[Ffmpeg-devel] libavcodec h264 decoder

Stefan Gehrer stefan.gehrer
Sat Dec 16 19:23:34 CET 2006


Loren Merritt wrote:
> On Fri, 15 Dec 2006, michael benzon chua wrote:
>> On 12/15/06, Stefan Gehrer <stefan.gehrer at gmx.de> wrote:
>>
>>> Interesting concept, but as soon as you use some real cryptography 
>>> instead
>>> of just toggling the sign I think the bitstream size will increase 
>>> as some
>>> seemingly random coefficients will be harder to compress than what the
>>> encoder outputs. Or am I missing something?
>>
>> I think this is the case as well. The size of the output .264 file 
>> increased
>> when I changed the dct values. I'm hoping the effect isn't too 
>> significant,
>> and that the encryption is still reversible.
>
> The size of the .264 shouldn't change if you just toggle the 
> coefficient's sign, because sign isn't compressed.
> Why do you want to mess with the guts of the codec, rather than 
> encrypting the whole stream?

Out of curiosity, I just played one of the Apple HD trailers and tried 
two things:
inverting all signs of DCT coefficients and randomly inverting signs. In 
the first
case, the video is still surprisingly clear, mainly looks like a 
photo-negative. In the
latter case, the image is quite distorted but still leaves you with some 
faint idea of
what's going on.
I could see an application here e.g. on set-top boxes: Instead of just a 
blank screen
telling the user that a channel is encrypted, you can give her an idea 
of what she
is missing, in order to increase the urge to subscribe to that channel.
Reminds me of the old days of scrambled analogue signals.

Stefan




More information about the ffmpeg-devel mailing list