Recovering JPEG images from Microsoft Office document

[Walter Ingraham]

I've got several files that seem to be MS Office documents (the header is
D0 CF 11 E0 A1 B1 1A E1).

Each file contains more than one chunk of data that appear to be JPEG
images with abbreviated headers (not JFIF).

Each chunk has an SOI (FFD8) and EOI (FFD9) marker. There are no APPn
markers, but there are SOF0 (FFC0) and SOS (FFDA) markers.

The problem is that the SOF0 header does not match the definition that I
know. Embedded in the header are the letters 'RGB'. I'm guessing (WAG)
that this is MS code to indicate that there is a standard pallette being
used.

Here is a typical example:
FF C0 00 14 08 00 60 00 41 04 52 11 00 47 11 00 42 11 00 41 11 00

Following the SOF0 header is the SOS header. It too contains the ASCII
for 'RGB':

FF DA 00 0E 04 52 00 47 00 42 00 41 00 00 3F 00

My question is this: Is there a JFIF header that I can append to the data
stream that will make these images readable?

Here is the complete header as found in a "chunk":

FF D8 FF C0 00 14 08 00 60 00 41 04 52 11 00 47 11 00 42 11 00 41 11 00
FF DA 00 0E 04 52 00 47 00 42 00 41 00 00 3F 00

Any thoughts?

[Robert Maas, http://tinyurl.com/uh3t]

Post by Walter Ingraham
Any thoughts?

What happens when you try to load the document into MicroSoft Word?
(If it's in MicroSoft Word, saving as Web page might split all the
images into separate files.)