Secret Watermarking Standard Congress May Require Sucks By Its Own Measures
from the doesn't-look-good dept
No wonder the company that makes the VEIL watermarking technology that's included in the "analog hole" bill don't want anyone to know the details of the technology. Even by its own measures, that watermarking technology sucks. Following Ed Felten pointing out the problematic nature of the secrecy around the offering, the company offered to release an executive summary its own test results to show how good the technology is. Beyond the obvious bias (and the fact that it's only an exec summary without any real details), there's an even bigger problem here. Ed Felten looks through the exec summary and notes that, even when the tests are set by the company, the watermarking technology appears to be awful. First, one of the keys to any watermarking technology is that it shouldn't mess with the content in a noticeable way. However, 29% of people did notice a difference in the watermarked versions. Even more important is that there are, clearly, ways to remove the watermark. While they won't say how, even the company admits that one method was successful 58% of the time -- and, as Felten points out, these are tests the company did themselves, which tend to downplay the real vulnerabilities. In other words, a big part of the reason why the company doesn't seem to want to detail the spec (that would be required by law if the bill passes) is because even it knows that it sucks.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Reader Comments
Subscribe: RSS
View by: Time | Thread
Better be careful....
[ link to this | view in thread ]
No Subject Given
CHICAGO, Feb. 1 (UPI) -- The surviving Baby Bells -- Verizon, Bell South and AT&T/SBC -- have disclosed that they may someday charge new fees to digital businesses, sites like Google and Yahoo!, that generate substantial traffic on the Internet. The explicit rationale? These firms are taking up too much bandwidth. But telecom experts tell United Press International's The Web that they are worried that such a "toll road" could take a toll on the future growth of the Internet.
"The mere mention of the words 'toll road' sound like government regulation is right behind," Chris Consorte, president and chief executive officer of Integrated Direct LLC, an interactive online ad agency based in New York, told The Web. "The minute we're talking about a bandwidth fee is the minute entrepreneurs begin to second-think great ideas and developing their businesses." By Gene Koprowski
[ link to this | view in thread ]
Re: No Subject Given
[ link to this | view in thread ]
Distortion yet again
This article is a complete distortion of the the tests presented in the cited summary.
To address the your comments:
The people that screened the broadcasts said they could tell the difference between the watermarked and non-watermarked signal 29 percent of the time, but the results of the test show a very different answer. They correctly identified the watermarked signal 49.9 percent of the time. Exactly what you would expect if there was no difference. There were two alternatives and they guessed right half the time.
The other test was not a test to see if the signal could be removed, it was a test to see if distortion to the signal would make it impossible for the decoder to recognize the watermark. The results were that some types of distortion did make it difficult to recognize. The document doesn't say whether the distortion rendered the signal completely unusable or not. I find it quite telling that you failed to comment on the fact that the system's false postive rate was greater than the standards set by the broadcasters and the consumer electronics industry.
There are a number of valid reasons why this technology (and the document for that matter) should be viewed with suspicion, but this article doesn't address any of them. If you want to be taken seriously, then you need to make rational arguments and not distort your sources to make Chicken Little arguments.
[ link to this | view in thread ]
Re: Distortion yet again
This says to me that perhaps that 29% swing is gathered partly from the difference between 50% and 33%.
I see more distortion in your rebuttal than the original argument. I am highly skeptical this technology would live up to what its expectations.
I like mike's summary. he included links which led to the articles (and in one were the test results) upon which he drew his conclusion.
.mrc.
[ link to this | view in thread ]
Re: Distortion yet again
[ link to this | view in thread ]