[MPlayer-users] [OT] AC3 vs. DTS

[Giacomo Comes]

On Tue, Aug 02, 2005 at 06:15:18PM +0300, Ivan Kalvachev wrote:
> On 8/2/05, Giacomo Comes <comes at naic.edu> wrote:
> > On Mon, Aug 01, 2005 at 05:14:33PM -0400, Rich Felker wrote:
> > > On Mon, Aug 01, 2005 at 03:58:11PM -0400, Giacomo Comes wrote:
> > > > On Mon, Aug 01, 2005 at 02:12:36PM -0400, Rich Felker wrote:
> > > > > On Mon, Aug 01, 2005 at 11:43:00AM -0400, Giacomo Comes wrote:
> > > > > > > > For the future I'm still waiting for the next big thing, since I'm not
> > > > > > > > that happy with the heavily protected new HD formats. I still want to be
> > > > > > >
> > > > > > > Heavily protected? Any cryptography where the attacker has the key is
> > > > > > > not heavily protected, it's just basic obscuring/obfuscation and will
> > > > > > > be cracked in no time, just like CSS.
> > > > > >
> > > > > > That's completly false. You are talking only about weak encription
> > > > > > algorithms like CSS. CSS was a "secret" algorithm. Once people understood
> > > > > > how it was working (knowing some keys) they found how weak it was
> > > > > > and how to guess the keys analizing the encrypted data (what libdvdcss does).
> > > > >
> > > > > You fucking didn't read what I wrote. If you can play the movie, YOU
> > > > > HAVE THE KEY. DRM/"content protection" is always obfuscation, not
> > > > > cryptography.
> > > >
> > > > It seems that I make you angry, because you start insulting people.
> > >
> > > Not insulting, just yelling. And it's because you called my statements
> > > false when you apparently have no clue about crypto.
> > 
> > I was thinking the same about you :-)
> > 
> > > > Anyway I read what you wrote. I just want to remember you that CSS
> > > > was not cracked in no time. It took some years after the introduction of
> > > > the DVD technology, and the crackers succeed because they found a key not
> > > > obfuscated in a software player.
> > >
> > > DVD was cracked way before DVD drives were common in computers. Before
> > > that, there was really very little motivation:
> > >
> > > 1. Compression software and bandwidth for distributing movies over the
> > >    net were severely lacking at the time, so the warez scene had
> > >    little motive. Most people were sharing analog-captured .asf and
> > >    .rm files if anything at all.. :)
> > >
> > > 2. Since few computers had DVD drives, no one cared if they could play
> > >    their DVDs on their non-windows computer.
> > >
> > > The digital media scene on the net is much more developed now, and I'm
> > > confident that competent people will attack whatever new crap they
> > > throw at us quickly and liberate it as well.
> > >
> > > > In your statement you say that knowing a key "will make the cryptography
> > > > be cracked in no time"
> > > > That's only valid if the encryption algorithm is weak.
> > >
> > > No, this statement is blatently false. Assuming you know the key (or
> > > have it somewhere and just haven't found it), the only factors making
> > > it difficult to crack the 'protection' are the level of obfuscation of
> > > the key and the level of obfuscation of the encryption (i.e. using an
> > > encryption method that's not well-known). Both of these are matters of
> > > security through obscurity. The weakness or strength of the encryption
> > > itself is always irrelevant when the only security comes from
> > > obscurity.
> > 
> > Looks like you were thinking about circumventing a protection,
> > but you were talking about breaking cryptography.
> > Circumventing a 'protection' is one thing, it can be easy.
> > Breaking a cryptograph algorithm is another thing.
> > libdvdcss does not circumvent CSS, it breaks the weak algorithm.
> > 
> 
> Well take a look here
>  http://www-2.cs.cmu.edu/~dst/DeCSS/Gallery/
>  http://www.cs.cmu.edu/~dst/DeCSS/FrankStevenson/analysis.html
> 
> CSS is weak algo (40bit) because USA doesn't allowed export of
> stronger encryptions at the time .
> 
> Anyway the point here is that CSS itself is not bruteforced. The weak
> point of DVD is the way keys are stored. As you can read in the
> annalys, it requares only 2^17 tries to find matching key (18 second
> bruteforce on P2-450MHz).
> Not only that, once a key is found we could get all available keys, as
> they are stored on the DVD too.
> 
> Well an chain is as strong as it weakest branch. 
> >From this point of view, no cryptography algorithm could be used as
> DRM scheme because the attacker and the reciver are same person.
> The reciver need the key to legetimely use the content . Once it have
> it, he could become attacker and copy the content.
> Even unsymetric key systems ( e.g. RSA ) could not help because
> attacker will have the decoding key (in order to use the content
> legetimely).
> 
> The only protection is obfiscation of the key.
> Encrypting the key just bring the question "What to do with the key
> for the key?"

CSS was a weak protection scheme, 40bit, not well used.
Obfuscation does not give good security.

But:
CSS was initially defeated once a not obfuscated key was found.
CPPM still resist, althought it uses the same obfuscation technique.

Conclusion:
In practice, obfuscation is effective for a certain amount of
time, and we all should be glad that they insist in using such
protection scheme (because at some point it will be defeated).

> 
> Wish You Good Luck
>   Ivan Kalvachev
>  iive
> 
> P.S.
> About DTS vs DD, I had read few articles (from both sides) it looks
> like the main difference is the LFE, on encoder side. DTS seem to be
> louder and this perceived as better.

That's the logical conclusion people should know.

Giacomo