CCE top field first bug. How to find the right field order and fix it

[Mozart]

Hi folks,

does anybody here remember these threads, talking about how to choose the field order?
thread1 (http://forum.doom9.org/showthread.php?threadid=9310)
thread2 (http://forum.doom9.org/showthread.php?threadid=8545)
thread3 (http://forum.doom9.org/showthread.php?threadid=8268)

Well, I had forgot this subject since that time. However, few days ago I did some backups with
noticeable choppy playback. And it doesn't matter if I use top field first on or off. Both are choppy.
There is no bug in DVD2SVCD. It is a CCE bug. Some people already know this. For me, CCE ever encodes "top field first",
even if it was turned off (I don't know where is the source of this bug. mb1 said that it is because of some codecs...).

So, what to do to avoid such choppy SVCD? It is simple, and very easy. You will need to use
Bitrate Viewer (http://www.tecoltd.com). It is freeware and you don't need to fill in any account.

rip a small chapter (warn: it cannot be the first chapter) using SmartRipper. After this, load the ripped vob with
BitrateViewer. It will show to you a lot of informations about the vob file. Veryfy if Field TopFirst: shows
"no" or "yes". If it said "yes", this means that you can use DVD2SVCD with no problem, because the CCE output will
have exactly the same field order of the source. If it said "no", the CCE output will have a wrong field order,
and you will need to fix it manually.

How to fix it:
1) encode the movie with DVD2SVCD and let it go to the end;
2) open the command line promp and go to pulldown.exe`s folder;
3) type the following:pulldown.exe "DRIVE:\encoded_movie_path\CCE_Encoded_Video_NTSC.mpv" "DRIVE:\encoded_movie_path\CCE_inversed.mpv" -nopulldown -tff even
4) rename the file from CCE_inversed.mpv to CCE_Encoded_Video_NTSC.mpv;
5) restart from the step which goes after video encoding, that can be pulldown or muxing and cutting.

Enjoy.

[Kedirekin]

@Mozart,

Great post. Always looking for tips to get the encode done right the first time.

One point - I believe the 'CCE always top-first' bug only occurs if you have progressive checked (CCE being smarter than is good for it). As long as you don't have it checked, I believe CCE honors your choice of field order.

Also, in step 5 -
that can be pulldown or muxing and cutting

Since it is an interlaced encode, there shouldn't be any pulldown in crash recovery.

[tai]

Mozart: I wonder how Birate viewer work for you but the bitrate viewer always tell me that dvd video is FieldTopFirst although I'm sure that it's not

[pacohaas]

@anyone who knows: is this a CCE 2.5 bug? still existing in 2.62? 2.64?

@Kedirekin: The Matrix(the movie, region 1) is an example of a film who's frame type is progressive and not top-field first.(according to bitrate viewer at least) It also has a nonlinear quantscale if that interests you as well.

also, in those past threads, it mentioned that the bug would be fixed in the next build(that was back at 1.0.3b5 I believe). basically what i'm asking is do I have a working process using CCE 2.6.2 and DVD2SVCD 1.0.7b7p1?

[pacohaas]

I just tested it out by reloading the ecl and looking at the settings and they were all as I had specified in DVD2SVCD. This is using the versions I last mentioned....hope this info helps somebody.

[Mozart]

@tai

there are two things to have in mind:

1) if the movie is "progressive" (BV will tell you if it is or if it isn't) there is no sence to talk about the field order, and BV usualy randowly shows "top first yes" or "top first no";

2) usually the first vob has a studio logo which sometimes has a different nature when compared to the whole movie (the logo can be interlaced and the movie progressive). It was because of this that I said "the analised vob cannot be the first".

Usualy I test all ripped vobs to verify if there is some kind of swap of order.