HDTV OTA Reception Issues
Not everyone will have problems, but my transition to over-the-air HDTV reception was not a completely smooth one. This page is meant to chronicle the current state of TV reception at my home.
This document last modified November 21, 2011.
The bottom line: everything comes in great, except for WNET Channel 13. This just happens to be my favorite channel - arguably, the flagship station of the PBS network. And I can't receive it at all.
Here's the background:
I've had trouble receiving VHF 13 (WNET) ever since the digital transition. Most of my UHF stations came in fine even with an attic-mounted antenna. This antenna is no slouch - the Winegard 7698p is the biggest, highest-gain, most-directional VHF-hi/UHF antenna that Winegard makes, and I've paired it with a low-noise Channel Master 7777 pre-amp before feeding the signal into a home-theater PC with two dual-tuner Hauppauge 2250 capture cards. So there's an external two-way splitter before each dual-tuner card's own two-way splitter. That's 6+ db of splitter loss, hence the desire to use a pre-amp.
I'm 37 miles due south of the transmitters at the Empire State Building. As per TV-Fool, I have a line-of-sight view to the transmitter from my now 30-ft antenna height, but TV Fool doesn't consider one large apartment building about a 1/2 mile north of me. That building may be casting a sufficient shadow to wreak havoc on my signal strength on a number of channels. Regardless, here's the current TV Fool signal chart for my location.
The length of this two-story house runs approximately north-south; it's actually canted about 15 degrees east of true north. The antenna, when previously mounted in the attic or now above the roof, is aimed north (0 degrees), which gives it a good view of the ESB stations, as well as WNJN to the west and WLIW to the east. These two PBS stations carry most of the content that I used to be able to get from WNET.
This is an old house (built in 1924), and it currently has three layers of roofing shingles on the roof. That probably exacerbated any problems from attic-related reflections (multipath), and may also enhance canceling reflections from any antenna mounted too close/low to the roof. When the antenna was mounted in the attic, it barely fit, with about two feet of clearance on either side when canted towards true north.
I also have a nearby low-power FM station (WYGG, 88.1) that is close enough to be a potential source of overload to either a tuner or a pre-amp. To eliminate that concern, I've installed a Pico Macom High-low Splitter-Joiner, with a terminator on the low side. This functions as a very effective FM-trap, dropping everything below VHF channel 7 by at least 20db. The offending FM station was reduced from 37dBmV to 4dBmV. (The Pico-Macom HLSJ is no longer available from Solid Signal, but this seems to be an identical part:
Blonder Tongue High-low Splitter-Joiner.
The screen shots that follow are from a Sencore SA1454 spectrum analyzer. All measurements were taken directly off the antenna, without the trap/HLSJ or preamp in line.
I have a few observations about the signals that I measured. First off, take a look at the third page of six screen shots. These are from WNET channel 13. The spectrum charts are the bottom three pictures. Note the way the signal drops off at the high-end of the channel for the attic-mounted antenna (in the first picture). My best guess is that this was from canceling reflections within the attic. It was that theory that prompted me to move the antenna out of the attic and up to the roof, initially on a five-foot, gable-mounted mast. Since part of the mast was mounted below the roof-line, the antenna was actually only about 30 inches above the roof-line. Regretfully, I never broke out a tape measure to get the exact height above the roof. But notice in the second spectrum photo - of the first roof-mounted antenna signal - that this roll-off shifted to the bottom of the spectrum. That tells me that the reflecting surface is further away than it was in the attic; a more-distant reflection would impact a longer wavelength, or a lower frequency. Now consider the wavelength of VHF channel 13. (See my wavelength/frequency table.) Channel 13 has a 55 inch wavelength. What if my 30-inch antenna height guesstimate was off by a few inches? What if the actual height was closer to 28 inches, corresponding to one-half of the wavelength at the beginning of the channel 13 bandwidth? A reflection from any surface at that distance could easily knock out the signal in that portion of the band.
The solution would be to raise the antenna even further off the roof-line, preferably at least one wavelength higher than the roof. Since the lowest frequency that I'm concerned about is VHF channel 7, that means at least 68 inches.
I went to Radio Shack a few days later and bought a ten-foot mast, and used the same gable mount. That has the antenna about eight feet above the roof-line. The results are obvious in the third spectrum picture of the Channel 13 series - the signal strength across the band has largely flattened out. Overall signal strength is still too low for reception, but the roof-reflections are gone.
Most other channels exhibit considerably smoother spectrum charts, too.
Another significant observation about the roof mounted antennas was that I actually lost signal strength when I took the antenna out of the attic. This runs completely counter to what should be happening. Normally, when you raise an antenna higher, you gain signal strength by some marginal amount - and sometimes not so marginal. But the signal strength readings are all over the map, and they don't even remotely correlate to what TV Fool says my expected signal strength should be.
To see this more clearly, I charted the actual, unamplified signal strength that I'm now getting on each channel, and compared that to what TV Fool expects that I should get, adjusted for the relatively high-gain Winegard 7698p antenna vs a reference dipole. Winegard shows the gain ratios in db for various channels, and I extrapolated from those figures to take some guesses about the gain of other channels that Winegard doesn't specify. For the most part, most channels should see 10 to 14 db more than TV Fool expects. This compiled data is visible in this signal strength spreadsheet. The last column sums up the story here: the readings are all over the map.
After discussing this with a number of people who are familiar with the issues, and a few who have experience with this particular brand or model of spectrum analyzer, we've concluded that there's a fault in my spectrum analyzer that's leading to incorrect signal strength readings. Fortunately, the spectrum charts and the carrier-to-noise readings appear to be reliable.
One final observation that I've noticed about the roof mounted antennas is that most channels exhibited better carrier-to-noise ratios than their corresponding attic mounted measurements. (See this carrier-to-noise ratio spreadsheet.) This better carrier-to-noise ratio, as well as the overall smoother spectrum chart, for each channel was manifested at the tuner in markedly better HDTV reception - pixellation and video errors are now far less frequent, and audio-dropouts are practically non-existent.
And having a home-theater PC, where recordings are made to a file on the hard drive, actually allows me to run some error correction and analysis software on a recorded program. Most recordings exhibit no more than two or three faulty video or audio frames, out of perhaps 100,000; cited another way, this can amount to under 300 erroneous bytes out of a 6GB file. This is excellent performance, better than many cable systems.
As I said at the top of this page, every channel comes in great - especially now that I have a roof-mounted antenna. Except channel 13.