Welcome to the Video Clarity February, 2009 newsletter. We are going to deviate from our normal column and discuss some frequently asked questions, but before we do. What is going on with the analog TV shutoff on February 17th?

The FCC and Nielsen reported that 6.5M Americans were not ready for the turn off of analog service. Commissioners Michael Copps and Jonathan Adelstein called for a delay to give the government more time to get "boots on the ground" to help with the transition. At least $200 million has been spent in recent years to explain to TV viewers how they need to prepare, but one key element of the federal government's transition plans hit a bottleneck in recent months. The National Telecommunications and Information Administration (NTIA), the arm of the Commerce Department administering the program, reached the $1.34B limit of $40 vouchers for the $40-$80 converter boxes. They are now sending out new coupons only as older, unredeemed ones expire thus freeing up more money. The NTIA had nearly 2.6 million coupon requests on a waiting list as of January 21, 2009.

Proving that there are two sides to every story, the US Senate recently passed a bill to delay the shut-off of the analog spectrum; while the US House did not. Why?  The four-month delay would come with a cost to local television stations. Keeping their old transmitters turned on would bring higher power bills and maintenance expenses. Paula Kerger, president and CEO of the Public Broadcasting Service, estimates that delaying the digital TV transition to June 12 would cost public broadcasters $22 million. The analog spectrum has been repurposed to several wireless companies and public safety agencies. Delaying the shift will cost them money as well.

What we can be sure of is that it will cost money during a time when everyone is struggling to cut costs. Many companies are struggling with how to and whether to test their video quality. We all know that testing is important, but it takes time, money and a plan. One needs to remember that the common goal is constant – make the customer happy. According to the adage, a happy customer tells 5 people about his experience; while an unhappy customer tells 10.

So we are going to take some time to reflect on how to streamline the video quality testing practices by answer some frequently asked questions.

Q: I’m rebroadcasting a signal or I’m an encoder/decoder manufacturer, how do I know if my incoming signal is any good?
A: Considerable research is going on with No Reference and Reduced Reference metrics, but none of these have been certified by the VQEG or ITU. Moreover, no metric is close to being certified. In the interim, a simple test can be done to reduce the problem complexity. If you take a source A/V signal (“reference”) and send it into your encoder/network/decoder (“processed”), you can compare the reference and processed streams and collect the following information:

• The Audio delay
• The Video delay (comparing the audio and video delay, you can check for potential lip-sync issues)
• Audio silence or Video drop outs
• Repeated Video frames or Audio samples
• Audio Quality drops below a set threshold
• Video Quality drops below a set threshold

The last 2 need a little explaining. If the reference signal has poor quality, then the encoder will have difficulty processing it. This is due to the predicative nature of encoders. How can you accurately predict an error? Simply said, you cannot so the encoder will force an error state and try to recover. The audio or video quality will by necessity drop. If the quality difference was X, a threshold can be set at Y percentage to detect these errors.

Detecting the error is not necessarily sufficient; thus, the reference and processed streams should be saved for further analysis. Below shows a simple pictorial representation of how to detect errors.


Video Clarity is introducing RTM, which continually reads 2 A/V streams, aligns them, checks the A (“reference”) signal for integrity, and does the A-B comparison on both the audio and video. Errors are reported and the streams are saved during an error state. The system is built into our ClearView Shuttle Broadcast Video Analyzer; thus, the error streams can be further analyzed using the full complement of testing tools with our ClearView analysis system. This includes full reference quantitative, objective metrics, no reference spatial and temporal metrics, and multiple subjective viewing modes so you judge for yourself.

Q: How do I test a new system development with respect to the previous one?
A: This depends on the intent of the new design – simply stated: does the new design change the quality or not?

If the quality is not supposed to change, then this is relatively easy. Simply compare the A/V processed results from the previous design with the current.  Two ways exist to do this:

• A set of test A/V streams can be inputted into the previous and next design and the results can be compared
• A set of test A/V streams can be inputted into the next design and they can be compared to the filed results of the previous design

In either case, scriptable commands using ClearView would create an automated solution.

Video Clarity is introducing RTM, which continually reads 2 A/V streams, aligns them, checks the A (“reference”) signal for integrity, and does the A-B comparison on both the audio and video. Errors are reported and the streams are saved during an error state. The system is built into our ClearView Shuttle Broadcast Video Analyzer; thus, the error streams can be further analyzed using the full complement of testing tools with our ClearView analysis system. This includes full reference quantitative, objective metrics, no reference spatial and temporal metrics, and multiple subjective viewing modes so you judge for yourself.

If the quality is supposed to change, then this is relatively easy as well, but it will take longer. To test the actual quality, you either need to use a quantitative, objective metric or subjectively review the results with a set of people.


Objective metrics have been tested and certified by VQEG and the ITU. They rate the video on either a DMOS (differential Mean-Opinion-Score) or JND (Just Noticeable Differences) scale, which reflects the consensus of a set of people. The ClearView analysis systems have many metrics built in to judge the video quality and include multiple subjective viewing modes so that you can judge for yourself.

Q: Couldn’t I just look at the IP/Compressed layer and not bother with the output of a set-top box (STB)?
A: Not really. STB have intelligence built within to perform error concealment and corrections. Errors will occur and what the user sees is the STB intelligent way to conceal them. What you want to measure is the complete user experience. Also aspect ratio and cadence errors are not easily found without fully decoding the signal back to uncompressed.