A normal mpeg2-encoded stream is far from perfect, the artifacts are clearly visible if you pay a little attention. Please don't tell me that my encoder is lame (that's for sound, not video). And I'm talking about adding another set of artifacts on top of that.
They are not going to futz around with software players this time. Hardware only, and tamper-resistant to boot. You will get "your" key with "your" player, and you will like it.
Say, $10 worth of protection will guard against an attacker with less than $1,000,000 to blow for five years. How many people in the world are willing to share at this point?
I say that you can't chain different codecs without either quality or size degradation. The emphasis is on two words: chain, implying there's more than one, and different, implying they're not instances of the same codec. I hope this is clear enough. Did I say two or more different codecs? This has nothing to do with pipes, files, processes, kernels, drivers, or your toaster. All you need is to encode a raw stream, decode it, and then encode again. By the way, the second codec should not be the same as the first, in case you missed that. Oh, and the compression ratio should be reasonable at both ends. It is completely immaterial what's between the two codecs: files, pipes, sockets, carrier pigeons, or stone tablets.
Imagine this: your DVD player is your primary keyserver, configured by the factory to recognise and serve at most 3 secondary key servers and at most 10 players. Each of your secondary key servers work with at most 1 tertiary key server and 3 players. You now have a system which is more than scalable and robust enough for home use.
Now you can move encrypted files however you wish, but your player will constantly ping your keyserver for the key, and if it gets worse than say 1ms average response time it stops playing.
OK so this scheme is full of holes too, but I don't want to give them any more ideas.
A general-purpose computer is much harder to contain than a simple one-purpose device like a DVD player. When they start making single-chip equivalents of modern desktops, you will have to start worrying about Palladium.
As of now, it is entirely possible and even feasible to make a very tamper-resistant DVD player. As in, you'll need $5,000,000 worth of equipment to break in. The technology is here.
people don't get it. THE COMPROMISED TV WILL GET A NEW KEY VIA NORMAL SOFTWARE UPGRADE. Probably transparently too, if it can phone home via your broadband connection. End of story.
The player(s) will get a new key instead of the revoked one. And if the players are done "right", you won't be able to reverse engineer their keys in your lifetime. Somebody else would crack them easily, provided that somebody else is NSA.
because the originaql stream is either analog or comes from a mpeg2 decoder (DVD, cable, sat, they are all likely using mpeg2.) You are able to re-encode it without much loss of quality because it keeps throwing away bits that were already thrown out in the original encoding process. Now if you chain two entirely different codecs, and try to keep file sizes reasonable, you might get a lower quality rip with lots of visible artifacts.
Only your player has the key to play your media. It is buried deep in silicon so you cannot feasibly extract it. You will need elaborate equipment worth of few million USD and lots of time and expertise, and when you're done they make that key obsolete.
You can copy the encrypted content however you wish, but it's only playable in your player, not anyone else's. So sorry, no $59.99 decryptors at Best Buy.
or even just hijack the analog signal to get 99.9% of the original.
Sure thing, and 300% of the original gigabyte count. Or get same file size with 50% quality, because compression artifacts from two different codecs will amplify each other. Or spend for years researching an exact duplicate of their algo (which only exists in hardware, remember?) and then they change it. Your choice.
I don't quite get why people are so optimistic regarding this nightmare.
it can be done. Of course the analog hole will be with us for a long time, and Far Eastern wholesale pirates are not going to go away either. But these two gaps are narrow enough for the industry executives not to worry.
This kind of job requires competent engineering, and I sure hope that people employed by the DRM industry are incompetent (or can hide their competence well enough:)
There's an interesting variation
on
Defining Google
·
· Score: 1
of the pirates problem. Replace "50%" with ">50%". Assume that, all other things being equal, the pirates don't want to execute their comrades. That is, if pirate X doesn't die and gets Y coins both in scenario A and scenario B, and less pirates are executed in scenario A, X prefers scenario A over B.
Slashdot Mirror
Sorry, I didn't quite finish reading your post before I hit reply.
I'm not an anal video freak but I notice these things sometimes. I do get reasonably good equipment though.
Get sued when someone burns himself trying to repair a broken set.
can't play HD content, so you need to get everyone to throw the old equipment out anyway.
A normal mpeg2-encoded stream is far from perfect, the artifacts are clearly visible if you pay a little attention. Please don't tell me that my encoder is lame (that's for sound, not video). And I'm talking about adding another set of artifacts on top of that.
They are not going to futz around with software players this time. Hardware only, and tamper-resistant to boot. You will get "your" key with "your" player, and you will like it.
Say, $10 worth of protection will guard against an attacker with less than $1,000,000 to blow for five years. How many people in the world are willing to share at this point?
I say that you can't chain different codecs without either quality or size degradation. The emphasis is on two words: chain, implying there's more than one, and different, implying they're not instances of the same codec. I hope this is clear enough. Did I say two or more different codecs? This has nothing to do with pipes, files, processes, kernels, drivers, or your toaster. All you need is to encode a raw stream, decode it, and then encode again. By the way, the second codec should not be the same as the first, in case you missed that. Oh, and the compression ratio should be reasonable at both ends. It is completely immaterial what's between the two codecs: files, pipes, sockets, carrier pigeons, or stone tablets.
Imagine this: your DVD player is your primary keyserver, configured by the factory to recognise and serve at most 3 secondary key servers and at most 10 players. Each of your secondary key servers work with at most 1 tertiary key server and 3 players. You now have a system which is more than scalable and robust enough for home use.
Now you can move encrypted files however you wish, but your player will constantly ping your keyserver for the key, and if it gets worse than say 1ms average response time it stops playing.
OK so this scheme is full of holes too, but I don't want to give them any more ideas.
As of now, it is entirely possible and even feasible to make a very tamper-resistant DVD player. As in, you'll need $5,000,000 worth of equipment to break in. The technology is here.
people don't get it. THE COMPROMISED TV WILL GET A NEW KEY VIA NORMAL SOFTWARE UPGRADE. Probably transparently too, if it can phone home via your broadband connection. End of story.
The player(s) will get a new key instead of the revoked one. And if the players are done "right", you won't be able to reverse engineer their keys in your lifetime. Somebody else would crack them easily, provided that somebody else is NSA.
because the originaql stream is either analog or comes from a mpeg2 decoder (DVD, cable, sat, they are all likely using mpeg2.) You are able to re-encode it without much loss of quality because it keeps throwing away bits that were already thrown out in the original encoding process. Now if you chain two entirely different codecs, and try to keep file sizes reasonable, you might get a lower quality rip with lots of visible artifacts.
You can copyright a particular expression of one. For ideas there are patents. Just so you know, and good luck.
a clean room environment and lots of expertise. Which are not exactly things most people can find in their basement.
Have this chip. It's a decryptor, a decoder, and a DA convertor all in one. The key is in the silicon. Go extract it.
does your stream comes from a DVD in the first place? I guess not.
and the PostIt sticker with the passphrase which was affixed to it. And family silverware, and car keys, your honour.
You can copy the encrypted content however you wish, but it's only playable in your player, not anyone else's. So sorry, no $59.99 decryptors at Best Buy.
The defence has no further questions.
There's not enough energy in the Universe to brute force a single 256-bit key.
They are on the same chip. Go ahead.
or even just hijack the analog signal to get 99.9% of the original.
Sure thing, and 300% of the original gigabyte count. Or get same file size with 50% quality, because compression artifacts from two different codecs will amplify each other. Or spend for years researching an exact duplicate of their algo (which only exists in hardware, remember?) and then they change it. Your choice.
I don't quite get why people are so optimistic regarding this nightmare.
This kind of job requires competent engineering, and I sure hope that people employed by the DRM industry are incompetent (or can hide their competence well enough :)
of the pirates problem. Replace "50%" with ">50%". Assume that, all other things being equal, the pirates don't want to execute their comrades. That is, if pirate X doesn't die and gets Y coins both in scenario A and scenario B, and less pirates are executed in scenario A, X prefers scenario A over B.
is mplayer and it runs as 'nobody', so nyah. Not that I would ever use Kazaa or anything...