Viewing and Configuring Results for CAV RGB Measurements

Viewing and Configuring Results for CAV RGB Measurements

CAV signals are simpler than Composite Video signals, CVBS, since the color components are separated in separate channels and not in a single channel using a complicated modulation scheme as in CVBS. However, due to normally higher resolution of CAV (HDTV) and thus larger waveforms, and due to that we now have to test three channels and not just one for most parameters, this tends to increase the test time for CAV. It is to a certain extend countered by the fact, than CAV requires test of fewer parameters because of the lower complexity of the signals themselves.

Color Bar

Some CAV test signals look quite different from those of CVBS and S-Video, for example the Color Bar. In Composite and S-Video, the color information is a modulated Chrominance signal. In CAV, the color components are separated in separate channels. The present section on RGB Measurements has the three colors Green, Blue and Red separated in individual channels. Each channel is normally limited by a 30 MHz bandwidth.

One example of a signal that looks different is shown in the following figure. This is a Color Bar measurement for a CAV-RGB signal.

Figure: The Color Bar for RGB has three channels, each shown by a separate color, Green, Blue and Red. Note the synchronization is a Tri-level sync rather than the Bi-level sync found in CVBS and S-Video.

The first different part is that the Sync signal for the RGB is a Tri-level sync rather than the Bi-level solution seen with S-Video and CVBS. For the lower frequency SDTV and EDTV CAV signals there is a Bi-level sync pulse similar to that of S-Video and CVBS. The Sync is also different in terms of that it triggers on the rising edge at the 50% level of Sync Low and Sync High (˜ 0 mV). In CVBS and S-Video the reference is the falling edge at the 50% level of Black and Sync Low (˜ -150 mV).

The Color bar signals look different from those of CVBS and S-Video. A White signal is created by all three colors at maximum level, i.e. 700 mV. Yellow is a combination of Green and Red, etc:

Figure 11: The RGB colors are combined to create White, the six colors and Black.

CAV Color Bar is a measurement of the amplitude of 8 color bars (white, yellow, cyan, green, magenta, red, blue, black) in a single specified component channel. Color bar levels should be measured for each component, e.g. RGB, to verify the brightness, contrast, and color fidelity of the video signal. They can also be used as a basic indication of non-linear distortions. Nominal levels can be predicted for each color and component and an error value calculated from the measured levels. There are 16 measurements for each color component type (RGB or YUV).

The Y signal is used as the signal to trigger on. In the case of RGB the trigger is on the Green channel.

Figure 12: The measured values of the RGB colors are available in NI VMS.

Figure 13: The Errors of the individual color channels are shown graphically.

Channel Delay

Since three channels are involved, an interesting feature is the delays between the individual channels. A Channel Delay measurement is available for CAV. It measures the delays between the individual channels, in the present case between the R, G and B channels.

The Channel Delay is a measurement of the relative timing of the three component video channels. The measurement analyses the cross-correlation between pairs of channels using a signal containing a large number of correlated peaks - the first 2 bursts of a multiburst, or the first third of a sweep. A negative delay from channel A to B indicates that channel A leads channel B. Only G-B (Y-Pb) and G-R (Y-Pr) delays are measured, the third delay is calculated from the previous two

Figure 14: The Channel Delay picture can be expanded by using the Zoom control, as in this figure. The present example uses a Multiburst test. It appears that the delay from G to B is about 31 ns, but much smaller for the G to the R signal.

Figure 15: The measurements obtained from the Channel Delay test. Only one test is required for all channels.

Horizontal Timing

The Horizontal Timing can be performed for Bi-Level synchronization as well as Tri-Level synchronization. The signal in the example is Tri-Level based.

Horizontal is a measurement of the amplitude and timing of the HDTV Tri-Level or Bi-Level Sync Pulse and the horizontal blanking interval. Back Porch is a measure of the back porch level before software clamping (DC restoration) is applied. The line sync datum is defined as the time of the zero-crossing (center) of the rising sync edge.

Horizontal Timing can be performed on any active line for most of the measurements of Horizontal Timing, but if the measurements:

• Start of Active Video
• End of Active Video
• Active Line interval
• Blanking Width

are to return values, a test pattern including an active signal, for example a Pedestal (e.g. White Pedestal), for the active area must be used. In the present test example, we use a White Pedestal that is found in lines 471 to 560.

Figure 16: The Horizontal Timing is checked in Line 500, using the H Timing parameter. The test above is for the G component of the video, and the test should be repeated for the B and R components as well.

Multiburst

The Multiburst is a measurement of the amplitude and frequency of 6 different frequency bursts in a multiburst signal, and basically it is similar to the Multiburst test for CVBS and S-Video. It represents the frequency response of the signal. Since it contains the multiburst response for all three channels, this test must be repeated for every channel to ensure the correct frequency response of all channels.

Figure 17: The Multiburst of the 6 frequencies for test of the frequency response.

Figure 18: The resulting frequency response. The response is calculated with reference to the Flag Amplitude, which for this measurement is automatic detected as the Sync amplitude.

Noise Spectrum

The Noise Spectrum is a measurement of the RMS level of noise in a quiet signal (i.e. no sudden changes in level). Either a Ramp or Pedestal signal may be used. Several filtering options are available to produce the measurement. The present example uses a Ramp.

Measurement is improved by averaging over several lines. Noise is one of the more sensitive measurements, so use typically 8-16 lines for the averaging.

Figure 19: The bandwidth of the noise is selected using the BW selector.

Figure 20: Note that the choice of bandwidth will often influence the noise measurements significantly. The present signal changes noise performance of almost 20 dB from a 1 MHz BW to a 15 MHz BW.

Non-Linearity

The Non-Linearity is a measurement of the non-linearity of amplitude increments in either a staircase or a ramp signal. The measurements provide the difference between the largest and smallest staircase step amplitudes, expressed as a percentage of the amplitude of the largest step. It also gives the difference between the amplitude of each step and the largest step, expressed as a percentage of the amplitude of the largest step.

Figure 21: The Non-Linearity measurement example uses a 5-step Staircase.

Figure 22: The Measurements of the Non-Linearity of the OmniGen signal.