December, 2015
www.us-
tech.com
Understanding Video Latency Continued from page 53
every 16 lines and transferred each chunk separately over USB, the laten- cy would be reduced even further.
Measuring Latency in a System A variety of methods can be
used to determine whether a sys- tem’s latency meets requirements. Advances have been made, over time, as researchers improve their meas- urement techniques. One method is a program that
displays a pair of colored vertical bars that oscillate back and forth. The phase of the second bar is adjustable relative to the first, so the motion of the second bar is delayed by a certain amount of time, simulat- ing latency. When a camera is pointed at the
first bar, the display shows the same bar delayed by the system latency. The goal is to make the position of the second bar in the program win- dow match the first bar in the dis- play, by adjusting the simulated latency value in the program. When the bars are in sync, the simulated latency gives an estimate of the sys- tem latency. The method is effective, however tuning the latency value while comparing two oscillating bars can become tiresome. Another method is to use a count-
er to calculate the time difference between the camera and the display. For example, a camera is pointed at a 1kHz counter situated near the dis- play. As the system is running, the operator takes a snapshot of both the counter value near the display, and another value in the displayed video.
Improving Cable Quality
Continued from previous page
superior low-resistance connections to the header, high voltage isolation exceeding 1500VDC, and rugged con- struction. Application of 4-Wire Kelvin
mea s urement techniques will im - prove the quality and reliability of cable and harness products. Precision resistance measurements of less than 0.1W will reveal wiring defects that are not visible in less sensitive meas- urements, including bad solder joints, faulty crimps, recessed pins, pin con- tact contamination, improper wire gauge, and stress-exuded wire. Resistive losses resulting from these defects in applications carrying cur- rent above 1A may cause excessive heat generation or fire, or in the case of measurement circuits which obtain input from precision sensors, may cause false reports or circuit misoper- ation. The 4-wire Kelvin resistance method not only makes it possible to obtain mW or µW sensitive measure- ments, but eliminates any effect of incidental resistance introduced by leads or the fixture. Achieving these benefits, howev-
er, requires test equipment with twice the number of test points than would otherwise be necessary, and a test fix- ture with two wires leading from the tester to every pin on the mating con- nector. Contact: CAMI Research, Inc.,
42 Nagog Park, Suite 115, Acton, MA 01720 % 978-266-2655 E-mail:
info@camiresearch.com Web:
www.camiresearch.com r
The program uses a low-latency
To calculate total latency, one looks at the snapshot and subtracts the dis- played counter value from the exter- nally placed counter. The snapshot allows taking a sample of both capture and display simultaneously. A final method requires more
control over the last stages in the processing pipeline, and a computer to run the measurement program. The camera source for video process- ing is aimed at the computer moni- tor. The end of the pipeline must ter- minate at the computer, through a network or USB/PCI/PCIe interface, and the program must be able to receive the video stream through a network socket interface (UDP or RTSP) or a video streaming interface (Video4Linux2 or Directshow).
API, such as the xv library on Linux, to display a window where a black colored background is pictured. When the program initiates a latency measurement, it records the initial timestamp and the display window turns white. The video pipeline cam- era is continuously capturing the window on the monitor, and the white image will wend its way through the video processing pipeline and into the computer. Meanwhile, the measurement
program is capturing the video stream from the network or video interface and recording the timestamp of each frame as it arrives. If the video stream is compressed, it is decompressed using the avcodec library. The meas- urement program uses an algorithm to analyze the luminance (brightness) of the pixels in the image. When the algo-
Page 55
rithm detects a sharp increase of brightness, it can calculate the latency by subtracting the initial timestamp of the window turning white, from the timestamp of the received video frame. This method is preferable to taking a manual snapshot, since the process is automated. A variety of technology exists to
help achieve lower latency. An exam- ple is the Sensoray Model 2253 com- pact USB-compatible audio/video codec, which has a low-latency pre- view mode. The Model 2263 supports a variety of analog and digital input formats, and has a low latency uncompressed preview on a host while sporting efficient H.264 video compression. r Contact: Sensoray Company,
Inc., 7313 SW Tech Center Drive, Tigard, OR 97223 % 503-684-8073 Web:
www.sensoray.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84 |
Page 85 |
Page 86 |
Page 87 |
Page 88