PDA

View Full Version : vop packed to non video packed?


johnhamler1
29th June 2006, 03:13
how do you change a xvid encoded with vop packed to non vop??without restarting the whole process again.

unskinnyboy
29th June 2006, 04:11
Err, "non vop"? "vop"? Sounds like you are speaking about N-VOPs characteristic of packed bitstream? If so, you can unpack them using MPEG4 Modifier (http://forum.doom9.org/showthread.php?t=78050). But not the other way around (means you cannot enable packed bitstream on a file encoded without it, unless you reencode).

johnhamler1
29th June 2006, 07:28
thanks a lot, I guess this is what i needed...

johnhamler1
29th June 2006, 21:09
How do you install it,does it work with another software???

unskinnyboy
29th June 2006, 21:29
There is nothing to install. It is a standalone executable file. Just open the avi in it, check Unpack and save the avi with another name. I didn't get the 2nd part of your question.

Cirion81
18th July 2006, 11:06
Err, "non vop"? "vop"? Sounds like you are speaking about N-VOPs characteristic of packed bitstream? If so, you can unpack them using MPEG4 Modifier (http://forum.doom9.org/showthread.php?t=78050). But not the other way around (means you cannot enable packed bitstream on a file encoded without it, unless you reencode).

Hello unskinnyboy,

why is it not possible to convert an unpacked bitstream to a packed bitstream? As far as I gathered from this forums XviD FAQ the difference between an unpacked and packed stream seems to be the order of I,P and B frames.

I am asking, because I own a standalone player with the Sigma-Designs-Chip EM8620L. The problem with this chip is, that videos containing 2 or more consecutive B frames lead to a jerky playback if the bitstream is unpacked. A packed bitstream though is handled smoothly.
So if there would exist a tool similar to MPEG4Modifier which could convert a unpacked to a packed bistream, I would be more than happy ...

Greetings and thank you for your answer!

Moitah
18th July 2006, 21:40
It is certainly possible, but MPEG4 Modifier's AVI reading/writing code doesn't allow for it to be done easily, which is why I haven't added it.

Cirion81
19th July 2006, 09:07
Hello Moitah,

any chance of this feature beeing implemented?
I don't know how widespread the EM8620L is (mine is in the Pinnacle Showcenter 200), but I know that the Buffalo Link Theater and some Kiss player utilize this chip as well.

Greetings

celtic_druid
19th July 2006, 10:02
Seems very odd. DivX doesn't use packed bitstream for > 1 bframe does it? If so do they also playback jerky? But you are saying that 1 bframe, not packed is ok?

Cirion81
19th July 2006, 10:36
Seems very odd. DivX doesn't use packed bitstream for > 1 bframe does it? If so do they also playback jerky? But you are saying that 1 bframe, not packed is ok?
Well, I must admit, I am not very much into the details of DivX, since I don't use it (therefore I can't tell you if there are any problems with DivX).
All I can say is, that XviD videos with 2 consecutive b frames and no packed stream play jerky. I will have to do some reencoding to test with 1 b frame, unpacked stream - or are there any test data files out there to download?

god2003
19th July 2006, 10:47
There is a french Test CD. http://divxtest.isstar.net/DivXTestCDv2.0.by.bbdoc_forum.surdvd.com.iso

you can get other mirrors to download (and bittorrent emule) here:
http://forum.surdvd.com/viewtopic.php?t=42412

It's in french but i'm sure you will figure out. The links are visible.

There is a translated page http://www.divxtest.com/DivxTest-in-English.html

Here is a list of the tests categories :
- A_ Medias and Disc type
- B_ Options / general Specifications
- C_ GUI : Menus options
- D_ GUI : Audio and image files options
- E_Connections
- F_Compatibility : Supported image types
- G_Compatibility : Supported audio types
- H_Compatibility : Supported video codecs
- I_Compatibility : Video containers and “bivx” tests
- J_Compatibility : separated subtitle files tests (for *.avi containers)
- K_Others unreferenced (optional tests)

A detailed manual is available on the CD itself.

foxyshadis
19th July 2006, 11:48
I'd ask for creating packed bitstream just to make editing files less painful (I love pulling favorite scenes out of stuff I watch, and I absolutely can't stand avidemux's terrible interface), except that vdub ignores packed bitstream now anyway, somehow. I don't know if it's ffdshow's fault or vdub's, but packed stuff gets treated almost the same as unpacked stuff now, and it's incredibly annoying.

You can use avidemux to create packed bitstream files out of unpacked, if using it doesn't drive you crazy.

Cirion81
19th July 2006, 16:05
@good2003: thank you for your tip.

@foxyshadis: this sounds nice. I had a good look through the documentation of avidemux, but did not find any reference how to pack an unpacked stream without recompressing the whole video - which I seriously would like to do due to quality reasons and a very slow PC ...
So maybe if you could give me a hint how to create packed bitstream files out of unpacked ... Thank you!

celtic_druid
19th July 2006, 16:21
AVIDemux can unpack. As far as I know it can't pack. Why should it anyway? It has no problem with bframes that aren't packed.

foxyshadis
19th July 2006, 16:34
Hmm, you're right, I don't know why I thought it did, I must have misread something. Seems odd that there are so many tools that can unpack and none that can pack again.

celtic_druid
19th July 2006, 17:07
Not when you think about it. Packed bitstream is basically wrong, so removing it makes sense. Plus there was the whole issue with MTK based SAP's not liking packed bitstream with > 1 bframes.

foxyshadis
19th July 2006, 17:56
I guess I'm approaching it from too much of a virtualdub/premier standpoint; if they had mp4 input capability or dealt with mpeg-4 properly (ala avidemux) I wouldn't care at all.

mod
19th July 2006, 17:56
The issues come with pbs, so it makes sense to remove, not to put..

Cirion81
19th July 2006, 21:10
@mod: as you can see, in the case of the Sigma-Design EM8620L issues come without packed bitstreams ...

I guess thats part of the general idea to make players incompatible with each others. Some manage GMC, some don't, some dispise b frames, some need them in packed/unpacked format. *sigh*

Bottom line I think is: no matter if packed bitstreams are making sense or not - there are standalone players out there which handle one or the other type better.

mod
19th July 2006, 21:40
as you can see, in the case of the Sigma-Design EM8620L issues come without packed bitstreams ...
Ahh.. yes there're always new interesting issues every day..

celtic_druid
20th July 2006, 04:08
MTK's issues with pb were fixed via a firmware. Make more sense for Sigma to fix it rather than everyone go around packing files.

It was an issue with MTK due to the popularity of such players (number 2 behind ESS I guess). First I have heard of the Sigma-Design EM8620L issue, so I guess they are not that common.

Cirion81
20th July 2006, 11:26
Hello again,

well I did some test with packed and unpacked bitstreams with XviD. here are the results

Packed bitstream:

1 consecutive b frame: smooth
2 consecutive b frames: smooth
3 consecutive b frames: smooth

Unpacked bitstream:

1 consecutive b frame: smooth
2 consecutive b frames: jerky
3 consecutive b frames: jerky

What I noticed immediately is, that with the test video I used (slow pan over a panorama), there were only about 3 noticable jerks in 40 seconds - way less than I usually get. I had a look at the file with MPEG4Modifier which gave me the following information:

Max consecutive B-VOPs: 2
1 consec: 0,90%
2 consec: 99,10%

I looked at the stream where the jerkiness occured, and found the following: while in the smooth playig part the stream was ...PBBPBBPBB... the jerks occured around ...PBBPBPIBBPBB...

So possibly either the changing of 2 concecutive B frames to one and back again or the I frame somehow irritates the decoder chip.

Since, as far as I learned the past days, the difference between packed and unpacked bitstream is the way the B frames are stored, the changing of the B frames rate seems to me to be the source of the jerkiness.

This I would also base on the fact, that a "normal" video contains a much higher rate of B frame rate changes than my test video. For example some information gathered from a video containing more action than panning, and playing much jerkier than the test video:

Max consecutive B-VOPs: 2
1 consec: 34,47%
2 consec: 65,53%

What remains to check is, if this problem is a firmware problem of the Pinnacle Showcenter 200, or a general problem of the Sigma EM8620L. Maybe someone owning a device with this chip could give me some feedback on that.

Moitah
5th August 2006, 05:00
MPEG4 Modifier is now able to pack.

Cirion81
6th August 2006, 23:48
WOW! thanks a million Moitah (also for the extra email) :thanks:

I will test the new version as soon as possible.

Cirion81
10th August 2006, 19:10
Hello Moitah, hello everyone

I ran all my test-cases again - which means I did "repack" those files with unpacked bitstreams using MPEG4 Modifier 1.4.1. Now they are playing as smooth as they are supposed to be. Thanks again Moitah!!

Pinnacle tells me (at least someone from the support front does) they want to fix this problem possibly with the next firmware update (there hasn't been an update since december 2005). Of course they did not specifically state, that this is indeed a problem of their firmware, and neither did they reveal if/when a new firmware is planned to be released.

So, they promised to solve this problem possibly ... someday ...

Moitah on the other hand developed a resolution within two weeks. I wanted to buy an AC3 update for my Showcenter - but instead I'll send the money to Moitah's paypal account.

Ah, and before I forget Moitah - if it is ok with you I'll notice Pinnacle of your "workaround", so they can add it to their FAQ.

-Christian

Moitah
10th August 2006, 19:26
Ah, and before I forget Moitah - if it is ok with you I'll notice Pinnacle of your "workaround", so they can add it to their FAQ.
Sure, that's fine.

And thank you for the donation :).