Re: Yet another x264 issue

Hi!

> $ wc -l /path/to/offending/logfile
81520 /home/vdb/Media/DVD/log_files/life_on_earth_4_1.log


> $ tail -n 20 /path/to/offending/logfile
in:81498 out:81499 type:b q:12.27 tex:282 mv:27 misc:259 imb:0 pmb:21 smb:1431 
d:-;
in:81503 out:81500 type:P q:10.00 tex:398 mv:191 misc:307 imb:27 pmb:2 
smb:1423 d:-;
in:81501 out:81501 type:B q:11.14 tex:236 mv:26 misc:258 imb:0 pmb:20 smb:1432 
d:-;
in:81500 out:81502 type:b q:12.17 tex:100 mv:17 misc:211 imb:0 pmb:9 smb:1443 
d:-;
in:81502 out:81503 type:b q:12.27 tex:243 mv:23 misc:262 imb:0 pmb:20 smb:1432 
d:-;
in:81507 out:81504 type:P q:10.00 tex:398 mv:191 misc:307 imb:27 pmb:2 
smb:1423 d:-;
in:81505 out:81505 type:B q:11.14 tex:236 mv:26 misc:258 imb:0 pmb:20 smb:1432 
d:-;
in:81504 out:81506 type:b q:12.17 tex:100 mv:17 misc:211 imb:0 pmb:9 smb:1443 
d:-;
in:81506 out:81507 type:b q:12.27 tex:243 mv:23 misc:262 imb:0 pmb:20 smb:1432 
d:-;
in:81511 out:81508 type:P q:10.00 tex:398 mv:191 misc:307 imb:27 pmb:2 
smb:1423 d:-;
in:81509 out:81509 type:B q:11.14 tex:236 mv:26 misc:258 imb:0 pmb:20 smb:1432 
d:-;
in:81508 out:81510 type:b q:12.17 tex:141 mv:18 misc:209 imb:0 pmb:10 smb:1442 
d:-;
in:81510 out:81511 type:b q:12.27 tex:282 mv:27 misc:259 imb:0 pmb:21 smb:1431 
d:-;
in:81515 out:81512 type:P q:10.00 tex:398 mv:191 misc:307 imb:27 pmb:2 
smb:1423 d:-;
in:81513 out:81513 type:B q:11.14 tex:236 mv:26 misc:258 imb:0 pmb:20 smb:1432 
d:-;
in:81512 out:81514 type:b q:12.17 tex:100 mv:17 misc:211 imb:0 pmb:9 smb:1443 
d:-;
in:81514 out:81515 type:b q:12.27 tex:243 mv:23 misc:262 imb:0 pmb:20 smb:1432 
d:-;
in:81519 out:81516 type:P q:10.00 tex:398 mv:191 misc:307 imb:27 pmb:2 
smb:1423 d:-;
in:81517 out:81517 type:B q:11.14 tex:236 mv:26 misc:258 imb:0 pmb:20 smb:1432 
d:-;
in:81516 out:81518 type:b q:12.17 tex:100 mv:17 misc:211 imb:0 pmb:9 smb:1443 
d:-;

Indeed, the "out" ends at 81518, while the message talks about an offending 
81519. 


> It happens randomly? If I understood correctly, sometime it should work
> fine. 
Exactly.

> Have you noticed any dropped frame for the failing sources? See
> [*] above.
There was this message at the end of the first pass :

[iodir.c] warning: left out 1 directory entries3:17,  ( 0| 8| 9)


BTW, sometimes, the second path shows this message close to the end :

[encode_x264.so] critical: 2nd pass has more frames than 1st pass (81219)                                                            
[encode_x264.so] critical: continuing anyway, at constant QP=26                                                                      
[encode_x264.so] critical: disabling adaptive B-frames

But it still finishes the transcode. The first pass says it has transcoded 
81223 frames, just as the second. 
So it would seem that somehow, 4 frames are not recorded in the log file 
during the first path, though transcode did read them.

	Richard