IFO-Edit 0.95 stripping chapters ?

[candyman_eu]

Hi,

I just wanted to author a DVD with the new IFO-Edit .95. My original IFO has 28 chapters and a duration of 2:02:15.
However when I multiplex everything together, the resulting IFO has only 23 chapters and a duration of 1:46:20.
So I'm a bit wondering what's up here. I also authored another movie (PAL) and here it works. I just checked a second NTSC one and it's also 20min shorter ... created the MPEG2 in CCE ... then used Pulldown and tried to author ... any clues ???

Thanks ...
Candyman
:rolleyes:

[jdobbs]

I'm getting the same results. My original IFO had 32 chapters. But after I author I get only 26 chapters/cell points. If I examine the "CELLTIMES.TXT" file it shows 32 entries. But I think it may incorrectly add an additional chapter to the end. In reality (I think) the first cell/chapter point should count as 0 so you would only need 31 cell start points in the file??? It appears that the last 6 or 7 cells weren't reproduced. (The entire movie is present -- it just seems to be the chapter points that are missing).

The reason I comment on this is: If this part gets exact, the entire DVD-9 to DVD-5 conversion could finally be completed with absolutely no expensive authoring software...

GREAT WORK DERROW!!!

Since this IFO has a one-to-one chapter/cell ratio I can use Chapter Extractor to show you (ignore the 1-2 frame differences -- from examining the IFO I think IFOEDIT is the one that is right and XTRACTOR is adjusting for 29.97):

Here is the output on the original using Chapter Xtractor:
0
6859
11092
15878
19808
22294
29175
39177
44318
47312
51159
61776
66600
70076
75972
80268
83965
86289
93122
99633
102402
107499
114706
124117
129301
132750
138366
143088
148626
153336
157232
160878

Here is what is in the CELLTIMES.TXT file:
6859
11092
15878
19808
22294
29175
39176
44317
47311
51159
61775
66599
70075
75971
80267
83965
86289
93121
99632
102402
107499
114706
124117
129301
132750
138366
143087
148626
153335
157231
160877
168261

Here is what I get as a final output IFO from IFOEDIT:
0
6856
11081
15869
19792
22273
29154
39141
44281
47266
51115
61715
66541
70005
75901
80195
83887
86213
93031
99536
102305
107392
114592
123998
129181
132622

[candyman_eu]

I also checked with PowerDVD the first and last VOB and indeed the movie seems to be completely present ... but why does IFO-Edit 1) stip several chapters and 2) even if some chapters are stipped, the length of the movie should not be altered or ?

@Derrow - Could you try to repro/solve this ? It seems to be only on NTSC movies ...

TIA! ;)

[jdobbs]

I'm guessing this is related to the difference between true framecount and the framecount resulting from RFF... The 26th chapter seems to actually be at the point of the credits in the file... The M2V I am using is native FILM.

[candyman_eu]

The M2V I'm using is the output from CCE and then running 'Pulldown input.m2v output.m2v -prog_frames p -drop_frame true' against it. Maybe it has something to do with the pulldown we use ?

[jdobbs]

No. The one I'm using wasn't changed nor did it have pulldown run against it. I just demuxed it from the original .VOB and reauthored (it was small enough to fit on a DVD-5). I was just trying out IFOEDIT.

[jdobbs]

Bump

[oggiedoggie]

jdobbs, what did you mean when you guessed that "RFF" had something to do with the offset?

I am seeing the same problem and am also eagerly awaiting a fix.

(later)

Oops, never mind, "repeat first field" flag, found it by searching the forums.

[jdobbs]

I'm only guessing. But I know the example I used was a 23.976 FILM based movie that runs on NTSC. It's possible Derrow is counting frames when he builds the VTS but doesn't adjust for the frames created by RFF... It makes complete sense looking at the numbers in the example above. 23.976/29.97 = .8 if you multiply the last chapter in the original IFO by .8 you get 134608 as the title length (in FILM frames). Notice that the last chapter that wasn't truncated was also the last one before you hit the end of the title in FILM frame counts...

[jdobbs]

@oggiedoggie

What I would suggest as an interim fix (only for FILM based sources) is to modify the CELLTIMES.TXT file manually with notepad or wordpad. Just change each value by a multiple of .8, and resave the file. That should put the cell times in correctly.

Unfortuately that won't be a final fix unless there is a one-to-one relationship between the cells and chapters. But it gets you closer.

[jdobbs]

BTW. I will add this as an option to the "Write Chapter Files" in IFOUpdate. That way you can produce the exact chapter format as the original. That won't help those who don't have the original IFO -- in that case you will have to generate it using VOBEdit. I would think that is why Cell points are used instead Chapter points to begin with (chapter points aren't represented in the VOB except as the cell that holds it's start).

[quantum]

Adjusting the celltimes is not going to fix the problem with the movie duration, which to me points to an underlying problem. I just did a side by side comparison with an NTSC movie I CCE'd then authored in both DVDMaestro and Ifoedit.

The Ifoedit version didn't play correctly if I ran a pulldown after CCE (it skipped and paused every few seconds). It played correctly if I didn't run a pulldown, which seems odd to me. DVDMaestro wants the pulldown version.

The Ifoedit movie shows 1hr23mins. DVDMaestro version is 1hr44mins. Chapters are messed up in Ifoedit version also.

[jdobbs]

Interesting. It seems to point to the same problem. If you multiply 1hr44mins times .8 it comes to 1hr23mins...

[Sy]

I've been doing the .8 multiply on the cell times for a while now... it works perfectly.. the new chapter stops come within a frame or 2 of the original. an easy way to do em is in excel. open the celltimes.txt file and copy all the #'s them open excel and paste it and in the colum next to it inster the formula "b1=a1*.8" (assuming the original celltimes are pasted in the a column) and just copy that formula to the entire column. then just click the decrease decimal button so that it rounds up the decimal numbers to the nearest whole value. Copy the new column and paste it back into the celltimes.txt file and save... takes me less than 30 sec.

Just fyi this is my method to rip just the movie from a NTSC DVD:
1. rip themovie to hard drive
2. run DVD2AVI and force film (23.976) and demux ac3 track
3. run vfapi convert to make d2v file avi
4. cce avi file twice to get multipass vbr at maximum possible size
5. edit celltimes.txt file with above method
6. author files with ifoedit .95 then get VTS Sectors
7. burn with prassi
8. watch movie

I've done dozens of movies with this procedure and have yet to come across a movie that I couldn't do... If anyone wants a more detailed guide just let me know.

~Sy

[jdobbs]

Bingo!!

[oggiedoggie]

I follow the steps here, Sy; thanks! Can I get rid of the VFAPI step and re-encode with TMPGEnc from the d2v directly?

What about problem discs like Disney and Star Wars Ep2, where different language titles and credits are done with multiple angles? (I think it's a RIP setting, or maybe using Ifoedit on the rip?)

I was using ChapterXTractor and replacing the celltimes.txt stops with its output. I thought of using Excel too.

Thanks everyone!

[jdobbs]

On multiangle titles I just pull out the angle I want using SmartRipper.

[Sy]

I use IFOEDIT .91 when using vob extras to remove multi angles.. doesn't seem to be working on .95
~Sy

[quantum]

Originally posted by Sy
I've been doing the .8 multiply on the cell times for a while now... it works perfectly.. the new chapter stops come within a frame or 2 of the original. What about the time discrepancy for the total movie length? Example, the actual move is 1hr44mins, the ifoedit version is shown as 1hr23mins in the player.