The view from my side, as both a broadcaster and a consumer of both broadcast and 'webcast' content:
>From what I've been led to understand in my time in broadcast*, the decision wasn't made because of power costs but because they (the Grand Alliance and the FCC) believed that 8VSB would work better over the US both due to terrain and propagation differences and due to our markets of television transmitters and need to place co-channel (same frequency) transmitters relatively nearby with minimal interference. Some argue that (C)OFDM would have been a better choice even taking those things into account. Which should have been chosen based on what criteria is a long discussion in and of itself but we've got 8VSB (at least for now). As for the mobile thing, I don't know if anyone had thought that far ahead. It seems silly now since I had a 2" rear-projection portable NTSC TV that was made in the mid-to-late 1980s (Sony Watchman) and now they are trying to 'solve' the mobile delivery 'problem'. ATSC M/H adds a lot of error correction as well as 'training data' to make the M/H stream 'easy' to detect and decode, even with Doppler and multipath effects (see ATSC A/153 part 2 for detail) - this reduces the data rate available for the 'main' (A/53) data significantly and does not create nearly as much available for the M/H (A/153) data. This sounds even sillier when you realize that ATSC standard A/49, first published in 1993, was a "Ghost Canceling Reference Signal for NTSC" (ghosting on NTSC is a symptom of multipath). As for 'webcasting' replacing RF broadcasting, I think we're a ways out if it will even ever happen in the general case yet alone every case. RF broadcast is very efficient, as previously mentioned. As a broadcaster, I push 19.393 Mbps of content 'into the air' for everyone around to receive at once. As a consumer, I have four tuners attached to what amounts to a few pieces of wire and I can receive roughly 80 Mbit of non-blocking data 'through the air'. My ISP provides me a downlink speed of roughly 10 Mbps. If I were in a larger market, I'd be able to receive even more data (non-blocking with more tuners or blocking from my POV if I didn't have enough tuners for every channel); even if the ISP provided downlink speeds scaled up similarly, there would be much more data available 'through the air'. The people who live near me have the opportunity to receive that same 80 Mbit of data without any transit costs. Can CDNs replace some of what is now broadcast? Likely. By reducing data rates (with better compression technologies as well as simply compressing more) and providing content that viewers want, they could (Netflix and others are already doing this with some content). There is, of course, a lingering societal question about "shared viewing experiences" for shows having set delivery schedules by broadcast. Live content and local content, however, will still be (in my opinion) best served by RF broadcast for some time to come due to both the inherent efficiencies in the system and the ease of localization for end-users. * The ATSC Digital Television Standard (A/53) was developed, documented, and formalized from the late 1980s through the mid 1990s (A/53 Part 1, Annex A describes the history). I wasn't working in television until 2000 or so and I wasn't doing television broadcast-related work until 2008. - Eric Eric Adler Broadcast Engineer
