The most destructive process for watermarks to sustain is file compression but many new watermarking technologies can survive everything from amplitude compression, bandpass filtering and D /A or A/D conversions to file compression and speed changes.

 

i doubt that but ok ...watermarks can be time stamps of an entire library of sounds ...in order to remove it you have to re-encode the whole library and make a new one replacing the original ...but you cant see this in a track, only in project ...

 

Watermarking, Vols. 1&2 by Mithun Das Gupta is a recent book (2012)

Here is one system:

Code:

http://www.bluespike.com/technology/giovanni/faq/

And another:

Code:

https://www.sit.fraunhofer.de/en/offers/projekte/digital-watermarking/

Code:

http://www.eetimes.com/electronics-news/4038234/Aris-and-Blue-Spike-pitch-their-rival-watermark-technologies

Got some more

Code:

http://www.petitcolas.net/fabien/publications/ieeespm00-evaluation.doc

Schemes:

Code:

http://lmgtfy.com/?q=digital+watermarking+schemes

(The let-me-google-that-for-you address is not meant as a snarky comment. It's just convenient to use.)

Piracy is creating all kinds of parasitic markets offering schemes to defeat it (for big money of course). I wonder if there’s a historical precedent for this type of behaviour in say the Middle Ages? It would be fun to know that.

 

@Feridan -- thanks for the additional links and research. I didn't take anything there as "snarky". Actually, I had originally googled "watermarking" but there's several million links that come up, and my hope was to find a few through people here that were relatively easy to understand if you're new to the subject. Both you and iSwingWood have given me a good start.

It's actually quite interesting, and far more problematic than I had first realized. What really fascinates me with it, is if information provided by watermarking would hold up in a legal environment. There are all sorts of issues involved when you have products communicating to other sources and people haven't been made aware of the intention of that communication. Even if you've just downloaded something, you still retain certain rights, sometimes to a surprising degree. I also have to wonder what use the watermarking will be if you have a large number of people sharing a sample library -- it gets back to the old thing, they can't start suing everyone. It should be interesting to see where all of this goes and how it winds up being used. It may turn out that its not worth the cost of the technology to the company's that are using the watermarking processes.

Anyway, thanks again for your help.

 

What do you mean? Wasn't this article about a man in Chile who was sued by CineSamples and ended up having to pay major damages to the tune of $370,059? I know it was a different scenario because he made a group buy with a few other people but still. And by the way some watermarking technologies are quite advanced so if you were to remove the watermark it would degrade the audio quality. I mean I'm not going to start analyzing and testing their algorithms (if I could even get my hands on them) to see if I could remove it as I have too much other important stuff to do but just the fact that they can find you is worrying enough, whether it happens or not.

 

I also don't believe in the "watermarking by sound".

So that they add some unaudible frequencies, for example.

Most sample libraries are in 44,1 kHz resolution, so there is not much of a frequency range to hide unaudible stuff.
(44,1 kHz equals 22,05 kHz in actual sound frequency... but as I understand, the frequencies above 20 kHz are not really used due to some issues with digital converters.)

Anyway... even if the sample library is in 96 kHz or higher, the end result (=published song) will most probably be in 44,1 kHz.
Or even far less, if some data compression (like mp3) is applied... which happens for most digital distribution, as you all know.

Add to all this all the editing/sound processing which is usually done to the raw samples... I seriously doubt there would be anything left from the "watermarking" after the mixing/mastering/sample rare conversion/data compression process...


But a way of watermarking I could imagine: they could hide an information inside the sample library, which makes it clear who the buyer was.
For instance, they could add the serial number of the product to the filename of ONE sample (or even just into the file details, thus hiding it even much better).
In a huge sample library with dozens of GB and several thousands of sample, that would be quite hard to notice.

Don't know if they do that, but that would be an option for "watermarking", I guess...

 

Really those people that don't believe in watermarking are operating on conjecture not fact. Who knows maybe you guys are right but I doubt that companies would be making such a significant investment in the technology if it didn't work. I have seen some of the algorithms and they're very complex but I haven't done further testing because honestly I have better things to do with my time. Rick Snoman actually came to the forums not too long ago and started firing off names of people that leaked his tutorials and mentioned that he utilizes watermarking and that it works quite well. If you're really interested then find someone that knows how to remove them and ask questions. That's the only way you're going to be able to make an informed decision.