Are you looking at an open operating system to help you close your mind? Forget the technical issues involved, I personally think you are talking to the wrong community: watermarking is a means of restricting distribution of visual material. Linux is not about restricting distribution of anything. Perhaps you should raise these questions on one of the Apple mailing lists where you will meet equally paranoid minds. Bringing these issues up on this list is out of line. Regards, nick "we have to make sure the old choice [Windows] doesn't disappear”. Jim Wong, president of IT products, Acer

What is the relationship? These are not watermarks. The above are methods to prove or disprove that any given person was the source of a given message. A watermark is a means of marking a resource such that it may not be used by anybody else without degradation of quality. Linux does not need an audio watermark, the whole point of linux is about the benefits of sharing. If you want to have a way of restricting distribution of audio then perhaps go fish in some other seas. Regards, nick "we have to make sure the old choice [Windows] doesn't disappear”. Jim Wong, president of IT products, Acer > To: nickycopeland(a)hotmail.com > CC: linux-audio-user(a)lists.linuxaudio.org > Subject: Re: [LAU] Looking for Audio Watermarking Advice & Tools > Date: Sun, 23 Oct 2011 12:53:34 -0700 > From: kevinc(a)cosgroves.us > > > On 23 October 2011 at 21:44, Nick Copeland <nickycopeland(a)hotmail.com> wrote: > > > > I'm looking for a Linux solution to watermarking audio > > > > Are you looking at an open operating system to help you close your=20 > > mind? Forget the technical issues involved=2C I personally think you are=20 > > talking to the wrong community: watermarking is a means of restricting=20 > > distribution of visual material.=20 > > > > Linux is not about restricting distribution of anything. > > > > Perhaps you should raise these questions on one of the Apple mailing=20 > > lists where you will meet equally paranoid minds. Bringing these issues > > up on this list is out of line. > > > Rocks, hammers, knives and guns can all be used for good or bad. So > could watermarking. I'm only after the tool, and I'll return to > being on topic now. > > Thanks.... > > -- > Kevin > >

On 10/23/2011 02:53 PM, Kevin Cosgrove wrote: It's my understanding that "watermarks" are more than this. That you can add a watermark to a CD and that if it is played on a computer it has a phone-home feature so that Big Brother will know that an (un)authorised person has been given control of the CD. I never did research to see if these black helicopters are real or not, However, my brother-in-law in Nashville used to get to listen to pre-released albums... but since he wasn't the official reviewer, he needed to play the CD's only in audio CD players -- since playing it on a PC would get him and his source in Big Trouble. -gabriel

On Sunday 23 October 2011 23:50:08 Gabriel M. Beddingfield wrote: wrote: Thats what you get from playing sony-cds on a windows machine. Once again they forgot to support both mac and linux users with that "feature". And that was neither a watermark in the audio, nor special track on the cd, just a normal windows boot sector executing a stinking trojan and opening your machine up to code execution from sony (and everyone who hacked sony or knew how to use that insecure backdoor opened by sony on your machine). Have fun, forget about watermarks, use proper licenses and worry about your creativity and not about keeping your audience from admiring you, Arnold

On 10/23/2011 05:07 PM, Arnold Krille wrote: ...and neither was that the thing I was talking about. -gabriel

On Sun, Oct 23, 2011 at 2:50 PM, Gabriel M. Beddingfield <gabrbedd(a)gmail.com> wrote: For a limited number of albums, and usage on windows, this may be true: http://en.wikipedia.org/wiki/Sony_BMG_copy_protection_rootkit_scandal So just put it in an audio CD player with SPDIF out. (So you can copy the watermark bit-for-bit :-) ).... To clarify some obvious confusion in previous discussion.:Watermarking is not DRM. Watermarking is used to find the source of leaked copies of copyrighted material, or, potentially, for tracking "viral" material released into the wild. For example, watermarking was used to name and shame -- http://ninjatune.net/ie/article/2011/jul/18/ninja-tune-on-leaks -- a certain journo who got a prerelease copy of albums and shared it... -- Niels http://nielsmayer.com

On Sun, Oct 23, 2011 at 09:44:58PM +0200, Nick Copeland wrote: No it is not. All a watermark does it to identify the source of the data it is embedded in, and in such a way that can't be removed easily. That does not imply any use of it. And by the way, restricting distribution of some material is perfectly legitimate, it's up to the copyright holder to do that or not, and there is nothing morally or otherwise wrong with that if he/she does. The very idea that, if something is made available under some conditions that should imply it is free for all, is as stupid as 'either you're with us or against us'. Ciao, -- FA

On 23 October 2011 at 22:09, Jeremy Salwen <jeremysalwen(a)gmail.com> wrote: I'd been wondering if some sort of spread-spectrum code would be an applicable technology. -- Kevin

On Sun, Oct 23, 2011 at 10:09:44PM -0400, Jeremy Salwen wrote: Not always. For example, a spread-spectrum signal can co-exist with others if it uses a different spreading code. And even if the algorithm is know, the actual code used can be kept secret. See above. And there are *much* more sophisticated ways than those discussed in that page. There are three steps involved in adding a watermark to an audio file, and they are similar to those found in any digital telecom system: 1. Encryption, 2. Coding - inserting redundancy to allow error recovery, 3. Modulation - representing the digital data in audio form, and of course the inverse sequence to recover a watermark: 3. Demodulation - extracting features from the audio signal to produce 'soft bits' (bits which are not hard 0 or 1 but are represented by a probability function), 2. Decoding - combining soft bits to produce 'hard' ones, 1. Decryption. As long as these steps are separate, having the source code reveals all about 2 and 3, and that is usually enough to allow removing a watermark without introducing significant damage to the audio. In digital telecoms all 'advanced' systems combine 2 and 3 to some extent, mainly to improve both bandwidth and power efficiency. There's been a lot of research into such this, and the times when e.g. 16-QPSK + Viterbi + Reed-Solomon was 'state-of-the-art' are well past. What is required for a robust watermarking scheme is to mix in (1) as well, and one way is using cryptographically strong spreading codes. Ciao, -- FA

On Mon, Oct 24, 2011 at 08:50:00AM -0400, Al Thompson wrote: Hiding data in WAV files or similar is fairly simple. It's a bit more difficult in lossy formats but still not rocket science. But a watermark needs more than hiding: * It should survive a change of format. * It must not be easily removable unless you have privileged information. And that can be quite hard. Ciao, -- FA

On Mon, Oct 24, 2011 at 10:00 PM, Al Thompson <althompson58(a)gmail.com> wrote: Al, I have a huge amount of respect for your technical knowledge but offhand remarks like this require some countering in an online forum. Its true that low bit rate psycho-acoustic lossy compression does remove some "good" information. But my impression is that at "appropriate" bit rates, an overwhelming majority of humans cannot differentiate between an mp3 and the original recorded audio. I'm not entirely clear what "appropriate" means, but I believe that its hard for most people to differentiate (in a double blind test) at 128kbps, very hard at 192kbps and essentially impossible at rates above 256kbps. Should I update my understanding of this? i'm relatively skeptical of any claims for audio steganography that can survive arbitrary psycho-acoustic compression. in addition, random permutations of the least significant bits of a PCM encoding will almost certainly eliminate or at least reduce the confidence level associated with the presence or absence of the watermark), in a way that will be inaudible to more or less anyone.

On Mon, Oct 24, 2011 at 10:33:13PM -0400, Paul Davis wrote: It can still be done, using 'high level' features of the audio signal. For example, a lot of music contains regular rythmic patterns. Small timing errors can be used to insert a watermark. Even if the result sounds different when compared to the original, that doesn't mean you could tell which is which - the watermarked file can be 'as good' as the original. And in fact in many cases it's not even necessary to do this: the content watermarks itself. The probability that another rendering of the same music would reproduce the same error pattern is as good as zero. So if you find this pattern the file must be a copy. A variation on this is used to check the authenticity of some recordings in forensic audio. Almost all 'surveillance' audio contains some low level 50/60 Hz. If the recording is done on a digital device then its clock can be assumed to be more stable than the mains frequency, so any variations in this can be measured and compared against a database (which some forensic labs have been building up over the last years - it's possible because power grids are coupled over very large areas, e.g. almost all of western Europe). The result indicates when the recording was made. 10-15 minutes is enough for a very reliable timestamp in most cases. Watermarks are not inserted directly as 'digital' information, (as could be done for steganography) precisely for this reason. Ciao, -- FA

On Thu, Oct 27, 2011 at 07:40:36AM -0400, Monty Montgomery wrote: Yes, the long-term stability will be very good, as the frequency is adjusted to keep clocks in sync. But there are small short-term variations all the time which you won't notice on a clock, but can be measured easily. And if the lights were switched off ... :-) Ciao, -- FA

On Thu, Oct 27, 2011 at 11:47:07AM -0400, Monty Montgomery wrote: More or less, yes. But since the grid is not a zero-impedance connection of everyone to everyone, phase errors (and therefore small temporary frequency errors) can exists between distant areas on the same grid. See for example <http://fnetpublic.utk.edu/anglecontour.html> Ciao, -- FA

On Mon, Oct 24, 2011 at 10:33 PM, Paul Davis &lt;paul(a)linuxaudiosystems.com&gt; wrote: Ill take issue with that slightly. Namely that if you want to be that technical, you have to explicitly state the conditions of the test. Listening conditions probably make the largest difference to someone's ability to discern the difference between a lossy compressed file and the original. With good listening conditions, including a relatively quiet environment and an even halfway decent pair of passive isolation headphones, you will likely hear some difference if you are looking for it. Then you get into how much someone would notice, but I suspect until you get to 256 most people under such conditions(Which are not even ideal) would be able to hear if they are listening for it. Note that it is hard for me to judge this as I would wager a fair guess my ears are much better trained(As I would imagine most on this list falls into this as well) than what would be considered the 'average' listener, so some assumptions I make are just that, while I can keep in mind this fact it means I am not a particularly good judge at times of what others will or won't hear. Then again I work under the rule of 'a sound system is only as good as the sum of its weakest components' and as a result will do everything as good quality as I can in most cases assuming that even if people don't realize there is a difference consciously, they might subconsciously. If they are not listening for it however, aka casual listening, then we are in a different realm entirely. I still suspect that for many enthusiasts at least 192 would be the minimum for MP3s, with 128 being the minimum for more efficient codecs(ie. AAC and Vorbis), likely why those numbers can be found in the various online music stores that are popular today, at least in a semi decent listening environment. Now you get to an actual good listening environment, and especially with enthusiasts in such, and your requirements will likely increase. Seablade

Sorry, Nick, but there's a difference between watermarking an audio or image file (to identify authorship) and trying to restrict distribution. Besides, if you're really looking for paranoids about restricting digital content, forget Apple and go visit Sony. ;-) Nick Copeland wrote: -- David gnome(a)hawaii.rr.com authenticity, honesty, community

Kevin Cosgrove wrote: I used to use watermarking on some of my photos. There's really no way to watermark a photo in a way that can't be essentially negated via graphic manipulations. I think that if the software needed to verify the watermark is open source, anyone will know how to identify the elements of the file that comprise the watermark. Then they'll also know how to remove them. -- David gnome(a)hawaii.rr.com authenticity, honesty, community

On Sunday 23 October 2011, Kevin Cosgrove wrote: This reminds me when the SDMI (Secure Digital Music Initiative) [1] tried to use audio watermarks among other technologies trying to stop piracy. AFAIK, the initiative was not a success [2]. Regards, Pedro [1] http://en.wikipedia.org/wiki/Secure_Digital_Music_Initiative [2] http://cryptome.org/sdmi-attack.htm

On Mon, Oct 24, 2011 at 4:46 AM, Lorenzo Sutton &lt;lsutton(a)libero.it&gt; wrote: http://www.controlgeek.net/blog/2010/12/1/spectrogram-birthday.html While they didn't use the inaudible spectrum, that was more because they didn't need to. Seablade