We have all been in this situation: you have two files which are subtly different, and you need to take a quick look in them. Let me point you towards xxd which is part of vim. Users of other editors can stop reading now! :-)
dpavlin@t61p:~/t61p/utf-8-accents$ xxd good 0000000: 506f 7669 6a65 736e 6f20 6172 6865 6f6c Povijesno arheol 0000010: 6fc5 a16b 6920 766f 6469 c48d 2070 6f20 o..ki vodi.. po 0000020: 4f73 6f72 750a Osoru. dpavlin@t61p:~/t61p/utf-8-accents$ xxd bad 0000000: 506f 7669 6a65 736e 6f20 6172 6865 6f6c Povijesno arheol 0000010: 6f73 cc8c 6b69 2076 6f64 6963 cc8c 2070 os..ki vodic.. p 0000020: 6f20 4f73 6f72 750a o Osoru.
Now, if I could only find peace of perl code which puts accents back to characters...
The really great thing about xxd is that it can turn a hexdump back into a binary file. I haven't come across anything that can do that since I wrote my own utility for the Sinclair QL in 1986...
I used it just last week to turn a captured serial port U-boot session of a hexdump of a flash region back into a binary file I could flash on another board:
clean up the captured output, write to boot.txt