Embedded characters in outlook emails

[Guest]

For a few of my users, I am seeing a set of embedded characters in their
email messages, which cause varying degrees of problems in other clients. In
a browser-based email client, the embedded characters are ÿþ, resembling a Y
with 2 dots above it, and a lowercase b superimposed on a lowercase p. In
some clients this will result in the characters being displayed at the top of
the email, and in others, the embedded characters appear to "expand" into
200-400 characters which do not render correctly in the font being used, and
appear as either boxes and chinese characters, or as boxes and a few symbols.

I have run multiple virus scans and do not detect any type of virus on
either the local pc or the exchange server. I suspect that these may have
been somehow embedded in the template used to create new email messages, but
have not been able to locate it at this point.

Any suggestions on what this is and where to look are greatly appreciated,
it is so far very limited in scope but has the effect of rendering email
unreliable for these people to communicate. (All of these are in-house users,
though the characters will travel to outside addresses as well as in-house)

 

[Pat Willener]

You are not telling us the full details;
- are these incoming or outgoing messages?
- what is the encoding of these messages?
- what Outlook version?
- what mail server (in case of incoming messages)?

'ÿþ' is 0xFFFE, which is not a valid character. Where does it come from?

 

[Guest]

Sorry, I thought I had covered those by inference. I did not mention the
server, however, which is running Exchange 2003.

These are outgoing messages, my users rarely send themselves incoming
messages (These are occurring when one of the affected users either
sends or forwards a message to another user, and the receiving user will see
the result. I have tested and the behavior is consistant when sent to an
outside address as well as to in-house users.)

Encoding of the messages is HTML, on the receiving end a user viewing in
text-only mode will see the larger number of characters in a form that the
selected font does not appear to be able to render; my assumption based on
the display is that it is showing a large number of high bit characters

Outlook version seems irrelevant in the receiving end, except in terms of
the display, I have users running Outlook 2003 for the most part, a few still
running Outlook 98, and all users have access to OWA 2003, though it is not
too widely used inhouse as a primary client. The OWA does the best job of
rendering the messages when the embedded characters appear, showing the "yb"
like sequence, then showing the remainder of the message properly, and is my
current work-around in this situation. Sending users are using either
Outlook 2003 or 98, though most have been upgraded to 2003. Client OS is XP
in all cases but one, which is still running 2000 professional.

The "yb" characters will be displayed if the message is rendered by the
client as "readable" at the beginning of the body, on a line by themselves,
in the same font size as the body of the message, but always in black, where
portions of the message may appear in colored text if the user has selected
that. (typically blue text in some places, black in others.)

No Stationary has been selected or anything else I have been able to
determine as the basis for embedding the characters.

Hopefully this provides some additional clarity on the otherwise baffling
issue; I have so far been unable to make any headway, even after viewing the
messages in a hex editor to look for any additional characters which are not
displayed, and as a captured HTML image. (The characters show up "outside" a
tag as vs between a pair, generally following a </BR> when captured as HTML)

 

[Pat Willener]

Ok, a little bit more clear now. A few more questions.
The user that sends the message (that ends up having 'ÿþ' characters)
sends the message in HTML format.
- what is the font s/he is using?
- what is the encoding of the message (Format | Encoding)?

What do you mean by high bit characters - characters outside the ASCII
range ( > 0x7F), or multibyte (Unicode) characters?