To view this page ensure that Adobe Flash Player version 11.1.0 or greater is installed.
ASK THE EXPERTS In today’s typical delivery network, a media operation might receive an original uncompressed stream from a satellite or an existing network, or a fi le from a post-production process. The operation then needs to be able to output these “adaptive streams” as a number of different profi les, in different bit rates to support any number of platforms – including high bit rates for HDTV. Although some of these streams might consist of SD content, they’re now primarily high-defi nition pictures that must be decoded on the fl y, recorded, ingested for QC testing, and encoded using the compression codecs required by the target devices. During the testing phase, the system must provide an automated means of testing a number of key attributes against industry-adopted standards. These include video quality, audio quality, lip sync, loudness, and ancillary content. What are the primary resolutions that they have to deal with? As we said, there is still a certain amount of SD content in the broadcast chain. However, most tools that have practical application in today’s media operations are focused on prevailing HD formats, while also including support for SD. We design our testing solutions around four key resolution levels. At the bottom are the streaming formats for handheld devices and PCs, and just above that are today’s most active broadcast-level HD formats, including 720p or 1080i. A step up from that is 1080p 50- or 60-frame rate, a format that’s often required for a live event that’s being transmitted over a dedicated network, in-production dailies, or an HD camera feed that needs to be transmitted for production purposes. Then above that is the 4k level. All of these resolutions are included in every tool we make. In addition, we maintain a wide variety of input and output capabilities to enable customers to test individual devices and then place them into a test network architecture, either at a live broadcasting plant, a content delivery network, or in the lab of the product’s manufacturer/developer. Considering that audio requires more types of testing than video, what type of audio measurement is required? In general, we’ve found that our customers require three key audio measurements. The fi rst is perceptual quality, which simulates a human perception test and creates a measurement that is as close as possible to an actual subjective study done in a standardized environment. An important benchmark for this is Perceptual Evaluation of Audio Quality (PEAQ), an ITU standard that enables the most accurate audio quality measurement in the industry today. Our products enable this testing by establishing a PEAQ record documenting the top-end perceptual quality level. Next, we offer audio performance measurements in our real-time systems that identify performance issues with audio devices or audio in the network chain. These tests enable operators to fi nd audio faults such as silences and glitches, and the testing system logs the failure and simultaneously records the offending video in a real-time test session. Audio performance testing also covers lip sync measurements with millisecond accuracy using the same logging and time-stamped recording function that captured the out-of-specifi cation portion the A/V signal. The third area is a test for audio loudness, which has become extremely important with the adoption of global loudness standards. Effective loudness testing involves applying loudness standards to every individual audio channel and reporting a measurement for a program as a ASK OUR EXPERTS 60 | TV-BAY MAGAZINE: ISSUE 86 FEBRUARY 2014 group. We also offer a true-peak audio measurement as part of the loudness testing. What are the differences between real time and non-real time testing? There are fundamental differences between the two. Real-time testing involves measuring performance over time of an audio and video signal as it goes through a device or over the air. With our solutions, an operator can monitor a signal according to certain test parameters and then determine its performance over time. We provide a reference signal and compare it to the downstream version, and then provide a very accurate measurement of video or audio quality without having to use invasive techniques such as placing tones or markers in the audio. We have also developed an exclusive technique to perform non-invasive lip-sync measurements, and we provide real-time loudness testing. When an event takes place outside of a set threshold that was predefi ned by the operator, our system records the stream in question for later review and creates a log of the failure measurement that is associated with the recording. Non-real-time quality testing is a repetitive process that is useful for fi ne-tuning a device codec design, setting or network path to achieve a desired level of quality. Typically, it involves recording sets of audio and video content that have certain error-prone characteristics, such as high motion, and then putting them repeatedly through a testing process and measuring the quality against a certain variable, such as encoding bit rate. In this manner, the operator can reset or revise the variable to determine the optimal setting of a given device for delivery to a given end-user device. POST YOUR QUESTION ONLINE: Search ‘tvbay’ Tel. +44 (0)1635 237 237 Email. firstname.lastname@example.org