On Fri, Sep 24, 2010 at 8:11 PM, Mathieu Bouchard
<matju@artengine.ca> wrote:
On Fri, 24 Sep 2010, Matteo Sisti Sette wrote:
(guess you mean p-frames and b-frames here)
I thought that m-jpeg was just jpeg applied to each frame independently i.e. all frames were I-frames (I had read that in wikipedia).
You're probably right, but then I wouldn't know what's the difference between Photo-JPEG and Motion-JPEG. Can Motion-JPEG-A be decoded with a Photo-JPEG decoder ? If that's the case, then Motion-JPEG-A would be just a different tuning of a Photo-JPEG encoder. (but Motion-JPEG-B can't be decoded the same).
I can't tell much more that isn't just another guess.
http://vjforums.com/showthread.php?t=12587or a shorter one:
According to the QuickTime specification, MJPEG A and MJPEG B are both
field-based, i.e. each of the two interlaced fields making up a video signal is
compressed individually, using the JPEG algorithm. The difference between MJPEG
A and MJPEG B is that MJPEG A supports markers in the bitstream whereas MJPEG B
does not. Photo JPEG is frame-based, according to the QuickTime specification.
To make things confusing, you sometimes see MJPEG A and MJPEG B also allowing
frame-based compression. This is the case in Adobe Premiere for instance. (
http://w2.alkit.se/servo/servofiles/)
Andras