Sunday, July 8, 2007

Blue-skying anime ownership

AIR TV Prelude
This crackpot post was motivated by another industry panel where piracy was brought up. It's a discussion of what amounts to a business model that is currently not possible for a variety of reasons. From a content production view, it calls for anime licensing companies to integrate with anime licensor companies at an unprecedented level, an effective amalgamation in all but name. On the technical side, much of the technology already exists. What's lacking is the will. Brace yourselves, this is a bit of a head-scratcher, after these messages.

* * *

Problems with personal storage of any kind

The problems associated with delivering the so-called R1 release are numerous and ought to be well known by this time. First, a licensing company has to step up to the plate at all, and take on the risk of licensing a work. Following that, the script has to be translated, voice actors hired, recording studio time put in, post-processing and disc stamping, promoting the work in the target market, the list goes on. A frequent complaint among anime buyers is the giant time lag between licensing agreement and first volume release. Even then, there may be spelling errors in the subtitles, or the video quality may be inferior to their R2 counterparts.

A personal quibble for me is that DVD (english) subtitles are inevitably yellow and single-line. Maybe it's because I use MPC as my DVD player as I haven't tried others. Regardless, it's ugly.

No matter what, you've bought the product and you're stuck with whatever problems are present. Then there is the problem of storing them. For one thing, if you really want to go all out to support a series and get the collector's box, it takes up more space than just the regular box. For another, the discs may not last forever.

I am hesitant to store anime these days. I may be technically inclined, but that doesn't obviate the fact that I'm also lazy. Burning DVD's is time consuming and they may not last forever (we'll see). Hard drives, despite being cheap and plentiful as external drive solutions, fail and you're not about to try and recover that data. I'm also not desperate enough to construct a RAID system out of multiple drives to back up my anime. What I want, and I suspect a lot of other people would like this as well, is someone to store my stuff for me.

Outsource your storage!

A trend today is that we want to, at least physically, own less and less. Pictures? Put them on an image host. Who needs CD's when that same music can go on to a bit of storage that may or may not have a player with it? Some things we don't even want to keep on our own computers now. Like images, we put our e-mail, our calendars, our feed readers, etc. on servers.

In a roundabout way, I have just cast my vote in favour of anime streaming. I won't have to back up episodes. I don't have to use up existing hard drive space or obtain new drives. Better yet, streaming is flexible. If the player application supports soft subs, I won't have to see ugly yellow subtitles. Video and audio quality can be broadcast grade or better.

There are caveats, though. Obviously I'm going to have to pay money. The question is whether such a service is worth the money as-is. The answer is no. While the service is nice, I want to own the series and all the rights that go along with it. If I want to download it to watch on the airplane, I should be able to. In short, no DRM. If I decide that I want to sell the series, I should be able to. That one's a little trickier to deal with.

Content production

This topic is a bit easier to deal with. Companies with already established licenses for long running series (Bleach, Naruto, One Piece, etc.) would probably have a much easier go at this. To reduce the time lag between japanese release and, say, english release, translation has to start as soon as possible. The shortest time reasonable is only attained if translation commences as soon as the script is released.

Close proximity to the source is also beneficial. Put translators in the same time zone as the original script writers. If they can be put in the same building, on the same floor, even better.

Use soft subtitles. In addition to the flexibility gained by dictating position, colour, font, and special effects for text, they do not force the video to be re-encoded, preserving the original video quality. Any errors can be quickly corrected without a re-encode as well.

Like the current trend in software development, emphasis should be placed on putting out an initial product first and then adding features and even subtitle corrections. Subtitles are all that is required first, followed by dubbing at a later point, and then special music videos and the like. Any consumer who purchases the initial release should be entitled to these future updates and features.

Content protection

While there can't be any DRM, there can be digital watermarking. Watermarks will probably find most success embedded in the video content itself. They must not degrade video quality, and must be robust to attacks via video filters. Such developments are now in progress and may be ready for commercial use.

Content delivery

Any of these companies will either have to retain the services of a content distribution network like Akamai, Limelight Networks, or Google (lol?), or develop their own distribution network capable of providing high, sustained, bandwidth, like a much speedier incarnation of Stage6 if they want to stream HD resolution. The need for a one to one correspondence between watermarked video and owner necessitates the need for individual video streams. That is to say, multi-casting at the router node or using peer-to-peer networks, in their current stage, will not work.

Watermarks will be generated and inserted into the video content on the fly, so the host will not need as many copies of the video as there are owners. However, there is no escaping the need for an incredibly large pipe. Bandwidth must be dirt cheap, and to maintain the streaming experience, users should probably have downlink speeds at or better than DOCSIS 3.0 cable modem standards.

Fibre-optic would be nice, in other words.

The user may opt to download an episode and walk away, or stream a lower resolution version which can be generated on the fly using a hardware rescaler. Modifications to the existing TCP/IP protocol, or the development of a completely new standard, would be desirable in order to maximize the payload bandwidth.

Distribution costs could be decreased via the use of next-generation codecs such MPEG-4 AVC, packaged in a subtitle-friendly video container such as MKV. The processing demands on the user would increase, but tough luck. The world is moving to 720p and 1080i anyway, and will require either dedicated hardware such as graphics cards, or processors with SIMD/vector processing instruction sets.

Bridging ownership in the physical and digital worlds

I forget where I first saw this, but it must have been from one of the figure collectors such as TJ Han or bj0rn. Rather than support a series by buying the DVD's, they preferred to buy figurines. They're more compact than a DVD set, you can take photos of them, and they're probably marginally less fragile than a bunch of DVD's.

Why not implant a security token in each figurine?

Each figurine would contain a private key — a digital certificate. The token could also be implemented using a pseudo-random number generator with a secret seed, but a private key coupled with wireless USB, Bluetooth, or RFID would probably be more seamless. In conjunction with an account number or password, the user would use two-factor authentication to access an anime series.

An example scenario

Suppose I want to buy Code Geass (I'm just throwing out an example. Don't brick me if CG is not your cup of tea). Because the licensing company put their translators in the same room as the script writers when they were still writing the script(!!), an initial, english subtitle only, release for the first four episodes is available for streaming or downloading. I go to the purchase page, where a number of character figures, as well as Knightmare Frame models, are on sale. I select one and pay for it. In doing so, not only do I get the figure of my choice, I get access to all episodes and any bonus features once they are released.

A couple days later, my figure arrives in the mail. Using my account number or password or whatnot, and the figure (via RFID), I gain access to the streaming site and player. The video is mine. I can have the player save the video container file instead of deleting it once another episode is run or the player closes. If I so choose, I can back it up to hard drive, or put it on my portable video player.

What if I want to sell the series?

So you want to sell the series, or give it to someone else. The first step is to actually hand over the physical figurine. This might be a sore point for collectors, but hey, if you really didn't like the series you ought to reconsider collecting figures from that series in the first place.

The second step is to transfer access ownership. The buyer will have to have their own account. A possible method is to have the buyer provide you with their account information, a publicly exposed piece NOT used to authenticate. Through an update page, the seller can inform the company that ownership has changed to a different account.

Note that these two steps may be performed in any order. The seller would probably want to confirm receipt of funds before transferring ownership of either the access rights or the figurine. Escrow services and the like exist today to provide seller protection.

I want to share my anime / Thwarting the evil-doers

Share away. Just make sure your buddy doesn't get caught with your video, because that watermark can and will be traced back to you. Consumers will not be treated like criminals, but discovered criminal actions will have a trail. As noted before, the robust watermark will discourage piracy to some extent.

If you have not willingly lent out your copy of the video (i.e. your account and your figurine were stolen), the situation should be treated in the same way as if you've lost your credit card. Your existing account will be suspended and a new one issued. Unfortunately, you will likely be responsible for the cost of replacing your figurine, although maybe the company will give you a break and strike the cost associated with the access rights.


In this current legal (copyright) climate, one would have to be crazy to trust any company with their property, especially property that they'll likely pay in excess of $10 for. For this to work, the consumer must trust the company not to be evil and won't one day change the terms of the agreement to extract more money.

On the content provider side, they would have to accept that most of their customers are not crooks, and that they are willing to pay for convenience and timely delivery.

On the technical side, a lot more people have to be willing to make online purchases. The bandwidth must be there on both sides. Everyone needs contemporary systems to play the video. Companies must have a the necessary infrastructure to manufacture or order these figurines, and any manufacturer needs to be able to embed a digital certificate into their products. I think RFID would be the way to go, although that necessitates the use of an RFID reader. Current secure authentication RFID tags are expensive, although their cost should come down in line with insecure RFID tags.

I want my pencil board!

Who's to say that RFID tags, which are flexible, can't be embedded into the middle of a pencil board? But the point is well taken. With the disappearance of DVD's, physical bonus features will for the most part disappear. That is the disadvantage of jumping into a series early, as variety will be limited. A possible solution would be to sell non-RFID versions of those items at a reduced price. I don't think most companies would let you trade up to a newer item.


Large pipes and peering agreements. Advances in digital watermarking. Advances in strong authentication. Advances in RFID technology. Advances in video encoding, decoding, and production. Secure transaction frameworks and fraud mitigation systems. All of the components exist if even in a limited form, today. Technically, this system is possible. The real question is, can both the consumer and producer come to an understanding of what it will take to move into the next phase of media distribution?