It's become clear that the right receiving antennas are critical for successful reception of DTV.
The FCC has set up a Web page to assist viewers with reception problems—www.dtv.gov/fixreception.html.
This page provides many useful suggestions on how to adjust an antenna for reliable DTV reception, along with links to additional information. One of the links is to a new FCC Web page, DTV Reception Maps
If you enter a street address or ZIP code, after some delay, you’ll get a list of stations, sorted by received signal level in dBm. The list is color-coded to show strong (35 dB above threshold), moderate (15 dB above threshold), weak (noise limited threshold), or no signal (below the noise limited threshold) conditions. The Individual Location Longley-Rice (ILLR) propagation model is used and interference isn't considered. Terrain extraction is at 1 km intervals. (Thanks to consulting engineer Charles Cooper for this detailed information on the workings behind the page.)
Clicking on a station’s call letters provides information on its DTV RF channel, predicted receive power level, and the direction to the tower. A link also provides a new map showing the expected change in coverage from analog to digital.
I gave this a try using a location where I've tested ATSC USB tuners in the past—the intersection of the Santa Monica and San Diego freeways in West Los Angeles.
My results highlighted one of the problems with the FCC antenna database: It’s based on horizontal plane azimuth patterns, which are highly distorted when mechanical beam tilt is used at transmission sites such as Mount Wilson, and it doesn’t include elevation pattern data.
As accurate as the ILLR model may be, the results were a case of “garbage in; garbage out.” Station KVEA (DTV channel 39), which operates from Mount Wilson with a low gain antenna at 54 kW ERP was predicted to have a 6 dB stronger signal than KPXN (Channel 38), which operates at 1,000 kW from the same location using a higher gain antenna optimized for coverage of Los Angeles.
In reality, KPXN is much stronger than KVEA at the tuner test location, so much so that some older tuners missed KVEA during channel scans. However, due to KPXN's horizontal plane radiation pattern (power straight out from the antenna; not what's hitting the ground) in that direction, the FCC predicts it to have a much weaker signal. I discussed this problem in my 2006 NAB presentation "RF Delusions
This is not to say the FCC DTV Reception Maps are useless. In the majority of cases where TV transmitting antennas are not on high mountain tops and where mechanical beam tilt isn’t used, the relative results should be quite accurate. The FCC Web site is not the only one affected by lack of accurate antenna data. Results from the TVFool.com
Web sites are based on the same data and will show similar errors--GIGO.
I'd be interested in readers’ comments on how well the FCC's Web page predicted reception at their location.