DivX 5.0.3


Recommended Posts

The first new DivX release in over eight months, built to integrate with the crop of DivX Certified hardware devices coming in 2003.

New:

* Support for interlaced video in the decoder. Preserving the interlaced fields of interlaced video may sometimes result in better video quality during playback.

* New de-ringing post-processing algorithm. The new algorithm improves the perceived video quality during playback. It is activated when the post-processing level is set to Level 6.

* Video Buffer Verifier (VBV) model (one and two pass encoding). The VBV ensures that the maximum peak bit rate never exceeds either the user's inputted "maximum peak" bit rate value or/and ensures that the encoded stream never violates the buffer of an MPEG-4 compliant decoder. This helps to prevent decoding failure in both hardware and software where memory may be limited yet compliant to the MPEG-4 standard as defined in ISO/IEC 14496-2:2001(E). In order for DivX video to be successfully delivered over a restricted channel in real time to a decoder, the encoder's rate control must ensure that the decoder's buffer is not violated. When this is done properly, then overflow and underflow will never occur and the encoder is said to be "VBV compliant." It makes no difference whether the video is encoded in 1-pass or many, in real-time or offline. It is the encoder's rate control that must ensure compliance. There are three main requirements of a VBV-compliant rate control algorithm, in order of importance:

1. VBV compliance. When VBV parameters are specified (size, initial occupancy, channel characteristics) the rate control should avoid producing video that is in danger of violating the decoder's buffer. Bitstream stuffing may be used in extreme cases to prevent overflow in true CBR systems. Underflow must be prevented by increasing frame quantiser or by dropping frames.

2. Target bitrate. We aim to meet the user's requirements as accurately as possible here. This is directly related to filesize.

3. Good-looking video. Rate control should aim to produce the subjectively best quality video given other constraints. Consistency is important here as changes in quality are very noticeable and video is often judged by the worse quality segment in the whole sequence.

* Nth Pass? encoding. Another new addition to the rate control arsenal, Nth Pass? encoding takes advantage of information available from previous analysis (or analyses) of the video sequence. Nth Pass encoding uses information garnered from previous passes through the video sequence, which can improve quality with each pass. There are four main steps to Nth Pass encoding:

1. Analysis (the first pass)

2. Map generation (where the results of analyses are processed)

3. Strategy generation (strategist examines map of the video sequence and creates a strategy that will best achieve the three VBV RC requirements)

4. In-loop nth pass rate control (uses information from the map and strategy to effect rate control that best fulfills the three requirements)

* Support for DivX Certified Profiles. With the new DivX Certification program, DivXNetworks is enabling third parties to create "DivX Certified" products that are rigorously tested and fully compatible with the entire suite of DivX? video technologies. There are four levels of official DivX Certified video products: Handheld Video Devices, Portable Video Devices, Home Theater Video Devices and High Definition Video Devices. These levels quickly and clearly communicate what type and size of DivX video are supported on the certified device. Now, when you encode video, you have the option to force your video to comply with one of the DivX Certified Profiles to ensure that your video will play back properly on certified products.

* DivX Decoder verification logo. There is a new semi-transparent DivX Logo that will appear by default for a few seconds any time you use the DivX decoder to watch a DivX video. This instantly lets you know you're watching a true DivX? video. You can turn the display of the logo on or off in the codec properties dialog.

* New internal application programming interface (API). The internal API of the codec was changed to better accommodate integration with third party partners. This only affects you if you are a DivX licensee and are working directly with the DivX object code. The external API (e.g., DirectShow, etc.) remains unchanged.

Improved:

* The motion estimation algorithm is better optimized for High Definition resolutions (up to 1080p, or 1920 x 1080 pixel resolution

* The pre-processing performance is now improved with with the IYUY 4:2:0 color space

Changed:

* Removed MP4creator and the MPEG-4 file output option due to some compliance issues. This will return in our next release once these issues are resolved.

* Removed the Intelligent IVTC functionality due to persistent problems caused by lack of variable frame rate file format support. This technology will be back soon in a new form...

* Changed the block skipping threshold for high quantizers, which improves visual quality when high quantizers are used.

Fixed:

* Fixed a bunch of little cosmetic bugs in the DirectShow decoder properties page.

* Fixed a bug where the last P frame before an I frame was not displayed when Smooth Playback was selected.

* Fixed a practically unnoticeable YUV->RGB16 color conversion rounding bug.

* Fixed a bug that was the cause of some of VirtualDub's "Cannot Start Video Compression" errors.

* Fixed a few issues with DivX 3.11 compatibility.

* Modified the decoder to deal with a bug in the old OpenDivX encoder that set the scalability bit incorrectly in certain cases. The DivX decoder can now detect this problem and play the content properly.

* Fixed a problem where some rounding operations were not done toward zero, as is required by section 7.4 of the spec. This fixed an intra case in MPEG-2 inverse quantization.

* Fixed a bug where some frames would be corrupted by an out-of-range error.

* Fixed several rounding issues related to MMX/SSE/SSE2 optimizations.

* Fixed an MMX bug in RGB32 color conversion.

http://www.divx.com/divx/divx_win_versions.php

Download: http://www.divx.com/divx/

Link to comment
Share on other sites

Cool! I don't understand why they only go up in minor increments for versions when these seem like larger releases to me....

I think they just want to be like Apple. ;) How 10.2 was a huge upgrade, but was only a .1 release.

Link to comment
Share on other sites

sheesh. only inferior mpeg1/2 codecs use multipass encoding.

divx3.11+gknot does perfectly fine with 2-pass.

and now they ? that ter:rolleyes:s:

:no:o:

2-pass is multi-pass:blink:k:

And if you think divx 3.11 is better than 5, you need some glasse:D:D They're just giving you the option of doing even more passes for even higher quality. The best of the bes:):)

Link to comment
Share on other sites

Yesterday after I started this thread there were no new replies in active topics for about 20 mins. I checked whos online, It showed only 2-3 users. hehe it was the sql worm.

So is this just an update to the player? Or is the codec updated to?

Yea... man the codec is updated.

Link to comment
Share on other sites

2-pass is multi-pass? :blink:

And if you think divx 3.11 is better than 5, you need some glasses :D They're just giving you the option of doing even more passes for even higher quality. The best of the best :)

yes 2-pass is multi-pass. but it's only 2pass

what i think is better is not for this thread, but it's common knowledge that divx3.11, 5 and xvid produce very close images and which one is better is a matter of personal preference.

and if you are going to agrue with me on this make sure you have some proof to back up your claim

Link to comment
Share on other sites

matrix-dx5-147712.jpg

http://www.doom9.org/images/codecs/matrix-...-dx5-147712.jpg

matrix-sbc-f07557.jpg

http://www.doom9.org/images/codecs/matrix-...-sbc-f07557.jpg

I included links incase doom9 doesn't do offsite linking... ;)

As you can see, SBC (which is pretty much the best you can get out of divx3) is much more blocky than divx5. Just look at the bad officer dude's face. Looks very blocky, not as clear. :) The dust also gets the block treatment in divx3/

Proof enough? :D Well, I think it looks better, at least :)

Link to comment
Share on other sites

timdorr, how convenient. you picked 2 shots where divx5 is clearly better. well, me quote you another part from doom9's article, specificly from conclusion http://www.doom9.org/codec3-4.htm

SBC has yet to be clearly beaten. From a crispness point of view it is still number one, however, you pay for that crispness with blocks. It is also still the hardest method to properly comprehend if you want to go beyond GKnot. However, if you've mastered Nandub I wouldn't yet throw all my knowledge overboard.

...

To summarize, DivX5 is still the solution for people looking for a quick 2-pass solution without too much trouble and who don't want to tweak their encoding sessions endlessly. SBC still remains the solution for the cautious experienced encoder who isn't yet willing to try something new. The something new would still be XviD though the 0727 build can clearly not be recommended. But XviD has had problems that eventually got fixed in the past so I wouldn't dump the codec altogether.. just watch out for an announcement that the problems discovered have been fixed unless you're willing to take the risk of ending up with a really poopty result. RV9 is the new codec on the block and it seems it's here to stay. If you have time why not try it on a couple of movies and see what you find out and if the codec can keep delivering if you use other genres where a lack of detail on backgrounds matters more.

like i said before, divx3, 5 or xvid is a matter of personal preference.

Edited by MxxCon
Link to comment
Share on other sites

To be fair, that article is over a year old you guys! 3.11 isn't even really MPEG4 and it won't have ANY support in hardware players. If you encode with 4.xx or 5.xx then you will be able to play it back on future hardware (and a few things that just came out). Try using 3.11 at lower bitrates and you'll see major blocking, use XviD or DivX 5.03 and you'll just have slightly less detail for a while before it ever starts to block up.

Link to comment
Share on other sites

excuse me but it is NOT over a year old.

that article was updated on 08/09/02 shortly after divx5.02 was released.

it's obvious if you use lower bitrate then you should apropriatly decrease resolution.

Link to comment
Share on other sites

3.11 is so mpeg-4, it's just a hacked version of the ones you find with Windows. It gives you more control over what you can do with the settings. SBC is still one of the best. Divx 5.0x hasn't pulled that much ahead of it, if at all. SBC preserves more detail while divx 5.0x smooths.

Link to comment
Share on other sites

Okay, fine, it's HALF a year old! 3.11 isn't MPEG4, MPEG4 didn't even really exist at the time. Basicly 3.11 was a hack of WMV 7 which was based off the early MPEG4 drafts but had some modifications done to it. Altogether it is vary different then any of the current MPEG4 ISO profiles (some of which are still being worked on). The point I was trying to make with using low bitrates is that it shows how the codec performs overall. In general, codecs are supposed to do the best they can to keep you from noticing artifacts. 3.11 NEVER tries to lower the detail in order to prevent macroblocks which is a major error on their part. I've done some testing and can say that 5.03 still isn't as good as XviD but it IS still quite better then 3.11. Btw, any time you're comparing codecs, be sure to leave post-processing off so that you can see how good the encoder actually is instead of seeing how well the decoder can BLUR the crap that the encoder made (in the case of 3.11).

Edited by pHuzi0n
Link to comment
Share on other sites

ffdshow is based off of XviD but with LOTS of additions to allow it to play back all sorts of formats. It is definately the best single decoder available and is quite faster then DivX's decoder.

Link to comment
Share on other sites

divx3.11 is as mpeg4 as you gonna get.

it does not produce mpeg4 compliant streams, but it is mpeg4 codec anyway.

low bit rate DOES NOT show how codec perform overall. it shows how codec perform at low bit rates.

your not going to watch movie that's encoded at 100kbit. so why test it at such insane bitrate? it doesn't mean anything usefull.

and when your comparing codecs TURN ON filtering. becuase once again, this is how you watch the movie. encoder and decoder compliment each other. they are designed for each other. encoder tries to make best possible picture, decoder fixes any remaining errors.

ffdshow is not based on XviD. it's based on libavcodec from ffmpeg

Link to comment
Share on other sites

Lmao, you're saying that it's MPEG4 even though it doesn't produce compliant streams! Let me clue you in, non-compliant = non-MPEG4. I'm not talking about super low bitrates, if the bitrate gets too low every codec will look like $hit. I'm saying that you should take whatever bitrate you normally use and then lower it slightly to see the first signs of loss. Any decent codec will try to lower the detail (by using a lower quantizer) long before it starts spewing out blocks. Not everyone uses filtering ya know. All it does is BLUR the image and for some people, bluring is bad quality (ie. they like their picture to be sharp). Besides that, hardware decoders DON'T post-process. And last off, here's a quote from the about page in ffdshow.

1. About ffdshow

ffdshow is a DirectShow decoding filter for decompressing DIVX movies.

It uses libavcodec from ffmpeg project or for video decompression (it can use

xvid.dll installed with xvid codec too), postprocessing code from mplayer to

enhance visual quality of low bitrate movies, and is based on original

DirectShow filter from XviD, which is GPL'ed educational implementation

of MPEG4 encoder.

Link to comment
Share on other sites

Yeah. I've used DivX 3.11, with SBC - and it still looked like crap.

The movie was LOTR - FOTR, and I tried to squeeze it into 1CD.

It worked okay with DivX 5.02, DivX 3.11 with SBC gave really bad

artifacts.

It was better than perhaps DivX 4, but when DivX 5.02 was released,

and Xvid came, things improved a lot. And just to add more to what

pHuzi0n said, you can't compare them with filters on.

I have friends who play back with mplayer in linux, some use ffdshow,

some uses DivX 5.02, and now some 5.03. The thing you want to see

is how well it is encoded, not decoded.

And also low-bitrate-tests are used all over. That's the way to show

how well Ogg Vorbis performs against MP3. You need to provoke the

encoder into showing artifacts, in order to make it easier to judge.

Does anyone know how to use Ogg Vorbis encoded audio streams in

DivX's?

Link to comment
Share on other sites

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.