"Cinavia"/Verance BluRay audio watermark protection

[Ghitulescu]

There will no be any technical solution to this.

The only workaround is to avoid the detection routine in the FW of the affected players.

The more-or-less-similar watermarking method of DVD-Audio is to day not cracked. Would it be someone kind enough to count the pending years? :) :)
:)
:)
Because this will be only solution to watermarking, the interesents have already devised a non-circumventing scheme, ie embedding this into the chips. The one-chip solution, ie no external routing/hackable signals, has already been envisaged, IIRC, in 1998. That would be probably the end of piracy.
Well, I rest my case :)

[rotty]

There will no be any technical solution to this.

The only workaround is to avoid the detection routine in the FW of the affected players.

The more-or-less-similar watermarking method of DVD-Audio is to day not cracked. Would it be someone kind enough to count the pending years? :) :)
:)
:)
Because this will be only solution to watermarking, the interesents have already devised a non-circumventing scheme, ie embedding this into the chips. The one-chip solution, ie no external routing/hackable signals, has already been envisaged, IIRC, in 1998. That would be probably the end of piracy.
Well, I rest my case :)

You may well be correct.

Slysoft's Peer made a statement some time back, saying that they intend to remove this stink.

Well havnt heard anything since then.

I may be possible to, if not remove it, disrupt it.

I would agree with you, that removal is very very difficult and even if we could, may leave us with an even worse effect to our audio.

DVD Video and BD is of course a massive market compared to the doomed DVD Audio, so is worth the effort.

Its interesting that all this comes with the UK news re copyright and right to backup. Dont know how that affects Cinavia.

Out of interest, have you avoided corrupted disks or have you purchased any.

[setarip_old]

@rotty Its interesting that all this comes with the UK news re copyright and right to backup. Dont know how that affects Cinavia.As you know, "Cinavia"/Verance protection in no way interferes with your ability to MAKE a backup copy - In its present incarnation, it "only" interferes with proper playback of the copy. So it would be my speculation that the UK situation will have no effect on further implementation of "Cinavia"/Verance protection...

[rotty]

@rotty As you know, "Cinavia"/Verance protection in no way interferes with your ability to MAKE a backup copy - In its present incarnation, it "only" interferes with proper playback of the copy...

I don’t believe you posted that.

What use is a backup that you cant play.

If that’s not interfering with your right to back up then I don’t know what is.

[setarip_old]

@rotty

That's the way the "game" has always been played.

But let's stay on topic ;>}

[Ghitulescu]

To summarise:
-it is possible to remove, in theory, the cinavia signalling on audio, by reversing the algorithms, which were published. The problem is how to identify the correct one (there are seven procedures, and of course, their combinations). And the reason why is not "possible" is also expalined in an older post of mine, it will severe damage the audio.
-the second option is to nvalidate/skip the detetction routins. I think this was done with PS3, but the battle was not yet won, again for reasons I explained in a former post.
-the most efficint method is to stop cinavai from being implemented. Quite hard now, as it's already mandatory in all BD domain.

Slightly off-topic clarifications: - as mentioned in one of my threads, the law in UK changed absolutely nothing (it brought UK to the level where Germany is, and this is still illegal).
- the requirements of the law that the end user should be able to use the movie s/he had legally bought were simply avoided by implementing the "Digital Copy" mockery. That definitivelly killed any your-own private copy right, since you have the official one. Since in the eyes of the law, there's no difference between an MP3 and a WAV/CD, or between a movie recorded with the cellphone in cinema and the master/BD/DVD issue of the same, it plays no role what type thereof is that digital copy.

[laserfan]

I rest my case
You are done. You can't post any more. Give it up. Please.:scared:

[Sharktooth]

-the most efficint method is to stop cinavai from being implemented. Quite hard now, as it's already mandatory in all BD domain.

No, it's easy, don't buy blurays. You should have never done that anyways...

[HWK]

No, it's easy, don't buy blurays. You should have never done that anyways...

Watermark is also included in DVD version. I will say don't buy players capable of detecting it.

[rotty]

Watermark is also included in DVD version. I will say don't buy players capable of detecting it.

Dont buy anything to do with Cinavia.

[dror.sharon5]

I'm far from an expert, so maybe my idea is so stupid I'll get laughed at, but here we go:
Why not edit the audio?
If the watermark is simply an inaudiable sound, that means it's simply a sound in a different freaquency. So why not just edit the audio in an audio editing program and remove any audio signle which is out side human hearing range?
I read somewhere that the watermark is compared to see that the movie is on it's original disc. If the watermark can't be found the movie should continue playing as usual, should it not?
Obviously such audio editing tools which will leave the audio intact are probably professional, heavy and cost A LOT of money, but still...

[Ghitulescu]

I'm far from an expert, so maybe my idea is so stupid I'll get laughed at, but here we go:
Why not edit the audio?
If the watermark is simply an inaudiable sound, that means it's simply a sound in a different freaquency. So why not just edit the audio in an audio editing program and remove any audio signle which is out side human hearing range?
I read somewhere that the watermark is compared to see that the movie is on it's original disc. If the watermark can't be found the movie should continue playing as usual, should it not?
Obviously such audio editing tools which will leave the audio intact are probably professional, heavy and cost A LOT of money, but still...

You should read the whole thread, unfortunately, because the answer to your ideas has already been given.

[dror.sharon5]

You should read the whole thread, unfortunately, because the answer to your ideas has already been given.

It's 56 pages, so... You know... THAT's not gonna happen...
Can I get the quick version?

[SamuriHL]

It's 56 pages, so... You know... THAT's not gonna happen...
Can I get the quick version?

You really want the quick version? Ok, if you insist. If you really thought you could just open up a compressed audio stream, strip out the signal, and save it, well, I don't know what to say. You honestly didn't think they'd have thought about that and perhaps made Cinavia slightly more "robust"? :) In short, what you propose is impossible. It's not some signal that they look for in the inaudible section of the audio. It's a transformation that's combined with the actual audio and is embedded like a tick. Removing it without knowing exactly what you're doing would destroy the audio. Plus it can't be done on the compressed track, so, you'd have to decode the DTS-HD/TrueHD/whatever audio track to LPCM before you even attempt it. Then you have to know exactly what you're looking for in order to reverse the transformation. And this is all just in theory. No one's even found out how they detect the signal yet, much less how to remove it. To complicate things, you're not just looking for one signal. Cinavia is extremely complicated and robust. It survives every kind of re-encoding imaginable. And there are several different types of signals that they can embed to instruct the player to do different things. In short, not a chance is it as simple as opening it up in an audio editor and removing all the inaudible "noise". LOL. This and many other seemingly "logical" ideas have been discussed to death. Every few months some ingenious type comes in and posts it as though no one's ever thought of it before. It's amusing, but, is definitely not going to lead to a solution. Far more has been discussed on this subject so if you do read through the thread you'll find a whole host of ideas that were thought of that don't work. Cinavia's not going anywhere any time soon. It's going to take a real reverse engineering effort from a company that's well funded and motivated before this thing dies. For now, concentrate on the detection part of the equation. A player that doesn't detect it = a player that can play Cinavia infected backups. That's about the only real solution we have right now. DVDFab can make backups that contain AACS encryption to fool Cinavia into believing they're a trusted source, so, the backup will play on a PS3. I don't personally consider this a good solution considering that your backups are A> protected, and B> still contain Cinavia meaning they could in theory look for a signal for "message 3" on a recordable media and flag that as untrusted. Boom, all those expensive backups are useless. This hasn't happened yet, but, it could. It just requires an update to the detection code in the player.

Anyway, I hope that answers your question. Many a person has brought up the obvious "easy" solutions before and none of them will ever amount to anything more than humorous fodder for discussion. Sorry.

[Andy o]

I thought the quick answer of "will it work?" would have been "nope".

[SamuriHL]

LOL. I was trying to head off the inevitable "but why?" response. :)

@seta, maybe it would be good to update the first post with links to the discussions in this thread for the "common creative solutions". I did something similar on SlySoft. It might cut down on these types of questions. I don't blame dror.sharon5 for not wanting to read through 56 pages just to find out where the idea has been discussed before. Just something to consider.

[k-c-ksum]

avoided cinavia discs to date. Is still the case that only the English audio track is infected.? Has anyone tried cutting the audio track into pieces of the english track and a foreign track so you have 1 track made up from both. you could have 5mins of the english track with a few mins of the foreign track then back to the english track until you have enough audio to test. Doesnt have to be perfect editing. Would be intersting to know if the clean audio segments is enough to stop cinavia warnings poping up.

[Eza17]

avoided cinavia discs to date. Is still the case that only the English audio track is infected.? Has anyone tried cutting the audio track into pieces of the english track and a foreign track so you have 1 track made up from both. you could have 5mins of the english track with a few mins of the foreign track then back to the english track until you have enough audio to test. Doesnt have to be perfect editing. Would be intersting to know if the clean audio segments is enough to stop cinavia warnings poping up.

Please see my earlier post #405. Cinavia is reliably detected within 4 seconds of audio so you can't simply remove it by changing parts. Remember the 20ish minute 'grace' period is arbitrary, cinavia could easily just kick in after a few seconds if they wanted it to.

[SamuriHL]

Please see my earlier post #405. Cinavia is reliably detected within 4 seconds of audio so you can't simply remove it by changing parts. Remember the 20ish minute 'grace' period is arbitrary, cinavia could easily just kick in after a few seconds if they wanted it to.

And don't forget, depending on the TYPE of Cinavia signal embedded, they do. Message 1 is darn near instantaneous from the reports I've seen. (Never seen that one personally). So yes, detection is very fast and they only do a few seconds to eliminate false positives.

[skwlsout]

Surefire cure for Cinavia,

http://www.skcneedle.com/uploadfiles/3.jpg

One in each ear, and then you can strip ALL the audio, cinavia included, then you enjoy all your media unencumbered by blasted watermarks.

[k-c-ksum]

Please see my earlier post #405. Cinavia is reliably detected within 4 seconds of audio so you can't simply remove it by changing parts. Remember the 20ish minute 'grace' period is arbitrary, cinavia could easily just kick in after a few seconds if they wanted it to.

they cant have it kick in after 4 seconds though even if they wanted to (well not on bluray backsups from a genuine source) :) thats why i suggested having an infected track segmented with clean audio

[Ghitulescu]

they cant have it kick in after 4 seconds though even if they wanted to (well not on bluray backsups from a genuine source) :) thats why i suggested having an infected track segmented with clean audio

Oh yes, technically, as someone before me said, cinavia is detected within split seconds. This is well documented, of course for someone caring to read the thread.

Why those 20 seconds?
If the movie starts, after the studio logos passed, hoopla! cinavia, now, that is FRUSTRATING.

[Guest]

@all

This thread is for discussion of Cinavia protection. OT posts have been removed. Please stay on topic to avoid strikes. Thank you.

[wastemoney2hd]

Hello all,

For my understanding :

1. When cinavia survives a recompression it is in the raw digital audio signal, there is no chance that it is in the compression itself. timeinfo, framecount...
2. If you want a mark (additional bits on some raw audio stream data) that does not have hearable effect, what are the options :
- make it not recognised by the digital to analog conversion
- insert it it a way that you do not hear it
are there other options ?
3. It is working on one (mono) raw stream, and does not use differences between multichannel audio

wastemoney2hd

[rotty]

Hello all,

For my understanding :

1. When cinavia survives a recompression it is in the raw digital audio signal, there is no chance that it is in the compression itself. timeinfo, framecount...
2. If you want a mark (additional bits on some raw audio stream data) that does not have hearable effect, what are the options :
- make it not recognised by the digital to analog conversion
- insert it it a way that you do not hear it
are there other options ?
3. It is working on one (mono) raw stream, and does not use differences between multichannel audio

wastemoney2hd


As far as I know, and I have not tried this, but I have read on threads that Cinavia survives microphone recording and therefore would NOT be at the digital level but at the analogue level.

Cinavia is hearable, but not necessarily noticeable to most, it modifies existing audio, but is of course an unwanted pollution of, and makes a joke of, so called HD Audio.

And no, it doesn’t use differential subtraction so it still works on mono track.

[wastemoney2hd]

hello rotty,

regarding :
As far as I know, and I have not tried this, but I have read on threads that Cinavia survives microphone recording and therefore would NOT be at the digital level but at the analogue level.

-> Did somebody verify this ???

regarding:
Cinavia is hearable, but not necessarily noticeable to most, it modifies existing audio, but is of course an unwanted pollution of, and makes a joke of, so called HD Audio.

-> 100% true. Somewhere in this thread there was a statement, that changing the pitch of the (analog??) signal for more than +/- 14% will make the watermark indetectable.
At least it has to survive 24p or NTSC to PAL conversations to work as intended.

And no, it doesn’t use differential subtraction so it still works on mono track. -> I did not see this info here. Did I miss it ?

Hopefully cinavia will NOT issue additional problems, as BD Format has enaugh problems to play ORIGINAL disks, mastering issues, manufacturing problems, player / firmware related and so on...

I have had more issues to play original BD Disks in 18 month compared to DVD in nearly 20 years. Most of the problems I could solve by firmware update, but finding a firmware that solves your problem and does not break the playability of other disk that worked before, is not acceptable.

wastemoney2hd

[rotty]

Hi wastemoney2hd

The fact that it was reproduced from mic recording was for Cinema priacy prevention.

As I said, I have not confirmed this and therefore can only pass on what I have read.

As for mono track, Peer at Slysoft had reported this about a year ago.

As for hearing the markings, therre are many many ways that the watermark is added, narrow frequency modulation as you have stated can be used, however, humans are very sensitive to pitch change like this, there are 12 semitones in an octave, 14% is over 1.5 semitones and sure could be heard even if for short period especially if you are aware. The length of the period (i.e. the shortest period this can be applied to the sound) is totally depedant on the frequency of the sound that you are modifying.

[Ghitulescu]

Oh, God....

Why don't you read the patents I posted long time ago....
Verance made this info public ...
Since cinavia is not found in the studio recordings, they could implement it in studio gear as well (for those having some thoughts), actually studio players from Pioneer already have this onboard. Since you're not producing anything (you're a consumer, not a producer), cinavia will hurt you not, you're supposed to play only originals. No law in any country I know of, alows you to copy a cinavied medium.
Game over. They are legally covered. Unless you, the people, change the law.

[setarip_old]

@Ghitulescu

Although you've yet again gone off topic with your statement that: you're supposed to play only originals. No law in any country I know of, alows you to copy a cinavied medium.
Game over. They are legally covered. Unless you, the people, change the law. Just so no one takes that as fact, I'll respond by suggesting to you that it's most likely there's no law in any country you know of that even MENTIONS "Cinavia" or Verance by name...

[Ghitulescu]

You missed the point, cinavia is part of an copy protection scheme, and as such any BD/DVD is illegal to be copied even if one actually can. Names play no role. If one breaches the law, then Verance will not sue him/her for breaching Cinavia, but for breaching eg DMCA in junction with one of the patents it has.
Cinavia cannot be technically removed. It may be "killed" but it leaves objectionable traces. One needs to know the exact procedure and to hope that the reverse path will give the original file. It can at most be circumvented by hacking the players, like it was the macrovision output of some DVDplayers (not all of them). Since the BD specs are on a permanent change, anything done today will be rendered useless tomorrow by FW upgrades or by making the players obsolete.

[asarian]

Cinavia cannot be technically removed. It may be "killed" but it leaves objectionable traces. One needs to know the exact procedure and to hope that the reverse path will give the original file.

There's no reverse path that will give you the original file: the Cinavia markings are mixed in with the original audio. It's not like paint in your hair that you can wash out. :)

Also, I maintain that Cinavia's Achilles Heel remains on the decoder side, where it has to determine whether the medium you're playing from is authorized to do so. I furthermore think trying to remove the entire Cinavia signature is kinda futile (and bound to fail; if only because you'd have the re-encode the original stream, at quality loss).

[Ghitulescu]

There's no reverse path that will give you the original file: the Cinavia markings are mixed in with the original audio. It's not like paint in your hair that you can wash out. :)

Also, I maintain that Cinavia's Achilles Heel remains on the decoder side, where it has to determine whether the medium you're playing from is authorized to do so. I furthermore think trying to remove the entire Cinavia signature is kinda futile (and bound to fail; if only because you'd have the re-encode the original stream, at quality loss).

This is EXACTLY what I said long time ago.
However, the thread is revived by newbies that are tired of reading its 1000+ posts and reask or resuggest the same thing over and over again.

I maintain that cinavia can be removed at the expense of audio quality. Like taking the paint spots from a T-Shirt with a scissors. Or like aggressive/automatic denoising of video or audio.

The attack must be performed at players' stage. Not on audio.

[Accident]

Although, personally I would take a slightly different approach. If you can detect, and eventually decode the cinavia message in the stream, and the cinavia message will most likely consist of sync, payload and checksum. Rather than removing it, I would aim to simply corrupting it, such that it has to be discarded as noise, or invalid. If cinavia can exist in the audio stream, without quality loss, then it can be very lightly modified, to be invalid, without quality loss.

[setarip_old]

@Accident

Hi! If cinavia can exist in the audio stream, without quality loss, then it can be very lightly modified, to be invalid, without quality loss.It would be great if you could perform and share with us a provable experiment, using a "Cinavia"-infested Blu-ray as your initial source material...

[HighLifeMan]

@Accident

Hi! It would be great if you could perform and share with us a provable experiment, using a "Cinavia"-infested Blu-ray as your initial source material...

I am going to do exactly this, or, at least, try to. The testing process will be quite time consuming because of the 20 minute thing and the very long audio processing times I'm anticipating. However, I'm pretty confident it can be done, but with how much audio quality loss, I don't know.

So I don't seem like some new guy that doesn't know what he's talking about jumping into the discussion: I read through this entire thread, the patents, and, something I haven't seen anyone else mention, the technical papers written by engineers at Verance.

For anyone that has an IEEE subscription, here is a paper I found very helpful, and that really shows how difficult this is going to be and how robust this watermark is:
http://ieeexplore.ieee.org/search/freesrchabstract.jsp?tp=&arnumber=4375954&queryText%3DAudio+Watermarking%26refinements%3D4294900078%26openedRefinements%3D*%26filter%3DAND%28AND%28NOT%284283010803%29%29%2CAND%28NOT%284283010803%29%29%29%26searchField%3DSearch+All
Rade Petrovic and Joseph Winograd have quite a few papers dealing with audio watermarks.

[Ghitulescu]

So I don't seem like some new guy that doesn't know what he's talking about jumping into the discussion: I read through this entire thread, the patents, and, something I haven't seen anyone else mention, the technical papers written by engineers at Verance.

For anyone that has an IEEE subscription, here is a paper I found very helpful, and that really shows how difficult this is going to be and how robust this watermark is:
http://ieeexplore.ieee.org/search/freesrchabstract.jsp?tp=&arnumber=4375954&queryText%3DAudio+Watermarking%26refinements%3D4294900078%26openedRefinements%3D*%26filter%3DAND%28AND%28NOT%284283010803%29%29%2CAND%28NOT%284283010803%29%29%29%26searchField%3DSearch+All
Rade Petrovic and Joseph Winograd have quite a few papers dealing with audio watermarks.
Of course they have :) they are the "inventors" of the patents Verance applied for.
The paper is a sort of dissimulated "marketing announcement". I wasn't really expecting they will say: hey, cinavia can be defeated and here's how we did it ;)
It states what should be by now obvious, that neither DVD-A nor cinavia have ever been defeated. This when both the encrypted and the original files were readily available. Imagine this will be several times harder when no original file is available ;)

The language used in the paper is less legal and more technical than that used in their patents. And they reissued the same thing I said before: the cumulation of several algorithms, to make the removal even more difficult, so the customers would object to the quality of the "decinavied" audio.

Under the fair use doctrine I will provide few sentences from the article:
Our optimism is based on analysis of past experience ........We also developed of a number of countermeasures that would either prevent known attacks (in 2007 nb), make them deteriorate the signal quality significantly, or prevent their repeatability.
...
In particular, multiple watermarking technologies should be deployed, and selection among them should be part of the (secret) stego key.
...
In order to prevent some attacks, or make other attacks more difficult, we propose a number of techniques that conceal watermark presence, location and properties.
...
Of course some of the goals are today only would-be-features, but they just started. IIRC, most of their patents have been already published after the paper and thus in force.

[asarian]

What's this (http://www.asus.com/Motherboards/Intel_Socket_1155/P8Z68M_PRO/#specifications) then?! It says "Blu-ray audio layer Content Protection". Are we seeing a Cinavia enabled mobo here?!

[setarip_old]

@asarian

Hi!

This appears to be a reference to HDCP:

http://www.digital-cp.com/files/documents/04A897FD-FEF1-0EEE-CDBB649127F79525/HDCP_deciphered_070808.pdf

[asarian]

@asarian

Hi!

This appears to be a reference to HDCP:

http://www.digital-cp.com/files/documents/04A897FD-FEF1-0EEE-CDBB649127F79525/HDCP_deciphered_070808.pdf

Oh good, thanks! They got me scared there for a moment. :) :thanks:

[rotty]

Hi I saw this link (below) posted some time ago.
The postee did not know if it was being kept up to date.
I looked today and it seems that it is being updated at regular intervals.

Perhaps the link may be worth posting on Twitter/Facebook etc if you have accounts.

http://en.wikipedia.org/wiki/Cinavia#List_of_Known_Releases_with_Cinavia_Watermarking

[setarip_old]

Is it just me or is "the silence deafening" from our friends at Verance regarding applications of "Cinavia"?

Perhaps "no news is good news"? (The list linked to by "rotty" is quite limited of late)

[SamuriHL]

Yea, it's a very bizarre situation to be sure. I really don't get what's happening with Cinavia lately. One would have thought it'd be adopted heavily by studios and player mfg's by now. Very strange.

[skwlsout]

I think there has been a shift in focus, away from technical measures targeting individuals, towards grander plans of control. Content holders are instead focused on lobbying governments to introduce site blocking legislation, with the idea that if they can choke of the supply of pirated material, then there business models will suddenly work again.

[Ghitulescu]

Is it just me or is "the silence deafening" from our friends at Verance regarding applications of "Cinavia"?

Perhaps "no news is good news"?

Yea, it's a very bizarre situation to be sure. I really don't get what's happening with Cinavia lately. One would have thought it'd be adopted heavily by studios and player mfg's by now. Very strange.

It's nothing strange in this. The implementation of cinavia is REQUIRED by the licence agreement to be SILENTLY slipped in into the FW upgrades, and these should be made irreversible. And cinavia is a mandatory part of the AACS which is mandatory for Blu-rays.

The cat with bells on the collar will never catch any mice.

I think there has been a shift in focus, away from technical measures targeting individuals, towards grander plans of control. Content holders are instead focused on lobbying governments to introduce site blocking legislation, with the idea that if they can choke of the supply of pirated material, then there business models will suddenly work again.

That will not stop the real hackers to do what they did also before the internet came ;)

[hello_hello]

It's nothing strange in this. The implementation of cinavia is REQUIRED by the licence agreement to be SILENTLY slipped in into the FW upgrades, and these should be made irreversible. And cinavia is a mandatory part of the AACS which is mandatory for Blu-rays.

Where did you get the "SILENTLY slipped" theory?
It's been "mandatory" for over two years. For something mandatory most manufacturers still seem to be happily ignoring it.
I assume the licence agreement must also stipulate a time frame for manufacturers to have silently slipped cinavia into all their firmware upgrades and current products. I guess as you appear to be privy to the details of the licence agreement you'll be able to inform us as to what that time frame is.

[Ghitulescu]

It's 2012, first of January. No device subjected to AACS licensing may be sold without cinavia. For products manufactured before that date, an individual agreement may be reached.

[setarip_old]

As I said previously, is it just me or is "the silence deafening" from our friends at Verance regarding applications of "Cinavia"?

It appears that, over the last few months, there have been less than a handful of new releases that contain "Cinavia"...

[Ghitulescu]

It appears that, over the last few months, there have been less than a handful of new releases that contain "Cinavia"...
Supposing one doesn't really want to buy all the releases and doesn't really want to upgrade a working player,

and given that it was not mandatory to put the logo on the back,

how can one positively identify those that contain cinavia, say in a shop?

[setarip_old]

how can one positively identify those that contain cinaviaBy going to the "usual suspects", dear soul - meaning those websites and forums that have been keeping track of this ;>}

[hello_hello]

It's 2012, first of January. No device subjected to AACS licensing may be sold without cinavia. For products manufactured before that date, an individual agreement may be reached.

How can the implementation of cinavia be required by the licence agreement to be silently slipped in into FW upgrades, while at the same time products sold before a certain date are subject to individual agreements?
It can't be both can it?