Howdy -
When is 80% of something about equal to 100%?
Now that I've got your attention ... ;)
Ordinarily you would expect that encoding fewer frames would
result in a .m2v file that was smaller than encoding more frames.
Right?
That's not what I am seeing and it's very puzzling.
What I did was take a ~88min capture from a tape that was known to
have originated on film. The data was captured as 8bit uncompressed
(in case you're curious that's about 105GB ;)). Stepping thru the
frames the 2:3 telecine pattern is obvious so using 'yuvkineco -F 1'
would work nicely.
Without any 3:2 removal the 158065 frames encoded to a ~30fps movie
with an average bitrate of ~6229Kb/s:
--------------
#!/bin/sh
N=sinbad
~sms/qt2raw /Volumes/LaCie/tmp/$N.mov | \
yuv4mpeg -w 720 -h 486 -x 422 -a 10:11 -i b -r 30000:1001 | \
y4mscaler -v 0 -I active=704x480+8+2 -O sar=src -O size=704x480 | \
yuvmedianfilter -f -t 0 -r 1 -w 12 | \
y4mdenoise -I 0 -t 3 -z 2 -m 30 -M 3 | \
y4mscaler -v 0 -O sar=src -O chromass=420_MPEG2 | \
bfr -b 20m | \
mpeg2enc -D 10 -G 15 -b 6400 -c --dualprime-mpeg2 -E -10 -f 8 -q 2 -K
tmpgenc -4 1 -2 1 -o $N.m2v
----------------------
(the 'qt2raw' program is a local work of art that can extract 8bit
uncompressed D1 (720x486) data from a Quicktime file and emit 4:2:2
planar data)
The "ls -l" + "mplex" summary of the 30fps encoding:
30fps (-R 0 --dualprime-mpeg2)
-rw-r--r-- 1 sms sms 4106845567 Jun 14 05:29 sinbad-30R0dp.m2v
INFO: [mplex] No. Pictures : 158065
INFO: [mplex] No. Groups : 10538
INFO: [mplex] No. I Frames : 10538 avg. size 61285 bytes
INFO: [mplex] No. P Frames : 147527 avg. size 23460 bytes
INFO: [mplex] No. B Frames : 0 avg. size 0 bytes
INFO: [mplex] Average bit-rate : 6229600 bits/sec
INFO: [mplex] Peak bit-rate : 6896400 bits/sec
Nice bitrate, good looking video.
Converting to 24000/1001 progressive (undoing the telecine ) reduces
the number of frames by 20% so the filesize/bitrate should go down
(and leave room on the DVD for a few extras). Correct? 20% fewer
frames should mean a smaller file, right?
So, 'yuvcorrect' (to convert to a top field first stream) and
"yuvkineco -F 1" were added to the pipeline and the "-p" option given
to 'mpeg2enc':
--------------------
#!/bin/sh
N=sinbad
~sms/qt2raw /Volumes/LaCie/tmp/$N.mov | \
yuv4mpeg -w 720 -h 486 -x 422 -a 10:11 -i b -r 30000:1001 | \
yuvcorrect -v 0 -T TOP_FORWARD -T INTERLACED_TOP_FIRST | \
y4mscaler -v 0 -I active=704x480+8+2 -O sar=src -O size=704x480 | \
yuvmedianfilter -f -t 0 -r 1 -w 12 | \
yuvkineco -F 1 -C yuvkineco.txt | \
y4mdenoise -I 0 -t 3 -z 2 -m 30 -M 3 | \
y4mscaler -v 0 -O sar=src -O chromass=420_MPEG2 | \
bfr -b 20m | \
mpeg2enc -D 10 -G 15 -b 6400 -p -c -R 2 -E -10 -f 8 -q 2 -K tmpgenc -4 1
-2 1 -o $N.m2v
----------------------
The encoded file is LESS THAN 1% smaller than just encoding the
~30fps capture without change!
24fps with 3:2 (-p -R 2)
-rw-r--r-- 2 sms sms 4070777716 Jun 13 10:07 sinbad-24R2.m2v
INFO: [mplex] No. Pictures : 126450
INFO: [mplex] No. Groups : 8430
INFO: [mplex] No. I Frames : 8430 avg. size 66178 bytes
INFO: [mplex] No. P Frames : 42150 avg. size 38428 bytes
INFO: [mplex] No. B Frames : 75870 avg. size 24952 bytes
INFO: [mplex] Average bit-rate : 6174800 bits/sec
INFO: [mplex] Peak bit-rate : 6893600 bits/sec
So with 20% fewer frames encoded why is the output file almost
identical in size???
To see if not using B frames would make a difference the same 24fps
encoding above was done again but "-R 2" changed to "-R 0":
24fps with 3:2 (-p -R 0)
-rw-r--r-- 1 sms sms 4042399041 Jun 15 01:21 sinbad-24R0.m2v
INFO: [mplex] No. Pictures : 126450
INFO: [mplex] No. Groups : 8430
INFO: [mplex] No. I Frames : 8430 avg. size 65807 bytes
INFO: [mplex] No. P Frames : 118020 avg. size 29551 bytes
INFO: [mplex] No. B Frames : 0 avg. size 0 bytes
INFO: [mplex] Average bit-rate : 6131600 bits/sec
INFO: [mplex] Peak bit-rate : 7050000 bits/sec
As usual leaving out the B frames produced a smaller file ;) True,
the difference is modest but there are times when an extra 30MB on
a DVD project makes the difference between fitting and not fitting on
the media.
Just one last test to run - that's the 30fps (no telecine removal)
using "-R 2" (instead of -R 0/--dualprime).
--------------
#!/bin/sh
N=sinbad
~sms/qt2raw /Volumes/LaCie/tmp/$N.mov | \
yuv4mpeg -w 720 -h 486 -x 422 -a 10:11 -i b -r 30000:1001 | \
y4mscaler -v 0 -I active=704x480+8+2 -O sar=src -O size=704x480 | \
yuvmedianfilter -f -t 0 -r 1 -w 12 | \
y4mdenoise -I 0 -t 3 -z 2 -m 30 -M 3 | \
y4mscaler -v 0 -O sar=src -O chromass=420_MPEG2 | \
bfr -b 20m | \
mpeg2enc -D 10 -G 15 -b 6400 -c -R 2 -E -10 -f 8 -q 2 -K tmpgenc -4 1 -2 1
-o $N.m2v
----------------------
30fps (-R 2)
-rw-r--r-- 1 sms sms 4129566072 Jun 15 20:11 sinbad-30R2.m2v
INFO: [mplex] No. Pictures : 158065
INFO: [mplex] No. Groups : 10538
INFO: [mplex] No. I Frames : 10538 avg. size 60002 bytes
INFO: [mplex] No. P Frames : 52689 avg. size 31916 bytes
INFO: [mplex] No. B Frames : 94838 avg. size 19144 bytes
INFO: [mplex] Average bit-rate : 6263600 bits/sec
INFO: [mplex] Peak bit-rate : 7085600 bits/sec
~2% larger than the smallest of the encodings (which was the 24
fps -R0 one).
Hardly seems worth the cpu cycles to do the telecine removal.
Anyone care to guess why encoding 20% fewer frames gives the same file
size as encoding all of the frames?
Cheers,
Steven Schultz
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
Mjpeg-users mailing list
Mjpeg-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mjpeg-users
