MJPEG Tools Svn

First bug - well, actually a typo in the usage() function.

Patch from Nicolas Boos to more accurately compute the available video bitrate

film-FX added for testing purposes...

The usual list of files for cvs to ignore.

Put the same "quant=0" in for SVCD STILLS as was used for VCD STILLS. The
comment that was present makes it clear what the goal of the quant=0 is.
Also, for the SVCD_NSR (user specified SVCD) set a 230KB video_buffer if
-V was not given by the user.

Add y4munsharp filter. Based on the GIMP's unsharpmask plugin.

interlacing information now read from stream. If not specified in stream
and not on the commandline, bail out...

corrected runtime options and help message...

Ok, here we go... What has happened so far?

1. switched back to circular full-search (with early termination) as it
seems to be the only MC-algorithm suitable for deinterlacing. See the
source, if you want to know more about it...
2. Completely implemented Generalized Sampling Theorem (GST).
3. Implemented Scene-Change-Detection
4. Removed all ugly compiler-warnings
4b.removed (hopefully) all memory-leaks/memory-problems
5. "Soccer-Test": passed... :-)))
6. better than my Phi**** 100Hz TV-Set-deinterlacer...
7. I have forgotten, to ...

Still to be done:
-----------------

- Man-page/Dokumentation (I hate that...)
- Fix runtime options
- field-dominance detection

Have fun!

Ok, folks,... all I guess I can do any further on this thing is clean the
code from all those ugly warnings and leftovers...