Log in

View Full Version : Is Xvid DEAD?


Pages : 1 [2] 3

Didée
18th November 2005, 13:11
sysKin: this (http://forum.doom9.org/showthread.php?t=86180&page=2) is an issue that needs to be fixed.
No, it is not. (answer (http://forum.doom9.org/showthread.php?p=739530#post739530))

raeltheimperialaerosolkid
18th November 2005, 13:51
Probably covered on the previous page by "see what other things can be fixed. DX50 userdata mess comes to my mind"

Well...maybe.
In the meantime I'll stay on the hardrock stable 1.0.3....:D

sysKin
18th November 2005, 15:43
Probably covered on the previous page by "see what other things can be fixed. DX50 userdata mess comes to my mind"
No, that's not related to DivX profiles.

Weren't we going to remove them completely?

Anyway, the answer is BOOOO----RIIIINGGG---- :P

celtic_druid
18th November 2005, 15:53
Weren't their some claims that when the DivX user data string was changed that it broke compatibiltiy with some standalones? DivX503b1393p instead of DivX999b000p.

In the current CVS the DXN profiles were I think renamed. DXN HT is just Home Theatre, but I never did a compile.

bond
18th November 2005, 16:10
In the current CVS the DXN profiles were I think renamed. DXN HT is just Home Theatre, but I never did a compile.why?

i think dxn profiles are good to have,cause it shows how xvid rules dxn :D

celtic_druid
18th November 2005, 17:20
Profiles are still there, just renamed:
Home Theatre NTSC, Home Theatre PAL, Cinema Plus NTSC, Cinema Plus PAL

Home Theatre = DXN Home Theater
Cinema Plus = ? Looks a bit like my MTK profiles. For instance qpel can be enabled.

So it still shows how XviD rules DXN and how they can spell theatre.

stephanV
18th November 2005, 17:38
uhm, both theater and theatre are correct :p

suxen_drol
19th November 2005, 03:10
uhm, both theater and theatre are correct :p

i tend to disagree!
perhaps we need i18n for british and american english.

Elias
20th November 2005, 15:12
XviD is not dead. However, progress is very slow; the 1.1 stable version was going to be released at least a month ago.

Anonymouses
20th November 2005, 16:47
i tend to disagree!
perhaps we need i18n for british and american english.

From the Cambridge dictionary:

Definition
theatre (BUILDING) Show phonetics
noun [C]
1 UK (US theater) a building, room or outside structure with rows of seats, each row usually higher than the one in front, from which people can watch a performance or other activity:
the Lyceum Theatre
a lecture theatre

Both are correct. :p

708145
20th November 2005, 18:07
I wouldn't say you're the only one, the ability to have AQ is there for a good reason I guess. But AQ is messy, takes bits, limits possible macroblock modes and can only be made on macroblocks as a whole. Besides, ASP's quantization - being linear - has huge steps on higher qualities (quants 1-2-3-4). That's unlike AVC.

Lambda, with its ability to save on motion vectors (VHQ) and drop coefficients (trellis) should be able to modulate quality quite a bit.
How can I know that :0 ;)

what about combining lambda with the switch to another quant range?
using a downscaled CQM one could shift the q2-q4 range to q10-q20 which results in much smaller steps.

is this a stupid idea?

bis besser,
T0B1A5

Kopernikus
20th November 2005, 20:12
Thats like Didèes 6of9 Matrix works.

But AQ in ASP is very limited. The difference in quantizer of two Macroblocks can differ only by 2, so if you decrease the stepsize, you cannot modulate the quality enough.

And the dquant field (3bits per MB allowing -2:+2 change) is not entropy coded, so there is a overhead of 600 byte per frame (720x576) with AQ. AQ also blocks the 4mv mode.

In AVC AQ is not so limited.

Manao
20th November 2005, 22:05
And the dquant field (3bits per MB allowing -2:+2 change) is not entropy codedIf dquant == 0, it takes no bits at all ( it's kinda written alongside the mb type ) so it's entropy coded.

It blocks the 4mv mode only if dquant != 0. Which means it's tricky to decide whether to change a quantizer or not ( because the next block might need to be coded at a different quant with 4 mvs ).

Varying lambda is a better way to do AQ ( at least, easier ), though it might not allow that must variation either. And it might also become quite messy when it comes to changing the lambda of the trellis.

DeathTheSheep
24th November 2005, 18:19
As of yesterday, work has been started on the XviD 1.2 CVS.

Whaddya think? AQ gunna be up and running in 1.1 or 1.2?

Cheers

Elias
24th November 2005, 19:09
As of yesterday, work has been started on the XviD 1.2 CVS.

Whaddya think? AQ gunna be up and running in 1.1 or 1.2?

CheersI think it's awesome news. Since the work on the 1.2 CVS has been started, will we see the 1.1 Stable release anytime soon?

len0x
4th December 2005, 17:31
I take it 1.1 is not going to be released at all and things move on to 1.2 ?

bond
4th December 2005, 18:06
I take it 1.1 is not going to be released at all and things move on to 1.2 ?first of all someone has to start coding ;)
syskin? :D

len0x
4th December 2005, 22:13
well, I'd say since 1.1 release branch was created almost two weeks ago much coding is not expected :) So shall we just ask C_D to build from that and call it 1.1 final? :D

SeeMoreDigital
4th December 2005, 22:24
well, I'd say since 1.1 release branch was created almost two weeks ago much coding is not expected :) So shall we just ask C_D to build from that and call it 1.1 final? :DJoking apart.... I say yes, why not?!

celtic_druid
5th December 2005, 00:30
I could rename it, but I wouldn't feel comfortable so you will have to use your imagination with the last 1.1.x build I did.

raeltheimperialaerosolkid
5th December 2005, 11:17
The topic that issued the problem has gone forward :)

here (http://forum.doom9.org/showthread.php?p=737709#post737709)

It says that the "packed bitstream" in the DXN profiles wasn't disabled even if the user unchecked it manually. I don't know if it's a problem of some certain build or if it's a source problem...

My 2 cents...

;)

No, that's not related to DivX profiles.

Weren't we going to remove them completely?

Anyway, the answer is BOOOO----RIIIINGGG---- :P

johnsonlam
6th December 2005, 13:35
XviD is not dead. However, progress is very slow; the 1.1 stable version was going to be released at least a month ago.

Progress very slow is NOT a "serious" problem, I can wait. I know not much developers now, but IMO they should send some news to Doom9's Admin.

There're too many news around, I don't have time to surf every homepage, for video and audio news Doom9 is my best portal thanks :thanks:

Sometimes I did think being a bit commercial can push the software development faster, like DivX.

loni_blues
8th December 2005, 03:37
celtic_druid,

Just a little doubt:
xvid.cvs.release-1_1-branch.2005.11.24.7z and XviD.1.1.cvs.exe would be both the same and "final" 1.1 release?
Thanks a lot,
loni_blues

celtic_druid
8th December 2005, 04:47
Same files in both yes. Final? Not sure.

tommy_vercetti
18th December 2005, 19:44
Answer to the original question, YES

boombastic
18th December 2005, 20:21
why you say that?you can't say it without explain!

nightrhyme
18th December 2005, 20:42
Any news from Syskin ?

sysKin
19th December 2005, 02:54
Any news from Syskin ?
Like this? http://forum.doom9.org/showthread.php?t=104257

gatekeeper_007
19th December 2005, 10:33
This reminds me of some other songs -
XviD is still alive, Xvid is havin' sex... (do u remember?) :p
The truth is that XviD has made for some time now a lot of people talkin' about it all over the world, and they all say one : that codec rocks.
Now the right question should be : will XviD ever die? (this will be a new movie "Hard to die IV"
I can say that I've read all the previous replys... but my opinion is that u didn't shearch enought before postin' - and that u should be more patient.
"we can't always see the stars but they are always there"
:thanks: to Koepi for this great codec.
respect.
I can't wait for a new release :cool:

Slitheen
22nd December 2005, 22:46
I do think that Xvid is going to be the one that gets left out in the not too distant future. With Divx and Nero Digital AVC codecs being the only ones still standing.

For this reason I've stopped using Xvid.

Elias
22nd December 2005, 23:21
I do think that Xvid is going to be the one that gets left out in the not too distant future. With Divx and Nero Digital AVC codecs being the only ones still standing.

For this reason I've stopped using Xvid.Why? If anything, XviD is still extremely popular. DivX is only commercially more popular, XviD has a larger fanbase. As for MPEG-4 AVC, x264 will probably beat Nero's lead sooner or later in popularity and quality.

Slitheen
22nd December 2005, 23:27
For me, it's a question of standalone support, and support that will be still there a few years from now.

Elias
22nd December 2005, 23:30
For me, it's a question of standalone support, and support that will be still there a few years from now.You don't get it... MPEG-4 in the mp4 container, it's all the same :) Really, there's no real difference (because mp4 doesn't store fourCC in the same way as avi). Just encode XviD right and put it into mp4 and it will work just as good as any Nero Digital ASP in mp4 on any Nero Digital certified DVD player. Same thing with x264.

Slitheen
22nd December 2005, 23:45
I'm sure you're right. But I'm the voice of the public remember. The only thing I understood from your last post was the MPEG4 bit ;)

Elias
23rd December 2005, 00:04
What I meant was, it doesn't matter what MPEG-4 codec you use, as long as they're in the MPEG-4 container (mp4), they all work equally good everywhere. Unless of course you use the unrestricted profile features (like e.g. lossless AVC with x264), which far from all standalone DVD players support. x264 is just as much AVC as Nero Digital AVC is H.264 and same with XviD; it's not any less MPEG-4 Part 2 than DivX.

Elias
23rd December 2005, 00:09
Progress very slow is NOT a "serious" problem, I can wait. I know not much developers now, but IMO they should send some news to Doom9's Admin.

There're too many news around, I don't have time to surf every homepage, for video and audio news Doom9 is my best portal thanks :thanks:

Sometimes I did think being a bit commercial can push the software development faster, like DivX.Did I imply that it was a serious problem? I'm using the XviD 1.2 CVS and it works great. So I don't see any real problem at all.

ChronoReverse
23rd December 2005, 15:26
Well, the cat's out of the bag now. XviD AVC!

Teegedeck
23rd December 2005, 18:45
Thanks for liking XviD but XviD is a team effort; please spare some nice thoughts for these core developers also:

Christoph Lampert
Pascal Massimino
Michael Militzer
Peter Ross
Radek Czyz

et.al.

bond
24th December 2005, 01:12
Well, the cat's out of the bag now. XviD AVC!xvid avc is not the official name till now, and of course its a totally different codec than xvid

Kostarum Rex Persia
24th December 2005, 01:46
But, why creating Xvid avc codec? What is purpose of that? We already use x264 freeware solution.

I think there is no need for xvid avc codec.

mod
24th December 2005, 01:49
But, why creating Xvid avc codec? What is purpose of that? We already use x264 freeware solution.

I think there is no need for xvid avc codec.
Where did I hear similar words? ..7zip.. :D

charleski
24th December 2005, 02:14
With Divx and Nero Digital AVC codecs being the only ones still standing.
Well for that to happen Ateme will need to finalise their current High Profile encoder and Nero will need to buy it! (And when you do buy it, give it a bit of advertising fanfare FFS, a lot of people are waiting to see what this encoder can do for themselves.)

Selur
24th December 2005, 07:45
But, why creating Xvid avc codec? What is purpose of that?
learning how avc really works,...

skal
24th December 2005, 08:02
xvid avc is not the official name till now, and of course its a totally different codec than xvid

?!

Bond, this is the second time i "?!" you today. Please, just refrain posting such assertive
*opinions*.

-Skal

sysKin
24th December 2005, 09:13
?!

Bond, this is the second time i "?!" you today. Please, just refrain posting such assertive
*opinions*.

Skal, if you guys are so secretive about it, and give no kind of informative announcement, "opinions" is all we've got left.

Bond was referring to information I gave him, which was the best information I had. If it was one year old and no longer true, you can't blame us for not guessing correctly.

[edit]Perhaps I should STFU about it, but I'd like everyone to know that I have some mixed feelings about the revelations - if a codec is referred to as "from XviD team", I have a weird feeling that people like me should not be finding out about it this way (this way == from rants that bond said something wrong because he was repeating after me).

bond
24th December 2005, 11:08
thanks for clarifying this syskin :)

Doom9
24th December 2005, 13:38
Would it make sense to give me a name for the comparison, the first time the codec was ever mentioned to the general public, then go back and change the name again? Imagine the confusion..

And to everyone: I'm beginning to suspect our Persian king is related to roccosi... remember him?

bond
24th December 2005, 13:39
i find "xavc" to be good :)

sysKin
24th December 2005, 14:22
Nah XviD2 is a great name, if the code goes opensource. I don't expect any confusion (stuff with "2" in their names are usually different).

bond
24th December 2005, 14:29
the "2" will not tell newbies that their xvid hardware player wont play their xvid "2" files ;)