Field-testing 8-VSB

An encouraging FCC Office of Engineering & Technology (OET) interim report on digital-television-receiver development, released last week, compares measurements of analog NTSC and DTV reception using the ATSC's 8-VSB modulation system.

From last April to September, at 51 sites in and around Washington, field tests were performed by FCC engineer William Inglis, traveling in a van with both a 30-foot mast antenna and two types of indoor antennas (measurements taken out of doors) and several third-generation prototype and production-model receiver chipsets.

David L. Means was chiefly responsible for organizing the field data and is the main author of the report.

The test results show that third-generation receivers are performing a lot better than first-generation units.

Yes, that's what we found. The results were very pronounced under conditions that gave problems to the first-generation receivers at the more typical sites. We deliberately picked sites that we thought would pose difficult reception problems, and these new receivers worked.

The study is at the half-way point, and you plan to test 50 more sites.

Yes. Right now, we're in the process of trying to accommodate more receivers that have become available. We only have six ports on our truck, however, so we're trying to decide which receivers to test. The plan is to swap some of the prototype models for chipsets that have been incorporated into consumer receivers. We're also trying to figure out how to handle the fact that we're changing receivers in midstream.

With the study only half finished, why release this data now?

We thought that we had some important findings, and the data so far indicated some clear trends, so we wanted to get the information out there.

One interesting part of the report was the testing of indoor antennas outdoors. But can reliable data be collected this way?

It was because of the logistics of the whole situation. If we had tested indoors, there would be wall attenuation, but that's difficult to predict and model. Eliminating wall attenuation at least gives you a handle on how much the reception is affected by the relative heights of the antennas and the types of antennas.

The tests compared NTSC and DTV. Is this a valid way to test performance of DTV reception? Why not test 8-VSB against COFDM or, perhaps, just 8-VSB coverage?

Well, at the time the test was undertaken [in April 2000], there was no way to do a good COFDM comparison. We undertook the test with what was possible with the resources at hand. Late in the test period, there was an attempt to get us involved in the NAB/ MSTV modulation test effort [comparing 8-VSB and COFDM with NTSC coverage], but, by the time the logistics were lined up, their testing was over. I don't see any reason to test COFDM at this point.

In general, what do your test results prove?

I'm encouraged by what we found. Speaking as a scientist, it's hard to make a quantitative assessment, other than the fact that the new receiver chips are X dB better. It's a clear indication that a lot of the problems that were observed early on were the result of early implementation of first-generation receivers. The main improvement is in the channel equalizers that companies are now building into consumer receivers.