Color bars and RexEdit RT
 

system is RexEditRT 2.93

Trying to set up a video chain so I can constantly monitor the video waveform and vector presentation. I have a recently calibrated Tektronix 520A that will operate in either display mode.

Using the color bar generation capability, I create 75%, 100% & SMPTE bars and place them on the time line.

Looking at the composite output from the BOB, I see that the bars look as expected in the 520A's waveform mode. But seriously lacking in amplitude in the vector presentation.

When I use the vector and waveform presentations that are part of the color correction filter, the waveform monitor agrees with the 520A. But the vector display shows the color end-points right in the small boxes where they should be. If I run the color bars from a Horita into the 520A and look at the vector presentation, they are where they should be. . . right in the small boxes.

So as a double check, I capture the Horita bars via the composite input and display them on the time line. They look just like the RE-generated bars including a very close similarity to the vector and waveform displays. And generate an output to the 520A that is very similar too. Good looking waveform, short vector display.

I must be missing something here. Any ideas or guidance?

Thanks

Are you sure that this "color bar generation capability" is actually
creating 100% compliant bars?

I have two sources, a Horita generator and the software-generated bars from RexEditRT.


I've done some more testing:

The output I'm using for the waveform monitor, composite video, is one I almost never use since DV for mastering or S-Video for VHS dupes are my normal output paths.

Further testing shows:
When I A-B color bars between the composite and the S-Video outputs from the output of the BOB, they look the same on a monitor. When I take the DV out to the DSR-20 and A-B the S-Video from the BOB with the S-Video from the DSR-20, they look the same. I can believe that one piece of equipment is bad but not two. More measurements are in order and now I've got to build a matrix of inputs and outputs on paper to see if there is something I'm overlooking

I've got to go back to the books and figure out why the luminance signal can be OK but not the amplitude of the color vectors as measured by the 520A.

I called Canopus Support to discuss this as additional testing had further confused me.

Their initial take is that the Composite output of the DVRex and DVRexRT may have a design fault. That of low chrominance.

They have asked Engineering in Japan to look into the matter.

With that information in hand, I went back to the tests and confirmed that the S-Video output is good and the Composite output is bad on all my DVRex boards. Since I've tried three different boards, I'm fairly confident that the composite video in the DVRex and DVRexRT is a bad design.

First excessive noise (The RT add-on board solved that) and now the realization that the composite output from the DVRex and DVRexRT is useless as a source of quality video.

My mistake was in viewing S-Video output on my broadcast monitor and measuring the composite video output. I'd forgotten that the comp video output into the monitor is driven by my DSR-20 which does not have a DV to Composite conversion problem. So everything looked rosy. But wasn't. Clearly the display color gets 'darker' when I A-B the color bars the way I should have in the first place.

So, not nearly as convenient but workable is to let my DSR-20 convert the DV signal to composite and look at that. Now the vector display shows the slightly hot (by about 10-15% chroma amplitude) output of the system.

Hopefully Canopus tech support will get back to me with some information and a solution.

Canopus Japan replied with the answer that the chroma is a bit 'short' out of the composite connector but that the monitor would automatically adjust for this.

The answer is what I expected, the comment that the monitor will adjust is BS. Broadcast monitors and vectorscopes are designed to show this type of problem.

I paraphrase their comment to: "The composite output of the DVRex and the DVRexRt is essentially useless."

Did I mention that the composite output of the DVRex is also known to be noisy?