Nielsen's Active/Passive Digital Meter
by Dave Zornow
Here are excerpts from my July 1998 column for Cable Avails magazine. For the complete story and lots more on the business of buying and selling Cable Television locally and nationally, call Cable Avails at 303 837-0900.
"A decade ago, the challenge was identifying who was watching," says Jim Spaeth, president of the Advertising Research Foundation. "Now the problem is identifying what they are watching." In a 500 channel world where every headend offers a different channel lineup, both researchers and ratings companies are concerned that the current ratings system can't keep up with the proliferation of viewing choices and sources. As digital cable expands and high definition TV becomes a reality, will we be able to trust network and syndicated TV ratings?
Nielsen's answer to this problem is the Active/Passive Digital Meter, also known as the "A/P Meter." It is being slotted to replace a 20-year old system called AMOL (automated measurement of station lineups) which doesn't include cable and never imagined today's technical challenges. AMOL requires broadcast networks and syndicators to add a code identifying the programming source and the date and time to the Vertical Blanking Interval (VBI) of TV signals. Nielsen uses media monitoring sites in each DMA to pick up the AMOL code and compare it to the broadcast source at its origination point. Using this approach, Nielsen can credit local viewing of broadcast programs to its national rating.
One of the big problems with AMOL is that it ignores cable. If a cable system experiences technical problems that pre-empt a cablecast of a network program, Nielsen can't identify the aberration. If a pay network runs a "sneak-peek" weekend promotion and doesn't verbally notify the rating service of the program change, it is likely to be missed. And in cases where a highly rated sports event replaces lower rated programming for a one-time-only event, cable network programmers can't be sure which programs' ratings they are seeing.
The A/P Meter promises to address this concern by electronically identifying what is being watched on each set in a peoplemeter household. The system uses an active video code (an advanced version of the AMOL code called AMOL-2) which is inserted into lines 20 & 22 of the video image. Using the A/P Meter located below the people meter on top of each respondent household's TV set, the code is identified and returned along with information about who is watching each program to Nielsen's Dunedin, Florida processing facility.
But what if the video code can't be found? Video compression may have stripped it out, or a vendor downstream from the programmer may have overwritten it. Nielsen's new system claims to have three levels of backup allowing it to identify programs with undecipherable video codes. If the A/P Meter fails to find a video code, it then searches for an "inaudible" audio code which duplicates the video code. If neither of these identifiers is available, the A/P Meter extracts an audio and video sample from the program to create a unique fingerprint - or signature - of the unidentified program. The signature is electronically uploaded to Nielsen's Dunedin facility where it can be matched to programming collected from national broadcast, cable and syndication satellite feeds and local TV station signals in 211 DMAs.
Syndicators, broadcast and cable networks have been contacted regarding installing the new program encoding equipment. About half of NHI's peoplemeter clients - about 20 cable networks -- have already installed the A/P Meter encoding equipment.
Although Nielsen claims success for the A/P Metering system in lab tests, clients are not convinced it can survive the challenges of variable signal strength, multiple set configurations and other problems encountered in typical TV households. To answer these concerns, Nielsen is conducting a regional field test in several Northeast DMAs from Boston to Washington, DC. The test panel will compare results in 250 households installed with the current metering system and the new A/P Meter to 250 households with only the A/P Meter. Results should be available by the end of the year.
Nielsen's new system is not without controversy. The portion of the TV signal Nielsen has proposed using for A/P Meter encoding is Line 22, located at the top of the actual TV picture. Nielsen's rationale for using Line 22 is that it is less likely to be lost in transit than Lines 20 or 21. The audio program code also changes the program content by hiding a code in a place where the human ear is unlikely to detect it. But programmers are concerned that both the Line 22 video code and the audio code will interfere with the picture and audio quality of what is being transmitted. Nielsen responds by saying these fears are unfounded.
SMART, the competitive ratings system from SRI, is testing a different approach to encoding and identifying programming. Where Nielsen uses Line 20, SMART uses Line 21, placing codes next to the closed captioning information and V-Chip data which are protected by law. Nielsen has staged an aggressive press campaign to discredit SRI's approach, pointing out that it has no backup system and that Line 21 is already crowded and potentially unreliable. Another difference program suppliers have griped about is the price of using each system; Nielsen charges programmers about $3200 per channel for the encoding equipment; SMART supplies the hardware to programmers at no cost.
To address the charges and counter-charges that each vendor has raised, the Advertising Research Foundation and the Media Ratings Council are conducting a comparative audit of each system. The purpose of the project is to evaluate the accuracy and reliability of each system under various conditions. The report will look at how each code is inserted; how encoding and decoding work; how factors, such as household size, VCRs, etc are likely to create program identification problems; and what the expected failure rates are of the new equipment.
The results of this report are critical to media buyers and sellers. Will the new systems improve program identification, or just add more noise to the process? Stay tuned. Your future ratings may depend on it.
Copyright 2003, Dave Zornow
Home | Services | Products
| Cable Network Software Demos |
& Newsletters | About TNG | Search
© 2003 TNG Research. All Rights Reserved.