Potentially bad news for TiVo Series3 users, or anyone using a CableCARD receiver with Time Warner digital cable. Multichannel News is reporting that Time Warner plans to deploy Switched Digital Video (SDV) on 50% of its systems by the end of 2007 to expand their HDTV capacity.
SDV? What’s that? First, some background. (I’ll try to simplify it.)
Cable systems face a bandwidth crunch. They have a specific band of frequencies in which to carry the video signals, with limits the total number of channels that can be carried. In order to expand their lineups and add more digital and HD channels, they need to have free space or find a way around the issue. There are a few ways to free up capacity.
One of the most common is dropping analog channels and moving them to digital. Six SD digital channels can fit in the ‘space’ used by one analog channel. Moving analog channels to digital frees up a lot of room for new offerings. The problem is it can alienate existing customers who are using cable-ready tuning devices, like TVs, VCRs, and older TiVos, as it requires a cable set top box or a CableCARD tuner. Customer who are happily watching and recording those shows tend to get upset when you force them to rent a STB or buy new receivers.
Another method is simply to make more room. Different cable systems, with different generations of equipment, have upper frequency caps of 550MHz, 750MHz, or 860MHz. Moving from the lower limits to 860MHz, the general high-end today, increases capacity. It is also possible to move beyond 860MHz, up to 1GHz, for additional capacity. But this requires an infrastructure investment and possibly replacing older STBs in the field.
Other options would include moving from MPEG-2 compression to MPEG-4 for digital channels, as satellite has begun to do. However, this would obsolete most of the existing digital receivers on the market. Most digital cable STBs do not support MPEG-4, nor do most CableCARD receivers. (The Series3 does have MPEG-4 capable hardware.) Using a higher-density modulation system, such as QAM-1024 instead of today’s QAM-256, has similar issues. While these solutions may be used in the future, they’re too problematic today.
Switched Digital Video (SDV) is an end-run around the bandwidth problem. It works similarly to Video On Demand (VOD). With normal cable services all of the channels are pumped into your home over the cable all the time. If you have 100 channels, and you’re only watching one, 99 channels of data are being wasted. Of course, there are advantages to this. The same content is pumped into everyone’s home. The cable plant is relatively simple – pump the same signal into all the distribution lines, and just split it as needed into each home.
Well, things started changing with VOD services. To be able to offer a menu of content at any time you had to be able to send just that content to specific households. So STBs started becoming bi-directional devices, and frequencies were reserved in the system for the VOD signals. When you want to watch something, your STB sends a request up to the head end, which sends back a message telling your box which frequency to look at, and it begins sending your content to you. So, over time, the cable systems became more intelligent with more investment in the head end and smarter clients.
SDV plays off of the same technologies, only a bit simpler. Instead of sending all of the channels all of the time, systems using SDV tend to be hybrid. Some core offering of channels is sent using the traditional method. However, lesser used channels are held in the SDV tier. These channels will vary by market, but tend to be foreign language channels, nth tier cable networks (The Crochet Channel, Cooking with Yams!, you know the content… 
), etc. HD channels are also popular in SDV, because HD channels consume a lot of bandwidth.
A set of frequencies is set aside for SDV content, as with VOD. When you tune to an SDV channel, your STB sends a message to the head end requesting the channel. If someone else on your branch is already watching it, then your box is told to jump on board. If you’re the first, then it assigns a frequency and has your box tune to it. So only channels being actively watched are using bandwidth on the system. (There are some potential issues – such as what happens when you tune an SDV channel and leave the cable box on indefinitely or what if enough people on a specific branch try to tune different SDV channels, exceeding the available slots? There are ways to handle the former, such as activity timers. The latter cable companies try to handle with statistical projections to ensure it doesn’t happen.)
What it comes down to is that there are really three viable solutions for cable companies to pursue – migrating analog content to digital, expanding bandwidth, and/or SDV. And all of these are being pursued to varying degrees by the different cable MSOs. Some MSOs have already gone 100% digital, but that isn’t viable in areas with a high analog cable utilization. Others have expanding their frequency range, but that can be prohibitively expensive for systems with a lot of old receivers, etc.
As you can see, it isn’t a simple choice. There are many factors involved, and the best option may vary from region to region.
SDVs main drawback is the requirement for bidirectional communication and intelligence in the client. This isn’t a big deal for users with cable STBs, they already have this for VOD. However, it is a big deal for those using CableCARD – be it in a TV, PC, or a TiVo. While all CableCARDs are bidirectional (always have been), there are different kinds of host device. Basically all of today’s CableCARD hosts are Unidirectional Digital Cable Product (UDCP), including the TiVo Series3. These devices can use CableCARD as a decryption token to access digital cable, but they lack the upstream communication for VOD & SDV systems.
The problem is that, right now, there isn’t a standard way for 3rd parties, like TiVo, to produce bidirectional devices with CableCARD. There is a hardware specification, OpenCable Host 2.0, but the software side is still tied up in heated debate. Right now the cable industry still wants devices to support OCAP, so the cable MSO can push their own software and UI down to the device. So, say you want to order a VOD movie, instead of your TiVo interface you’d get an interface that looks like your cable box. The consumer electronics (CE) industry doesn’t like this, for obvious reasons. The CE side wants a standardized interface that they can call with their own front end; so each device can integrate and differentiate the services. For example, TiVo could integrate VOD into their Universal Swivel Search or the conventional Find Programs and it would be transparent. They can’t do that with the OCAP mandate. Clearly, I favor the CE industry approach. I think the cable MSOs should offer content as a service, and not try to make a grab for the UI.
This feud has tied up the CableCARD 2.0 specifications for years now, and it shows no sign of abating in the near future. The FCC may need to step in at some point to force the issue, as they did with the UDCP specification. However, for now, everything, including SDV, is still in limbo.
This leaves the consumer stuck in the middle. If your cable MSO decides to use SDV you need to decide if you’re going to use one of their STBs or forgo watching the SDV channels. Your TiVo Series3, CableCARD TV, the new Moxi box coming out this fall, etc – all of them cannot access SDV with what is available today. The good news is, so far Time Warner is the only major MSO to embrace SDV in a big way. The others have focused more on the analog migration and/or frequency expansion options.
Potentially better news came last month when Tom Rogers, TiVo’s CEO, testified before congress. He stated that TiVo, and others, are working with the cable industry toward a solution to the SDV problem with unidirectional devices. And a National Cable & Telecommunications Association official confirmed his statement. (See my past posts for more information.) A number of potential solutions have been discussed in the community, from not-yet-unveiled hidden bidirectional capabilities in the S3 (suspicion revolves around a FPGA chip on the board near the CableCARD slots), to a potential PCMCIA-style add-on that uses one of the CableCARD slots (presumably with an M-Card in the other to maintain dual-tuner use), to an external USB dongle that connects to the cable, to communication over the broadband connection. (I listed those in my personal least-to-most likely order.)
How much this impacts you depends on, firstly, if you live in an area that is, or is going to be, using SDV and, secondly, if any of the SDV channels interest you in the first place. For some this will be a non-issue, for others it will be a major factor in their entertainment choices. And there is the ray of hope from last month’s statements, and longer-term hope that the CC2.0 deadlock will be resolved.
I originally picked up the news from ZatzNotFunny, who picked it up from EngadgetHD, who picked it up from the initial Multichannel News article.
