royalties, the technical aspect is important here. AV1 (or maybe better said, Daala) will bring that too unlike VP9 and previous alternative developments, but in many ways it goes in parallel to HEVC and only shares some of the goodies, so it is important the "MPEG" technology gets developed further. After all it has been MPEG/ITU technologies that have brought almost all the compression progress up till now - where would we be without AVC today? (And before that, without mpeg2).

Ideally we would see the best from the both camps combined but that is probably not possible. (AOM/Google side insists on royalty free model and probably won't allow stuff patented by them in the competing commercial standard, which would technically be possible).

Last edited by mandarinka; 23rd January 2017 at 15:39.
mandarinka is offline   Reply With Quote
Old 23rd January 2017, 23:01   #5  |  Link
nevcairiel
Registered Developer
 
Join Date: Mar 2010
Location: Hamburg/Germany
Posts: 10,342
Quote:
Originally Posted by turab View Post
Also, when there's a rate reduction of 0.1%, for example, how do we know it's significant? It could be a rate increase for a particular video source.
Stats like this are typically an average over several sources. Most tools are optional, so if on one particular source it doesn't work, either the encoder or the user can choose just not to use it.
__________________
LAV Filters - open source ffmpeg based media splitter and decoders
nevcairiel is online now   Reply With Quote
Old 6th February 2017, 14:21   #6  |  Link
iwod
Registered User
 
Join Date: Apr 2002
Posts: 756
I just want a decent image format to replace jpeg.
iwod is offline   Reply With Quote
Old 6th February 2017, 17:24   #7  |  Link
CruNcher
Registered User
 
CruNcher's Avatar
 
Join Date: Apr 2002
Location: Germany
Posts: 4,926
jep that war doesn't even really started yet, i mean it's partly already prepared but not yet really fought

but still JPEG does it's job and with the right PQ approach it still can be rather efficient

https://github.com/danielgtaylor/jpeg-archive
__________________
all my compares are riddles so please try to decipher them yourselves :)

It is about Time

Join the Revolution NOW before it is to Late !

http://forum.doom9.org/showthread.php?t=168004

Last edited by CruNcher; 6th February 2017 at 17:28.
CruNcher is offline   Reply With Quote
Old 7th February 2017, 03:26   #8  |  Link
iwod
Registered User
 
Join Date: Apr 2002
Posts: 756
Quote:
Originally Posted by CruNcher View Post
jep that war doesn't even really started yet, i mean it's partly already prepared but not yet really fought

but still JPEG does it's job and with the right PQ approach it still can be rather efficient

https://github.com/danielgtaylor/jpeg-archive
I would trade 10x Jpeg's computation power for 50% less file size at same quality.

We have had so much improvement in Video and not the same for images.

Anyway if HEVC is any indication, it is likely be the last popular MPEG codec. AV1 may not win, but hopefully AV2 + MKV.
iwod is offline   Reply With Quote
Old 7th February 2017, 09:34   #9  |  Link
Jamaika
Registered User
 
Join Date: Jul 2015
Posts: 697
The problem is that the encoders to the images of the latest generation JPEG are paid. I mean, I don't see modern sources on github. The advertised free software AdobeDNG is so old that it does not pay to take hands. I omit fact that difficulties in compiling various systems MAC, Windows and Linux. There is also no free software to convert JPEG(standard), TIFF, PNG to DNG, JPEG(DNG) or DNG <-> JPEG(DNG). Not to mention the such things as signal HDR. Container JPEG by every year arrives. I note also that the comparative tests of images are worth nothing. These free programs have half the features and capabilities than paid programs. Who needs pictures HEVC? I don't know, I think that the forum maniacs.
Jamaika is offline   Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 12:24.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.