Here's an unrealistic utopia. A dream scenario for what the music industry would like to be possible.

The year is 2009 and the following happened:
- the majority of the CDs now are unrippable. The only way to copy them is to play it and connect a cable from the amp to the PCs soundcard(analogue recording).
- online music stores now use uncrackable DRM.

Then average-joe has the following choices:
- buy it online. However, then he cannot play it on his cd-player and it will only play with "certified" players. He also cannot easily use it in a family or circle of friends with multiple playback devices.
- buy it as a CD. However, then he cannot play it on his portable - unless he goes through the hassle of recording it manually to the PC.
- download it illegally from a filesharing network. Then he can play it everywhere and share it easily in a family or friend-circles.

Pick your choice.... buy an inferrior product or get a better service for free.


In the meantime: bandwidths on the internet reach a state where decentralized encrypted anonymous networks achieve reasonable speed.

Copy-protection is an utopia which if it would become true would be for the worse of all parties.

- Lyx

This very well could happen, but hey, if you can spend 10000$ on a liquid crystal on silicon hd widescreen television and a surround sound speaker system that only supports 'fraunhofer digital', then you can hop on a plane to taiwan, or pay some geek to install some hacked firmware.

I think bootlegged music will always suck, except maybe on D*r*ct C*nn*ct type networks.

I think their utopia goes a bit further.

Pay for each time you play a song. If you have friends over, they each have to pay too.

Average joe doesn't care about that. He's satisfied with reencoded 128kbit CBR MP3. Besides, when the effort to encode a music CDs rises, then the quality of the rips on filesharing networks will also automatically rise, because the percentage of knowledgeable people who do them will rise. Also, in 2009 lossless will be the standard and the quality of the rip will only depend on the process of manually recording the protected CD.

- Lyx

Here's some more food for thought:

Another front on which the music industry is fighting is to make development of filesharing apps illegal. Ignoring the ambigious definition of what makes a filesharing app and what not (theoretically, a browser, IM or email-app can also be used for filesharing), lets unrealistically asume that they are successfull in doing so, and developing filesharing apps will be outlawed.

Switch of scene: Decentralized, encrypted and anonymous networks do already exist in 2005. Because of the current bandwith limits of most internet-users they are ineffective and therefore unpopular. But this problem will solve itself in the next years and for sure until 2009. The 2nd generation of such networks is already in development and does not just act as a filesharing app but instead as a network layer: you can run anything from IM, Webservers, E-Mail-Servers, CVS, etc. on them - on top of the normal internet.

Now lets return to our initial asumption: development of filesharing apps becomes outlawed... then one of the following will happen:

1. if the crackdown happens early (before bandwidth capacities will rise enough), then this will be a major blow to development of filesharing apps. However, they will soon recover and in a best case scenario(for the music industry) they will flourish again 2-3 years later - long before 2009 - because they will simply use the above mentioned networks for development and distribution.

2. if the crackdown happens late, then filesharing developers will fast adapt to the change with almost no interruption.


Summary:
- the tighter copy-protection, the lower the sales and the higher the amount of illegal downloads
- the more governments lock down the internet, the less controllable it becomes
- the harder the laws, the less effective and the more irrelevant they become


- Lyx

It's a fact that the disc labels in general have treated badly to the artists. The only language that the disc labels understand is money..

Normally, the disc labels pay to the artists around aprox. only 9% (or least) of the cost of the total sale of a CD...

Another interesting point is that the disc labels attack directly to the music consumers with very high prices, bad sound (CDs bad mastered), menaces for the filesharing users..

God!! , Incredible lack of intelligence...

As consequence of these abuses some artists have considered to record with new independent disc labels, that offer more benefits than the traditional ones. This idea every day extends more...

I believe that the old and abusive model it is going to be supplanted (in a future, of course) by this new one.

Prediction: Eventually things will come full circle and DRM-free music will become a competitive marketing strategy

excuse me, but if the music is unrippable/uncrackable, where will the illegal copies come from?

The analog hole, Playing the music and recording the out put. Will analog technology
make a come back and kill off Digital media ?

Someone, i.e. the person you quoted, already mentioned analogue ripping methods, but I'll also mention the industry insiders that already release films/albums onto p2p before they are actually available to buy.

I think a valid question here is, can uncrackable DRM exist? I would require hardware protection, a la trusted computing initiatives.

The playstation and ps2's hardware protection was cracked via modchips. There are even stores at the mall in the US that will install modchips into your gamecube, ps2, and xbox allowing you to use them as media centers, emulators, and divx players. Is that illegal?

I bet once dvd audio is cracked (the real thing, not ac3/dts), all of a sudden people will start buying dvd audio at a ridiculous pace. People aren't going to go from usable, copyable, rippable cd's to a restricted, uncopyable format, that you can't really backup or bring anywhere with you, because you don't have a backup to abuse.
Before DVD Video there was no previous Digital Video standard, which is why people bought into DVD at a record pace, it's picture

Think Longhorn and Intel.

If this has to do with RIAA general, then good ;-D. Small time and indepedent artists will get their music heard (independent labels will rise from the ashes). My gripe with people in general is that I believe the RIAA is greedy and at their level 40 billion+ isn't enough pure egotistical nonense, but it's like biting the hand that feeds you all of the time . I have friends that go to me "ohh I can just go and download that on the internet" my reponse "well that's great, but I think music really looses it's true value in general. Your not really getting a piece of the artist if you are just dowloading their commercial release all of the time" For people it's not really about music it's just about commercial release for background listening in general and for RIAA it's all about the mullah so that's that haha.



It's a never ending battle. . Why the hell can't people just boycott the RIAA for once and bring them to their knees. That will show them. Screw you we don't need your artists to appreciate the music. It's a shame these labels are allowed to basically monopolize in this country and own rights to people creative input.



On both fronts . If labels charged an honest price like most independent ones do and a percentage of the profit went to the artist in general, while at the same time allowing people to download their artists material we wouldn't be in this situation.

No, it is no valid question. The whole idea of the thread was a fictional scenario which made it clear from the beginning on that it is unrealistic.

The question is not "IF" - the question is "WHAT IF?".

- Lyx

Already happened. Think allofmp3 - and just yesterday when browsing around i stumbled over an indie online music store which was in the buildup phase which marketted itself clearly and in bold letters that it will not use DRM and let you decide the format and quality. Others already did it(i.e. magnatune).

My prediction is that more and more indie labels will offer a DRM-free alternative, so that roughly it someday comes to a polarization:

indies(no-DRM) -----> <------ majors+some major indies(DRM)

The majors will have money on their side - while the indies will have common sense, technology, quality and time on their side.

Who takes the bet? ;)

- Lyx

Just a quick aside...I doubt DVD-A hasn't caught on because it's uncrackable. The people who intensely care about freedom to rip, DRM, etc, represent a small minority of music lisnteners, even if they are a majority on this board. I think it hasn't caught on because the CD-A format is good enough for now...people don't want to start buying in another format that it might become obsolete. CDs hold as much audio as you'd need for an album, are as good quality as anyone will need (in two channels at least), and widely available and hugely compatible with hardware. How many people are going to seek out DVD-As and then seek out a way to play them, when CDs are good enough for 99.5% of music listeners?

One well made DRM-free analog-copied version of a DRMed track is all that's needed... and let's be honest, it's likely to be easy to create a high-enough quality copy for joe-public's needs!

That's true, it's mostly audiophiles buying dvd-a.
It would require a number of unlikely things to happen. Joe Sixpack would have to buy a high-end stereo to get the benefits of dvd-a and a high-end soundcard just to play his ripped dvd-a's. Joe Sixpack would have to start using higher quality audio equipment on his stereo, and there would have to be a merger of the pc and the hifi stereo, that is cheaper than both separately. Then there needs to be an aac encoder that is universally transparent at a defined setting.

The reason the scenario you suggest probably won't happen for music or movies is that the people releasing movies and music are in competition with each other - so they find it very difficult to sit down together and work towards what you're assuming is their common goal.

When people like Sony make music, movies, and hardware that copies them, even the three different parts of the same organisation can't agree on what should be done.

Look at the format war over HD-DVD. Think about the (even bigger) cooperation needed to make world wide DRM systems work properly.


btw, if your "dream" scenario came about, then all new players would have watermark detectors in them - your analogue copies would be useless because they wouldn't play on any recent device. You might think that you'd choose to keep your old devices, but for how long is that really practical for most people? It's like telling a teenager now that they can't have an iPod for Christmas, but will have to make do with an old cassette walkman instead.


What would happen in your "dream" scenarios is an interesting question. Would major record companies charge us more to buy (or just listen) to music? Or would people abandon the product as being entirely unreasonable, and flock to indies? Or just give up listening to music at all?

I think you'd see some pricing models that were very advantageous for consumers short term - but the end to "buying" music for keeps, unless you want to pay a lot more.

Cheers,
David.

Thats a reasonable point. But that will probably not happen. Why? Because for it to work copying of media GENERALLY needs to be banned in all devices. It needs to be outlawed that you can record some noises in your house or phone-answering-machine and also playback them in whatever format you like. I doubt that would happen, so equipment *will* exist which can play unDRMed media. That in turn means that as soon as you strip the watermark or circumvent it someway, it will play in new devices.

I'm not knowledgeable enough in that department but when a player needs to detect watermarking and block playback if it isn't yours, doesn't that mean that we have the same vulnerability as with DRM? That the player needs to have the "key" and therefore the "key" is in the "attackers" hands?

But you raise a very interesting point: DRM doesn't work without watermarking because of the analogue-hole. If watermarking could be made "uncrackable", then that may indeed work.

- Lyx

Yeah, right... but only the half way. I think that SACD and DVD-A hasn't been cracked till today just because there is no real interest between the potential users... People who buy DVD-A or SACD are mostly audiophiles who aren't interested in any kind of lossy encoding, to say the truth, they consider it inferior. And they have enough money to saturate their hobby.

Everything can be cracked, there's no doubt about it... I think it's very hard - even nowadays - to invent some kind of "uncrackable" copy protection. And analog Macrovision was "cracked", too...

Hi Lyckx,
I 100% agree with you, I take the bet on your side.

In my opinion, if there will be a real need, everything WILL be cracked, DRM..whatever... The encryption, created by one smart human, can be reversed in the same way by another smart one

"Every watermark can be hacked and every watermark will be hacked, if there is sufficient motivation and opportunity." So says Christian Neubauer of Germany's Fraunhofer Institute ¬ leading developers of state-of-the-art watermarking technologies.
http://www.mpg.org.uk/newsletter3/water.htm

A watermarking being an inaudible signal added to the content, you are able to inaudibly retrieve it or blur it.
You can compare several versions in order to identify the common signal. A good psychoacoustic model should be able to erase it, being inaudible (if it’s audible, it’s another problem).
And finally, don’t forget, one copy cracked, and everybody is provided…

Also watermarking (and particularly complex watermarking) is expensive to encrypt for online distribution. It’s one of the reasons the majors don’t want to use it extensively. They use it for some pre-release (journalist, radio, …) though. Some journalist has been caught.

That's a typical argument from someone who hasn't cracked a watermark! :-)

If you read the reports from people who have cracked watermarks (e.g. in HACK SDMI), the reasons watermarks fail are quite different.

The reasons why ultimately they will all fail are simple, and quoted earlier: you've got to give the hacker the key to play the product.

Cheers,
David.

Here are two interesting pages:

Why DRM Sucks
Why DRM is Great !

Hi 2B,
You right, I never craked a watermark (and you?) and I'm pretty sure I won't need to do it
I did read the Felten report long time ago now. Now it doesn't mean that my statement is false.
The only statement I'm not sure about is the psychoacoustic model able to crack it.
Could you tell me why it won't work? I'd like to know.
Thx,
Cheers

[quote=marcan,Mar 30 2005, 11:27 AM]

[/quote]
Hi 2B,
You right, I never craked a watermark (and you?)
[/quote]

Of course not - that would be against the DMCA and could get me thrown into jail when I travel to the USA!

Seriously, I was given some watermarked content to play with at university (though I never did find out where it came from - my Professor was very secretive! However, thinking about it, it was probably just someone else's PhD project) - quite easy to detect and crack, though psychoacoustics wouldn't help. It wasn't a particularly good watermark, but it was still interesting to play with.

[quote]
I did read the Felten report long time ago now. Now it doesn't mean that my statement is false.
The only statement I'm not sure about is the psychoacoustic model able to crack it.
Could you tell me why it won't work? I'd like to know.

[/quote]

There are a lot of inaudible changes you could make to an audio file. Pick any "transparent" encoder from the HA list - they'll all do quite different things, all inaudible. Or just make a program to add noise right up to the masking threshold - you'd expect that to trash any inaudible watermark, wouldn't you - because it's destroyed all the inaudible space in the signal?

But it's also inaudible (though easily detectable) to speed up a track by, say, 0.1%. Just by listening, I can't hear the difference, so the change is (by definition) inaudible. If you run the little program I've just suggested, and add noise right up to the masking threshold, the track is still 0.1% fast, which is still detectable!

Obviously no watermark consists of just increasing the speed of a track by 0.1% (in fact that's a great way to break many primitive watermarks). But it illustrates how you can think you're trashing all inaudible content in the track, yet there's another inaudible property - another inaudible dimension if you like - that you haven't touched by your psychoacoustic based crack.

How many inaudible or near-inaudible dimensions do you think there are in audio? Unless you can think of them all, and totally trash them all, without affecting the audio quality, your blind "psychoacoustic hack" approach won't work.

Further more, the watermark might be periodic (in some domain - not necessarily time) - and the watermark encoder and detector know this periodicity, but you don't - that means they can average across this period, and uncover information that you would think had been buried beneath your inaudible noise/changes.


Contrast this with an intelligent hack, where you know something about the watermark - even just the domain it works in, or the periodicity (in the above example) or whatever. Then you can carry out a surgical attack. If you're good, you can actually restore the audio to the state it was in before being watermarked - in which case, it could sound better after the hack. You're never going to achieve that with blind psychoacoustic "trashing of inaudible parts".

Hope this makes sense - it's all quite apparent from the hack SDMI challenge responses - from the information in that paper, it's possible to simulate how the watermark might work, and to realise just how robust it is until you make a surgical attack, when it falls apart.

It's supposed to be security by obscurity, but that's never a very secure approach when the algorithm is patented, and therefore not obscure!


If the detectors weren't crackable, and the algorithm was totally novel and wasn't published anywhere (patent or journal), then it would be exceptionally difficult to crack. I'm not sure how that works as a business model though. There appears to be a school of thought to patent enough of the algorithm to stop anyone copying it, but not enough to help anyone to crack it.

The problem is, with that kind of secrecy, there's probably a huge flaw in the algorithm which no one involved has spotted. If there isn't, then the weak point becomes the inventors themselves - can they keep a secret?

However, who would deploy a system based only on the assurance of the inventor that it was uncrackable?

Cheers,
David.

Thanks 2B, interesting as always,

Now let's push a little bit further.
Say we make two different mixes of the same track.
We can easily identify each one by ear or thanks to a fingerprint technology.
It’s difficult to avoid the identification without mess up the content.
Therefore we do have here a mean to identify the owner of the mix if this one is unique for each person. And you can not avoid it without altering the content. In that sense it would be incrackable.
Now my question is what is the smallest variation you can introduce in the mix in order to be sure that it can not be craked?

Thanks 2B

Cheers,
Marcan

Very interesting thread. Thanks, Lyx.
From the perspective of one who's been on this board since 2001, and with 2009 only another four years off, I should point out how little major changes have taken place in the formatting of *most* cd's, and the design of most cd drives and players, in the past four years. Based on this, I seriously doubt copy-protected cd's will take off in a big way in the next four years.
DRM may be a different story, as the rise of online music-buying services and the huge growth of hard-drive-based portable audio players has created the opening for DRM to be used and improved. But when I think that Lame 3.90.3 has been around since 2001, it kind of puts the "all this technology is developing at such an amazingly rapid rate" in perspective, eh?