Log in

View Full Version : "Cinavia"/Verance BluRay audio watermark protection


Pages : 1 2 3 [4] 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Peer van Heuen
1st August 2010, 10:28
I read on fengtao's link that it survives pitch-shift.

Yet, I have read on SlySoft forum that pitch-shift defeats it.

http://forum.slysoft.com/showpost.php?p=273114&postcount=739

Bit of confusion there.

Simple: there's pitch shift and there's pitch shift.
You can speed up by 5% or you can speed up 10 times. Of course there's a difference.
All watermarking methods will stop working at a certain speed factor.
Usually the detectors are required to do additional checking for higher/lower speeds, typically the range up to +/-10% is specifically tested for.
In other words: the watermark is not robust against pitch shifting by nature, this robustness has to be achieved "manually" though the detector.

mike20021969
1st August 2010, 11:25
All watermarking methods will stop working at a certain speed factor.
Thanks.
That's what I was sort of wanting to ask. Of course, no-one would want the audio on a movie to sound like everyone was on Helium - or the opposite.
(I think it was asked on the SlySoft forum what the pitch was changed to when the watermark avoided detection).

SamuriHL
1st August 2010, 15:12
I read on fengtao's link that it survives pitch-shift.

Yet, I have read on SlySoft forum that pitch-shift defeats it.

http://forum.slysoft.com/showpost.php?p=273114&postcount=739

Bit of confusion there.

The quote you read on SlySoft was from me and was based on info posted earlier in this thread.

mike20021969
1st August 2010, 16:38
The quote you read on SlySoft was from me
I knew that :cool:
Didn't realise about the earlier mentioning (far too many pages/replies to trawl through).

SamuriHL
1st August 2010, 16:41
I knew that :cool:

Right, but, it was the part about the fact that I got that information from an earlier post in this thread that wasn't known apparently. :) See here:

http://forum.doom9.org/showpost.php?p=1421548&postcount=118

rotty
1st August 2010, 17:19
hi all, I assume Cinavia is likely a very low frequency signal. For those who have a copy and want to test it, run the audio through a steep high pass filter with corner frequency of 20 Hz or so - i would be very eager to learn if that has an effect

After hearing that f shift destroys the detection I thought the same as you. I too would be interested in notching f below 20Hz. and seeing the result.

The only thing that puts me off the idea that this would work is that as I understand it, this protection is supposed to work with camcorder copying in Cinema's, If that is the case, then it''s unlikely to to be the solution as most camcorder mic's would probably not even pick those frequencies up.

It is however possible that the system used on BD is not that same system.

It is worth trying, if only to eliminate, but not forever eh.

It is also worth considering 3rd to 5th harmonic modulation. i.e. differential between them / phase / AM.

mike20021969
1st August 2010, 17:59
it was the part about the fact that I got that information from an earlier post in this thread that wasn't known apparently. :) See here:

http://forum.doom9.org/showpost.php?p=1421548&postcount=118
Apologies.
I got my wires crossed. I misread and thought it was the 60+ page SlySoft thread, not this one (hence why I thought too many replies/pages to trawl though. Oops!)

I must get new reading specs :D

setarip_old
1st August 2010, 19:00
@rotty

Hi!It is however possible that the system used on BD is not that same system.Exactly.

As I said in my Post #146 to this thread: It will be interesting to see how many varieties of "Cinavi" eventually appear - and what tools are developed and how complex or simple they will be...

setarip_old
3rd August 2010, 22:47
Although I personally lack the requisite skills, from what I've read at various BluRay industry related websites and user forums such as Doom9, it seems to me that rather than waiting for a probable software player "Cinavia" "update", someone should currently focus on the first known PS3 firmware "Cinavia" "update".

On the other hand, further to the suggestion by the esteemed "neuron2", we've likely only seen a trial balloon, very basic iteration of what "Cinavia" can do. Therefore, it may be beneficial to the community if such initial efforts to overcome it were done in as much of a non-public way as possible.

mikeathome
6th August 2010, 19:57
If you have time, try convert it to a mono signal, so no phase difference between left and right. Need to do this on all 'stereo' channels (Main L+R, Surround L+R for 5.1, + surround back L+R if you have 7.1). Takes some time. If you insert center and LFE only it should not survive either.

I am pretty sure this will kill it, but will kill the audio impression too.
It would explain why it survives all manipulation because you do the same manipulations on both stereo channels and do not change the phase relationship between L and R.

If you embed a signal on L and R channel, with a 180-phase shift relationship it'll be inaudible but still be present and survive all manipulation. This is regardless of the frequency you use.

mike

mike20021969
6th August 2010, 20:40
If you have time, try convert it to a mono signal...I am pretty sure this will kill it
I don't think it will. It wont be that simple.

rotty
6th August 2010, 21:31
If you embed a signal on L and R channel, with a 180-phase shift relationship it'll be inaudible but still be present and survive all manipulation. This is regardless of the frequency you use.

mike

L R 180 out of phase is how old Dolby Pro Logic works, i.e. rear sound encoded in stereo signal. This info is then expaned (non linear) to produce rear mono (normally 2 speaker mono rear).

Peer van Heuen
9th August 2010, 15:08
If you have time, try convert it to a mono signal, so no phase difference between left and right. Need to do this on all 'stereo' channels (Main L+R, Surround L+R for 5.1, + surround back L+R if you have 7.1). Takes some time. If you insert center and LFE only it should not survive either.

I am pretty sure this will kill it, but will kill the audio impression too.
It would explain why it survives all manipulation because you do the same manipulations on both stereo channels and do not change the phase relationship between L and R.

If you embed a signal on L and R channel, with a 180-phase shift relationship it'll be inaudible but still be present and survive all manipulation. This is regardless of the frequency you use.

mike

Definitely not.
Cinavia can be detected in monophonic audio and also can be detected by looking at one channel of a multichannel sample.

rotty
11th August 2010, 14:43
Definitely not.
Cinavia can be detected in monophonic audio and also can be detected by looking at one channel of a multichannel sample.

Hi Peer

My question to you is not when or if, it is are you confident this can be removed from audio stream and removed seemlessly at that.

setarip_old
13th August 2010, 02:39
"Cinavia" was not present on a Sony BDP-N460 with Firmware version M02.R.123...

rotty
14th August 2010, 22:03
Hi Peer

My question to you is not when or if, it is are you confident this can be removed from audio stream and removed seemlessly at that.

The no response answers the question.

SamuriHL
14th August 2010, 22:54
The no response answers the question.

Or that he's not around at the moment. ;)

DrinkLyeAndDie
15th August 2010, 00:13
Or that he's not around at the moment. ;)

... or more importantly, why does he have to?

We know what they say about assumptions...

It's just a matter of time before someone finds some way around it.

setarip_old
15th August 2010, 00:45
@rotty

Hi! My question to you is not when or if, it is are you confident this can be removed from audio stream and removed seemlessly at that.

Although I too am desirous of hearing as much meaningful "Cinavia"-related information as possible, your post is confusing to me. Please clarify - What distinction are you trying to make regarding "if" versus "are you confident" this can be removed?

Are you trying to elicit a percentage/confidence level as a response?

BTW - (In addition to the valid remarks #167 and #168 posted here by two Slysoft Forum moderators) In light of the fact that "Peer" is an employee of Slysoft (unless something's changed recently), he may be limited as to what degree of detail he can discuss here regarding things such as "dealing with Cinavia"...

SamuriHL
15th August 2010, 01:18
I'm sure that when Peer has time, he can clarify more if he chooses to. Just because he didn't give an immediate reply does not mean you should jump to rash conclusions. He could be working on important stuff, or he could be lounging on the beach with rum punch for all we know. LOL! Just know that the absence of a response is not meaningful.

setarip_old
15th August 2010, 01:23
@SamuriHL

Perhaps you'd be good enough to make it clear (in your post #170) as to who you are suggesting is "jumping to rash conclusions" - and just which conclusions you believe are "rash"...

SamuriHL
15th August 2010, 01:27
I think it was fairly obvious that I was referring to rotty's "no response = no" comment. I think that's a pretty rash conclusion to jump to. We can't always be available to jump on any forum to answer a post immediately. :) That's all. I meant nothing more than that. I certainly wasn't referring to your comment at all.

Anyway, back to the subject at hand, I've still yet to see any CONFIRMED discs besides the US release of The Losers that contains Cinavia. I would have expected at least one or two more discs by now but maybe the theory that they were testing the market is correct and they're now getting ready to add it to all the good titles coming out this fall. /speculation :D

setarip_old
15th August 2010, 01:50
The following thread appears in the "Audio" subforum here:

http://forum.doom9.org/showthread.php?p=1425518#post1425518

Although the subject matter starts out pertaining to monaural DTS tracks, it broadens as the thread evolves (Read it all the way through) - and the knowledgable posters may coincidentally be providing some insight regarding "Cinavia" techniques.

It may, therefore, be useful to those posters to this "Cinavia" thread whose posts have made it clear that they have far more technical knowledge of audio than I do ;>}

Peer van Heuen
15th August 2010, 04:39
The no response answers the question.

a) I'm on vacation with limited access to the internet
b) I really don't feel well answering such questions as they are directed at my personal opinion and whatever answer I would be giving wouldn't weigh more than a mere guess or personal feeling.
At the same time, I'll be quoted having said xxx or yyy like I'd made a promise.

Anyway: so far we handled every single problem, so based on that you can do your own guessing at whether I "am confident" or not :)

jdobbs
15th August 2010, 05:43
a) I'm on vacation with limited access to the internet
b) I really don't feel well answering such questions as they are directed at my personal opinion and whatever answer I would be giving wouldn't weigh more than a mere guess or personal feeling.
At the same time, I'll be quoted having said xxx or yyy like I'd made a promise.

Anyway: so far we handled every single problem, so based on that you can do your own guessing at whether I "am confident" or not :) Well said, and for what it's worth you've got my confidence. Gee, and after you spend an inordinate amount of time in order to remove it like some would like you too, we can all sit around backing up "The Losers" over-and-over again in celebration.

Enjoy your vacation. This one can obviously wait til you get back. ;)

SamuriHL
15th August 2010, 15:01
a) I'm on vacation with limited access to the internet
b) I really don't feel well answering such questions as they are directed at my personal opinion and whatever answer I would be giving wouldn't weigh more than a mere guess or personal feeling.
At the same time, I'll be quoted having said xxx or yyy like I'd made a promise.

Anyway: so far we handled every single problem, so based on that you can do your own guessing at whether I "am confident" or not :)

LOL! Great answer, Peer. Now get back to the beach and some rum punch! :D

rotty
15th August 2010, 15:57
@rotty

Hi!

Although I too am desirous of hearing as much meaningful "Cinavia"-related information as possible, your post is confusing to me. Please clarify - What distinction are you trying to make regarding "if" versus "are you confident" this can be removed?

Are you trying to elicit a percentage/confidence level as a response?

BTW - (In addition to the valid remarks #167 and #168 posted here by two Slysoft Forum moderators) In light of the fact that "Peer" is an employee of Slysoft (unless something's changed recently), he may be limited as to what degree of detail he can discuss here regarding things such as "dealing with Cinavia"...

The "if" is because Slysoft may be able to beat Cinavia but for various reasons not want to release that tech. Therefore I was not interested in "if" or "when" it would be released, just if he was confident he could beat it.

mike20021969
15th August 2010, 16:07
just if he was confident he could beat it.
According to James at SlySoft, yes.
We're going to remove it.
Source: http://forum.slysoft.com/showpost.php?p=272342&postcount=624

DenkiWasabi
16th August 2010, 02:45
While the confidence from the slysoft team is definitely reassuring, but am rather curious what is the anticipated approach? Since it's been repeated here in several instances that this is not a case of encryption or decryption like BD+, AACS, etc... Did SlySoft gang attain audio or DSP engineering expertise to tackle the audio itself? Or will the solution most likely come from exposing future software players with this support? Really just want to know if we have the original HD audio to look forward to or will everyone be forced to a non HD audio rips.

setarip_old
16th August 2010, 06:03
@DenkiWasabi

Hi!

With such "Slysoft-specific" questions, you'd probably be better off posting them at that company's own forum...

rotty
16th August 2010, 17:03
a) I'm on vacation with limited access to the internet
b) I really don't feel well answering such questions as they are directed at my personal opinion and whatever answer I would be giving wouldn't weigh more than a mere guess or personal feeling.
At the same time, I'll be quoted having said xxx or yyy like I'd made a promise.

Anyway: so far we handled every single problem, so based on that you can do your own guessing at whether I "am confident" or not :)

Hi Peer

Thanks for your reply.

By the way, I was not after your personal opinion, I was interested if you and the team had looked at the problem and knew how to beat it.

Thats all I meant by the word "confident".

Here's wishing you all the best in that endevour.

Jeremy Duncan
18th August 2010, 01:30
"For those wanting and waiting for AnyDVD HD to remove this protection, consider what we have so far:

1. AACS. This encrypts just about all of the data on the disk. Without removing this, most of the data just looks random, and nothing can be done with it. Licensed players must remove this encryption in order to play the content. AnyDVD HD removes this encryption too.

2. BD+. This damages small parts of the data on the disk. Without removing this, most of the data is there to be played, but it is missing a few parts. When played back, there is visible damage to the video stream (and audio?). Licensed players remove this protection so that it can be played without this corruption. AnyDVD HD removes this protection too.

3. Cinavia Audio Watermark. This adds a supposedly unnoticeable echo to the decoded audio. Without removing this, the audio sounds just fine (assuming it is indeed inaudible). Licensed players DO NOT remove this watermark, but instead just look for it. If they find it, they act accordingly, halting playback or just muting audio. Since nothing out there is officially removing the watermark, it would be very hard for Slysoft or anyone else to come up with a way to remove it.

In the end, I agree that this is too big of a job for AnyDVD HD, and should be left to another tool. Instead of just replacing data on a sector by sector basis, removing this watermark will involve the complex task of analyzing large portions of the audio in its decoded form (which means a TrueHD and DTS-MA decoder is needed), removing the watermark from the decoded audio (if that is even possible without damaging it), and then re-encoding the audio again (which means a TrueHD and DTS-MA encoder is needed). Even if the just left the data as raw PCM, that would mean an even more complex re-authoring of the disc data would be needed. Not to mention, if you decode, alter, and re-encode a lossless codec there's no guarantee it will fit in the exact same space as the unaltered audio.

So the question is, will Slysoft attempt such a tool, can they succeed, and will they sell it as a separate product, or give it away for free? Only time will tell. For right now, the best thing to do is to simply avoid using Cinavia enabled players, and those looking to get into Blu Ray should carefully consider their playback device if they intend to play anything other than original discs." - Matt from Anydvd forum

If they make this "Cinavia" available to current software players then it will be broken? How? look in the method it's read and where a part that holds integrity and can break protection if it's broken: copy that part, clone that part.
In current form blu rays are cracked from copying the behavior of the memory, so that the hdcp thinks anydvd is that part of the memory and doesn't realise anydvd is only a Clone of that part of the HDCP system.

But if the Cinavia made tha HDCP HW read the audio then there is nothing inbetween the audio and hw to clone, no sw part needed just hw needed, the hw sees a error that breaks Cinavia and sends a flag to the part downstream of the HW checker and the error message pops up.

Since the media moguls are simple they won't know that once Cinavia is released to Powerdvd then Anydvd can immitate the part Cinavia depends on. :devil: Just wait, Cinavia will be broken the day it's released for Powerdvd/TMTP :)

Guest
18th August 2010, 01:40
Thank you for your incoherent rambling.

SamuriHL
18th August 2010, 01:44
Yea, I don't quite get it either. It's copied verbatim from a post on SlySoft forum and I disagreed with parts of it there, too. I can't really make any sense of it tbh.

Jeremy Duncan
18th August 2010, 02:32
I will repeat what I said in a more coherant manner, I was trying l33t speak but my age showed I didn't know how. :cool:

Here is a quote I found in the anydvd forum:

"They were relying on AACS being strong enough to withstand attempts at breaking it, they were wrong. IIRC a little flaw in a software player allowed the keys to be read from memory, and pandoras box was open."

link (http://forum.slysoft.com/showpost.php?p=271344&postcount=290)

"I don't think getting the watermark out would be feasible. I think the best bet right now is figuring out how the hell the player is detecting the watermark when bitstreaming. If it's doung a double duty of decoding + bitstreaming, that's freaking annoying and if implemented in software players would mean even more extra bloat (not that in hardware players is not bloat either). "

link (http://forum.slysoft.com/showpost.php?p=271821&postcount=439)

"There's a disc ID that can't be copied onto a backup that is used as part of the AACS decryption process. The only reason "protected backups" work is because AnyDVD has the title key for that particular disc. The disc *ALWAYS* gets decrypted and if applicable BD+ protection removed before the player gets it when dealing with backups. I think people get confused with the idea of a "protected backup" somehow magically working with all the protection in place. It simply means that AnyDVD is working its magic on the image in realtime instead of when you ripped it. To the player, a protected image is no different than a ripped, unprotected image. "

link (http://forum.slysoft.com/showpost.php?p=271890&postcount=480)

So this tells me that anydvd acts in realtime by - getting the key + using the key = anydvd acts as key holder in the aacs method.

AnydvdHD then is a form of mimicking, or cloning to impersonate a device to another device through the use of unauthorized id theft.
it stands to reason then, that since the aacs doesn't know that the hdcp id is stolen they will go through the software again in the microsoft operating system.

The cinavia protection is based on the use of a key that is read, and this key adds bloat, so the key will be put through memory that can be read by anydvd.
I think the key is not a single string but a mathermatical model based on folding the audio so the key matches when the fold is correctly in place, thats why it takes 20 minutes for the protection message, it'folding to the preset fold the cinavia is looking for.
This fold though is bloat and so it's put through the memory and even though its folded it still can be read if the design of the fold is met. Then the audio can be immitated and cinavia given a false fold key.

The thing is the error of id theft is through sw players using memory that can be robbed of their id. this error is where the fold can be found.

Guest
18th August 2010, 02:45
The cinavia protection is based on the use of a key that is read, and this key adds bloat, so the key will be put through memory that can be read by anydvd.
I think the key is not a single string but a mathermatical model based on folding the audio so the key matches when the fold is correctly in place, thats why it takes 20 minutes for the protection message, it'folding to the preset fold the cinavia is looking for.
This fold though is bloat and so it's put through the memory and even though its folded it still can be read if the design of the fold is met. Then the audio can be immitated and cinavia given a false fold key. It's still incoherent. But please, DON'T try again.

Jeremy Duncan
18th August 2010, 02:57
It's still incoherent. But please, DON'T try again.

I will try one more time then. Have some patience with me.

The key is hidden is the audio, the audio is mathematically treated like a video, and the video is treated with watermarks over a period of 20 minutes that when folded in a summary mathermatical design creates a key that the player reads and accepts as authentic.
That is why after a new video is put through the fold is already made and a new fold is easier to id as unauthentic because the part the fold should be folding is not present.

I will reword that in case it's too difficult, in point form as I know this is easier to read:
a.) the audio is converted to video using math, this is so the converted audio can be given visable watermarks.
b.) the visable watermarks are displayed over a period of 20 minutes and are incoherant, they must be remuxed. Once they are remuxed they add up to a string that is read by cinavia.
it is impossible to guess the math to remux them, it would take brute force calculations and each test would take 20 minutes for each string tested.
if the string is very long then it would become impossible to guess tring easily.
c.) once the string is made after being remuxed the unremuxed string is still there to id new video to see if it's being converted to a new unremuxed string, that's why the newer video is easily id's more fast than when the video was first played. it's looking for the hash that's being formed.
d.) it's no use looking for the audio mark, it's invisable, it's designed to be counted by abstract methods so strange even my mind can't grasp how they did it. the only way to get the hash is to get in the stored memory where it id's the new video being created and comparing it to the previously made string.

Guest
18th August 2010, 03:09
Thank you for your explanation. It is all perfectly clear to me now.

setarip_old
18th August 2010, 04:01
@Jeremy Duncan

Hi! In current form blu rays are cracked from copying the behavior of the memory, so that the hdcp thinks anydvd is that part of the memory and doesn't realise anydvd is only a Clone of that part of the HDCP system.Sounds nice, but I believe your logic is flawed because it's based on limited knowledge of what's available for decrypting BluRay discs:

While, as I understand it, AnyDVDHD and DVDFab Passkey use drivers to allow realtime playback of original BluRay discs (what you've described) on systems that otherwise would not allow it (e.g. firmware intentionally not updated, Internet connection intentionally avoided), MakeMKV, as well as AnyDVDHD and DVDFab Passkey, are each capable of actually decrypting and ripping fully functional full disc copies of Blu-Ray discs. These can then be burned to disc and/or used to create .ISO image files for playback - Discs for both standalone players and PC software players and .ISO image files mounted for PC software playback.

The burned discs and .ISO image files play properly because, just like a homebrewn Blu-Ray disc would appear, they no longer contain any copy protection - NOT because of anything that's been placed in memory or anything that is "part of the HDPC" system...

BTW - There is nothing, thusfar, to substantiate that 20 minutes is "required" for the "Cinavia" to do its work with Blu-Ray discs. So far, there has been only one Blu-Ray title ("The Losers") confirmed to incorporate "Cinavia" protection and, if I remember correctly, only one person to note that the "Cinavia" message appeared at approximately 20 minutes. For all we know, it could have just as well been set to appear within one SECOND, as it purportedly does with "Cams"...

It would be great if you'd purchase "The Losers" BluRay disc and tried applying your theory to it - and then reported your results back here. Quite honestly, although I follow the gist of your postings, much of what you say is beyond my ken.

Ghitulescu
18th August 2010, 10:13
Cinavia is an interesting issue apparently. The way they intend to DRM the copyrighted work was proposed some 20 years ago by Philips and it was rejected ;)

Some interesting passages from Verance

However, one clear disadvantage of such systems is that the operator provides each consumer with a high quality (typically digital) master copy of the content.
The most viable marking solution embeds information into the content, thereby reducing the likelihood of alteration or removal of the marking information.
Digital watermarks are designed to carry auxiliary information without substantially affecting fidelity of the host signal, or without interfering with normal usage of the host signal.
Unfortunately, however, for the unscrupulous consumer (e.g., pirate), this provides a means to generate illegal copies of content with little effort. Letting the intended recipient know that this embedded information exists may help deter a potential pirate from making illegal use of the content. It will also allow enforcement agencies to track the source of many copies.
While access to an encrypted content may be limited to entities with proper authorization and decryption keys, once a content is decrypted, it may be readily copied and disseminated. This is particularly true for multimedia content that must inevitably be converted to audio and/or visual signals (e.g., analog format) in order to reach an audience. Watermarks are particularly well suited to plug this so-called 'analog hole'.

I find it worrying that the industry considers that is an disadvantage for them to provide us with high quality works (on which we pay a premium price). This is why they changed the "definition of quality" to match the format and not the content, ie the audio quality of something being defined by being in 192kHz/8 channels (a telephone conversation becomes this way a high definition audio if recorded in 192 kHz) or the video quality of a video being FullHD , 1080p (even if it comes from a webcam). The reason being most probably this. No more quality content, but what a nice package.

Even worrier is the corolar, that they believe that the copyright will be extended to last forever.

Guest
18th August 2010, 19:21
OT post removed.

This thread is for discussion of Cinavia, not general discussions of copyright and DRM.

setarip_old
27th August 2010, 22:12
For an interesting read about the apparent underlying (pending) patents regarding "Cinavia", click on the following primary link and then the links in the post by "John Doe":

http://forum.slysoft.com/showpost.php?p=276015&postcount=857

keysersose
31st August 2010, 20:19
I have made a BD-5 ISO file of the Losers DTS-MA track with a blank Video - it's about 3.1 GB and can be burned onto a DVD-R..

I sent it to Hotfile.com - if anyone wants to download it in order to test for Cinavia on their player please shoot me an email.

I would appreciate it if someone that does have a Cinavia player could download it and confirm the disc does trigger the Cinavia..

mrkeysersose@gmail.com

rotty
31st August 2010, 21:10
I have made a BD-5 ISO file of the Losers DTS-MA track with a blank Video - it's about 3.1 GB and can be burned onto a DVD-R..

I sent it to Hotfile.com - if anyone wants to download it in order to test for Cinavia on their player please shoot me an email.

I would appreciate it if someone that does have a Cinavia player could download it and confirm the disc does trigger the Cinavia..

mrkeysersose@gmail.com

Its the re encoding of the audio thats going to be the bummer with this.

If cracked, it will probably have to be detected/removed at the analogue level and re encoded.

We will probably have to make do with either AC3 or LPCM audio as a replacement. (no public DTS MA/HD encoders).

Now I am saying probably, but its worth keeping in mind.

Jeremy Duncan
1st September 2010, 06:03
I will tell you how it works, I think:
- The audio has marks in it that are creating a id tag.
- If the audio has the id tag it is authenticated over a period of time, some say 20 minutes for audio that has the code in it.
- If the audio tag has been tagged and is being verified then the tag stream is cut off there is nothing to authenticate, then the cinavia flag goes up.
So if you took audio that took 20 minutes before to get the cinavia flag, then took 5 minutes of thataudio then mixed in 15 minutes of audio with no cinavia, the flag will go up faster than 20 minutes.
Or, if you took a stream that has cinavia but the rest of the audio video stream is not the same quality as the original the audio is tagged but the authentication see's the code is altered from the original so the flag goes up earlier, say a minute or so.

- How to beat cinavia, possibly, or theoretically.
You identify it can cut off streams it finds as unpure compared to the original.
Then, you see how much stream it takes to qualify as unpure compared to the original (1 minutes, 10 minutes, depending on the quality of the ripped material).
Then, you use the stream it has not verified as unpure, find it in the machines memory, and IF the memory is authenticating the stream, see what is authenticating it - if it's a key or what.
That's why using the SW to find the memory then to find the key in the memory is vital to cracking cinavia. Luckily that's how aacs was cracked so if they did it before then they can do it again once Powerdvd or Totalmediatheater updates to use Cinavia.


The hope is once the key is found it can be use to replicate the authentication then mux it into the audio without sacrificing quality so the cinavia finds and and accepts it since the foundation is based in memory, the memory has no conscience and only accepts the physical data it recieves.
But if the audio goes straight to hw then the cinavia is found then passed on to memory, then there is no straight path the the memory and then, well it maight be to strong to crack. :confused:

Ghitulescu
1st September 2010, 12:08
I will tell you how it works, I think:

You don't have to think nor to imagine things, the cinavia watermarking is documented in their patents, which are free and far from being too many ;). BTW However, one clear disadvantage of such systems is that the operator provides each consumer with a high quality (typically digital) master copy of the content.this is an excerpt from one of their patents.

One cannot remove the watermark without damaging the audio, it works like a logo-remover tool.

There are 2 solutions against cinavia:
1. obtain a player that doesn't obey cinavia and resist hard to upgrade it
2. file a class suit (in the US) or petition the EU parliament against cinavia and its use since it unduly restricts the fair use rights

k-c-ksum
1st September 2010, 18:01
has anyone cut the first 20mins out and remuxed to see if cinavia kicks in straight away? Maybe the first 20 mins doesnt have the watermark and it was only embeded from 20mins onwards?

rotty
1st September 2010, 18:42
I dont think many people realise how much this will impact us all.

This is soon going to be the only thing people are talking about on here.

Just keep a mental note of where we are on the time line for the first batch of Cinavia disks (Late Summer 2010).

Ghitulescu touched on fair rights and he's so right, if anybody thinks there is going to be a quick fix for this then I think they are being very optomistic.

For a fix that also keeps the original audio stream type, then, well......

I will not buy a Cinavia Disk. As more and more disks appear then this may well mean I dont buy anymore Blu Ray disks, well so be it.

SamuriHL
1st September 2010, 18:58
That entirely depends on how you consume your media. People need to remember that this is done by the player. The player MUST support Cinavia detection and then ACT upon it. There is no other mechanism for this to work. I can assure you the open source players that do not take their marching orders from the AACS LA will not be implementing Cinavia. :D Current versions of the big 3 software players do not support Cinavia detection, either. So, back to my original statement...if you consume your media on an HTPC, then there is nothing to see here. Is it likely that future versions of the big 3 will add Cinavia detection? Probably. Until we reach that point, keep a backup of your most recent player. Restore it when you need to play a Cinavia infested backup. It's the hardware players and streaming devices that are going to see the biggest impact. That is, once those devices start supporting it, once discs start using it, and you wish to play backups on those devices of discs that contain it.

Ghitulescu
1st September 2010, 19:14
1. obtain a player that doesn't obey cinavia and resist hard to upgrade it

I can assure you the open source players that do not take their marching orders from the AACS LA will not be implementing Cinavia. :D Current versions of the big 3 software players do not support Cinavia detection, either. So, back to my original statement...if you consume your media on an HTPC, then there is nothing to see here. Is it likely that future versions of the big 3 will add Cinavia detection? Probably. Until we reach that point, keep a backup of your most recent player. Restore it when you need to play a Cinavia infested backup. It's the hardware players and streaming devices that are going to see the biggest impact. That is, once those devices start supporting it, once discs start using it, and you wish to play backups on those devices of discs that contain it.
That is solution one. :)

But, they have a winning card: time.

The most efficient solution is the second one. Because it kills the winning card....
2. file a class suit (in the US) or petition the EU parliament against cinavia and its use since it unduly restricts the fair use rights

It's time to bill (in the States) for the 28th Amendment (much in the sense of the 6th const. art.) for a separation of the state from the media content producers ;) like it is supposed to be between state and church.