A Word is Worth a Thousand Pictures - An Alternate View on Digital TV

I published this article back in 1999 and rediscovered it today, to a chuckle. It's still relevant as we continue to try and figure out the value proposition of datacasting and launch innovative ATSC 3.0 receivers like BitRouter's 3pi for the Raspberry Pi. See https://bitrouter.com/products/3pi/. Here you go, imagine it is 1999 as you read this...

We have been slaving over interactive and digital TV for years now. For one reason or another, we just cant get it to become as popular as analog TV or the Internet.

Several trial deployments of interactive TV found that in the US it would cost an average of $2,000 to wire up each house and yet peak take rates of movies were at $3.49 per movie. That is, you could charge no more per movie than a tape rental at Blockbuster.

Then, along came the Internet, suddenly 40 million people in the US alone found a reason to shell out at least $20 per month to have access just for service. Now they are spending a lot more on subscription and online shopping.

So, what is the clue here? In my opinion, the core issue is the cost to deliver each byte to the consumer versus its relative value. A standard definition digital movie takes around 2 GB of storage on an average and you can charge $3.49 to deliver it. That is around 0.1745 cents per megabyte. My online Wall Street Journal subscription costs $60 per year or 23 cents per day. Each day I read the journal I download around 2 megabytes of data. Let us say 5 megabytes to be safe. That is around 4.6 cents per megabytes. That is, a written word is worth a round 26 times more than a moving picture. And, I am willing to tolerate an asynchronous slow speed Internet!

The lesson, in my opinion, is that the real money in digital TV will be made sending INFORMATION down the high-speed pipe. Hence, the importance of datacasting, DSM-CC, ATSC T3/S13, DVB-data, ATSC T3/S16, ATVEF, DVB-J and ATSC-DASE.

DSM-CC is a standard that defines how interactive and broadcast data can be broadcast and retrieved. Broadcasts are done by using primitive techniques of sending data on a MPEG program ID or more complex methods like data and object carousels. Interactive data requires a back channel.

The ATSC T3/S13 sub-committee took the DSM-CC spec and defined descriptors and enhancements which define how terrestrial data broadcast will work in the US. The T3/S16 sub-committee made similar use of DSM-CC for interactive purposes.

The DVB Project took the same DSM-CC specifications and defined their own enhancements and modifications to define how broadcast and interactive data access will work in Europe. Thankfully, both ATSC and DVB standards in this area are remarkably close, being based on the same core DSM-CC standards.

ATVEF, DVB-J and ATSC-DASE (T3/S17) try to define how iterative applications should work on receivers. DVB-MHP (Multimedia Hope Platform) is the pre-cursor to DVB-J and is based on MHEG (Multimedia Hyperobjects Experts Group) plus Java. Both DVB-J and ATSC T3/S17 are based on the Java TV API work being done in collaboration with Sun. ATVEF is a Microsoft/Intel initiative to use XML technology instead of Java. All three standards accomplish the same thing.

ATSC PSIP (T3/S8), DVB-SI, DSM-CC, ATSC T3/S13, T3/S16 and DVB-data are all enablers for interactive applications using ATVEF, T3/S17 or DVB-J. In the end the success of DTV and profits from it will depend on whether it can deliver more than just moving pictures.