WMClear
Description
Description from WMClear's homepage on Sourceforge.net:
wmclear utility is able to automatically remove Windows Media Video
script commands from WMV files.
wmreader utility is also included, which is able to extract Windows
Media script commands and their timings from WMV files into an
ordinary text file."
Why Do We Need this Webpage?
I encountered a few critical bugs with wmreader, and found the
WMClear code posted to SourceForge.net did not have support options
or other options for submitting patches. And as such, this
page will host some patches required for wmreader to work properly.
Also for those using other operating systems, this page might help
those trying to watch the Martin Dickey's CSE 142 Media Player WMV
videos, which contain embedded ASF scripts for automatically
changing the JPEG slides during his lectures. The wmreader
utility will print out these time markers.
Usage of WMReader for CSE142 video.wmv Videos
1) Apply the wmreader_fixbadcast-20131214.patch
and build, or "make clean && make"
2) Run the utility on the video.wmv file, and you should now have a
slides.txt file. If it's still zero bytes in size, manually
remove the slides.txt file and rerun wmreader.
3) Notice there are duplicate entries for loading the slide JPEG
image. I think the second is to ensure the image loaded.
Can omit the second redundant entry (containing a one second offset
from the initial call to load the next image) using, "sed -n 'p;n'
slides.txt".
4) You should now have a ASCII Text file containing the times to
load the next image. The next step is to figure out how to use
an image viewer to load JPEG images at specified intervals, while
playing the video within your favorite video player such as
mplayer. I was thinking, maybe start two mplayer processes,
one to watch the video and the second mplayer process to process the
JPEG slides using the slides.txt times. A mencoder process
should be able to process the JPEG files into a video file, but
think simply piping the mencoder output to mplayer might be more
easier for users.
Patches
wmreader_fixbadcast-20131214.patch
- This doesn't completely fix the casting of the pointer to pointer
integer, as now the STDOUT is treated as a binary file when piped to
a file (ie. "wmreader video.wmv" > output.log) instead of an
ASCII Text file. As such, piping to grep (or "wmreader
video.wmv|grep Slide") no longer works. This needs additional
work! You should at least get a slides.txt file after running
wmreader, or remove the old slides.txt and rerun wmreader.
(Now just need to fix why STDOUT and slides.txt file has double
slide entries. See TODO "STDOUT and slides.txt contain double
...")
TODO
1) Hard job. Needs indent/reformatting?
2) STDOUT and slides.txt contain double start time entries, with an
offset of one second for the duplicated slide. The first
initial slide start time is the correct start time. The second
time is incorrect, according to the same video rendered through
"Windows Media Encoder Editor", which only shows the initial slides
start time. I still do not know why there's a second entry for
every slide's start time with one second offset. Likely to
ensure the slide loaded the specified JPEG file, and can be
discarded. Use "sed -n 'p;n' slides.txt" to print every other
line of the video.wmv.