Originally Posted by AnthonySterling
I'm all out.
The headers are identical, barring the content-length, so it's definitely down to the retrieval of the PDF content I'd say.
Are you using a different method of obtaining the source data? What about output buffering, could this be in play?
I am using the exact same function to get the source data as I am from admin..that's what makes it so strange!! Not sure about output buffering..how can I test that?