Tropican
31st December 2006, 20:56
Hello all,
Well, the 23.976 > 29.97 pulldown with 15 GOP length has been beaten to a pulp, but I couldn't find anything on 25 > 29.97 pulldown with the said GOP length.
Shouldn't it work?
15 x 29.97/25 (1.1988) = 17.982 = 35.964 fields
Even if you round up, you still get valid values for NTSC DVD video.
Yet here's the result of attempting to multiplex with Muxman:
MuxMan version 0.15P
Accepted video C:\MPGs\VTS_03_1.m2v.pulldown.m2v size = 4231466501
Accepted audio C:\MPGs\VTS_03_1 T02 2_0ch 256Kbps DELAY 0ms.ac3
13:38:30 Begin multiplex VTS01.
Title Segment List
Segment_1
Maximum audio duration 232218 fields.
Starting scene Segment_1_scn1 at 00:00:00:00
P-STD buffer underflow by 2514 bytes at 83876559, sector 517751.
SeqEnd at FC370E01.
Bytes remaining in buffer = 2514.
Bitrate - avg: 9138623, min: 381911 (lba 0), max: 9988932 (lba 201221).
Shortest GOP has 2 fields, longest GOP has 36 fields.
Fields: 232220, VOBU: 6493, Sectors: 2160943.
Video decoder buffer underflowed 1 times.
Check c:\muxman.log for details.
Resulting DVD will probably not play smoothly.
I would ignore such a problem and see if there's actually a problem in playback, but Muxman doesn't produce IFO or BUP files when this occurs.
Note that multiplexing with a GOP length of 12 does not result in any errors.
The dgpulldown readme has an interesting comment about the issue:
Note that when applying pulldown for authoring a DVD, as for PAL->NTSC
conversion, you have to ensure that the GOP size is not too large after pulldown.
There is a text file included in this distribution with further details.
For complete safety, encode your MPEG2 file with a GOP size of 12 or less.
The text file mentioned that's included is about 23.976 > 29.97 conversions, however. (and though it comes to the right conclusions, seems to be wrong AFAIK as the DVD spec measures things by the displayed frames, not the physical frames in the bitstream, right?)
Well, the 23.976 > 29.97 pulldown with 15 GOP length has been beaten to a pulp, but I couldn't find anything on 25 > 29.97 pulldown with the said GOP length.
Shouldn't it work?
15 x 29.97/25 (1.1988) = 17.982 = 35.964 fields
Even if you round up, you still get valid values for NTSC DVD video.
Yet here's the result of attempting to multiplex with Muxman:
MuxMan version 0.15P
Accepted video C:\MPGs\VTS_03_1.m2v.pulldown.m2v size = 4231466501
Accepted audio C:\MPGs\VTS_03_1 T02 2_0ch 256Kbps DELAY 0ms.ac3
13:38:30 Begin multiplex VTS01.
Title Segment List
Segment_1
Maximum audio duration 232218 fields.
Starting scene Segment_1_scn1 at 00:00:00:00
P-STD buffer underflow by 2514 bytes at 83876559, sector 517751.
SeqEnd at FC370E01.
Bytes remaining in buffer = 2514.
Bitrate - avg: 9138623, min: 381911 (lba 0), max: 9988932 (lba 201221).
Shortest GOP has 2 fields, longest GOP has 36 fields.
Fields: 232220, VOBU: 6493, Sectors: 2160943.
Video decoder buffer underflowed 1 times.
Check c:\muxman.log for details.
Resulting DVD will probably not play smoothly.
I would ignore such a problem and see if there's actually a problem in playback, but Muxman doesn't produce IFO or BUP files when this occurs.
Note that multiplexing with a GOP length of 12 does not result in any errors.
The dgpulldown readme has an interesting comment about the issue:
Note that when applying pulldown for authoring a DVD, as for PAL->NTSC
conversion, you have to ensure that the GOP size is not too large after pulldown.
There is a text file included in this distribution with further details.
For complete safety, encode your MPEG2 file with a GOP size of 12 or less.
The text file mentioned that's included is about 23.976 > 29.97 conversions, however. (and though it comes to the right conclusions, seems to be wrong AFAIK as the DVD spec measures things by the displayed frames, not the physical frames in the bitstream, right?)