Bits Theoretical Measured 8 58.31dB 57.9dB 7 52.29dB 51.9dB 6 46.27dB 45.88dB 5 40.25dB 39.86dB Calculated versus measured SNR
Will the 15db correction factor used with SNR peak-to-peak luminance/peak noise still be a good number for HD? Would this change if the resolution went from an 8-bit HD to a 10-bit HD sample? Is there an exact way to calculate this factor?
Michael Robin responds:
Read my article on digital video noise in the March 2005 issue of Broadcast Engineering, page 22. There's a table comparing the calculated SNR with the measured SNR using a 15dB correction factor. The measured SNR correlates well with calculated SNR.
Beyond the headlines
I was amused by the July 18 edition of Broadcast Engineering's Beyond the Headlines e-newsletter.
One article, titled “What happens to TV spectrum after the return?,” discusses the emergency public use of the returned spectrum. Another article discusses Rohde & Schwarz supplying transmitters for Qualcomm MediaFLO.
Why was this amusing? Because Qualcomm is planning on putting “up to 100 channels of high-quality video and audio programming” on the spectrum currently occupied by UHF TV channel 55. That's right in the middle of the spectrum that is supposed to be returned for “emergency public use.”
Although I respect the honorable Senator McCain's intentions, there is the appearance of impropriety when, according to the second article, there is no pressing need for the “emergency public use.” However, there is a company that would like to start getting a return on its investment by selling its digital programming.
That brings up another point: Why are traditional broadcasters being forced from the upper UHF channels if another company can just come in and use one of them for broadcasting “video and audio programming?” Also, are there going to be different rules for traditional broadcasters and Qualcomm? After all, both send “weather forecasts, sports clips, cartoons and the like.” Will Qualcomm have to send out EAS alerts and keep a public file? It will be interesting to watch.
The opinions expressed are my own and not that of my employer.
Mark D. Bulla
Chief Engineer, WNUV-TV
The key difference between broadcasters and Qualcomm is that what we do is free to viewers. Qualcomm will offer a subscription service.
Don't forget: Qualcomm isn't one of those big, nasty mega content owners that's trying to rule the world either. Perhaps that justifies Senator McCain's actions. Can you say free air time?
Why are 75 percent color bars used in the PAL system?
Michael Robin responds:
The 75 percent color bars are used in cases where some element in the TV distribution chain cannot handle 100 percent color bars. The difficulty in handling 100 percent color bars is due to the frequency division multiplexing of NTSC and PAL chrominance information with the luminance information, resulting in excessive video signal amplitude and transmitter overload.
Because standard camera-generated video signals are unlikely to reach chrominance signal levels equal to those of 100 percent color bar signals, under normal operating conditions, the transmitter will not be overloaded. Early videotape recorders also had difficulties in handling 100 percent color bars.
Current digital equipment and systems don't have this problem, so A/D and D/A converters are aligned using 100 percent color bars signals. Problems may arise when synthetically-generated video signals, resulting in excessive analog video signal amplitudes, reach an analog NTSC or PAL transmitter. These problems will disappear with the imminent demise of analog television.
Q. Which of the following NTSC test signals can be used to measure chroma/luma gain and delay? Color multipulse, FCC composite, Modulated bar, Multipulse 100, NTC-7 composite
A. All of the above
Gregory Chambers, WSIL-TV
Test Your Knowledge!
See the Freezeframe question of the month on page 8 and enter to win a Broadcast Engineering T-shirt.
Send answers firstname.lastname@example.org