in reply to Problem with UTF-8 email

I'll agree with zwon -- you haven't given enough information. It would help if you could show a snippet of input data and resulting output, along with a minimal (but runnable) perl script that shows how you converted that input to that output.

It'll also help if you present the data in the form of hex byte values. The thing that you say is the "weird character" that whitespace is converted to is not a "character" -- it's a string of up to 14 bytes, two of which are in the ASCII range ("3" and space, which happen to be adjacent in the string: ... 0x33 0x20 ...). I won't hazard a guess as to what the other bytes might represent, because I'm not even sure whether the string you posted is an accurate copy of the output you got. (That's why the hex byte-value dump is important.)

So, your question can't be answered yet, because you haven't shown us what the input looked like or what sort of code is creating the output. And it's not even clear what the output really is.

In case it helps, here's a perl one-liner for generating a simple hex byte dump:

perl -lpe '$_=unpack("H*",$_);s/(..)/$1 /g'