Veiled Threat or Veiled Bride?

Veiled Video

Justin Thyme, industry insider and crackpot, provides anonymous analysis and commentary.

Think broadcast flags on steroids. That’s what the MPAA has in mind by embedding a digital information stream directly inside the images you watch on television. The motivation of the MPAA is to place the same sort of broadcast flag mechanism as CGMS-A in a location that requires no special procedures or equipment of the video distributors to enable. Here’s how we bypass this trivial protection and why such weak protections are our friend, not our enemy.

Veil Technologies
is mentioned as a specific technology provider to accomplish this task in a MPAA draft resolution being circulated. If the draft becomes law, then consumers will be confront Veiled video within 12 months of its passage. Making devices to bypass Veil could land you in jail for 5 years for a first offence.

From my casual reading of the patent on Veil II, the basic idea is that since video is typically broken into two interlaced images they will make one image slightly different than the paired image in luminance (brightness). The human eye blends the two images together so that we’re not aware of the change, but those variations are easily identifiable by machines. When you subtract the first frame from the second, they are left with the digital signal plus some noise as the image slightly changes. The trick is utilizing a noise cancelling algorithm so that only the digital stream is left. It is a very low bit-rate, much less than what the oldest modems could do but it is more than sufficient for broadcast flags, or text messages like coupons, advertiser internet addresses or filmography notes.

If my reading of the patent is correct, then it is likely that there is a set of current video devices that will defeat broadcast flags if transmitted in this form. Interleaved video at low resolution is undesirable to many users, and boxes that reinterleave using reverse telecine, especially if it blends the adjacent lines in a scaling operation, would obliterate the Veil information. These devices are not cheap, but the market for very low cost boxes using cheap video processing chips doesn’t yet exist. I am pretty confident you’ll see these kind of boxes sold for $24 on ebay for “upscaling” your video from HBO, or VOD services, which will come at the cost of your ability to get filmography metadata and broadcast flags directly from the images. Most consumers won’t be crying a river.

So why am I endorsing it this flagging scheme? There are many watermarking technologies that are much more robust than this proposal. It is preferable that the industry settle on a weak protection like this than a strong one. The studios should sell this as a way to embed metadata on content that cannot get accidentally separated during recording or retransmission multiple times. For example, a TiVo could extract this information directly and know for certain that the show being played is the show the user requested to be recorded. TiVo could also display filmography info directly from text in the show, or via URLs from the studios. The studio could market their tie-ins for direct DVD sales and upcoming theater releases of sequels, etc, etc.

Goodness and light for consumers and the studios. So wise up studios… Go ultra-light on the flags portion of this proposal. No need to be in your face about it, ease us in. Talk about the advantages, make some allies in the CE industry who endorse it (like TiVo), put in lots more restrictions when they are permitted to be used. If you decline and the public brands this as a sneaky smoked filled room maneuver of the evil grandpa suing MPAA to jam broadcasts flags down everyone’s throats, then it will carry a significant political risk for legislators where elections are won or lost by a few percentage points.