Descending horizontal bands; not the usual suspects

[Hutch 0]

I have read the message about cloned cards bought on ebay. I bought mine direct online from AnyKeeper , so I believe that mine is legitimate.

I have a system with 5 Samsung CCTV cameras connected to an ISIS-1624 MagicRadar/ISIS DVR card in a PC. The problem is that with just one of the cameras I get darker horizontal bands moving slowly down the monitor. It takes each band 17-18 s to traverse down the screen and the spacing of the bands is not consistent but shows a regular alternating pattern of c. 3/8, 3/4, 3/8, 3/4, etc. of the screen height. This problem camera is a later higher-spec model (Samsung SCB-5000P) compared with the others (Samsung SHC-737P) that do not show this problem (and a older Fujitsu CG-311 also works fine). I have tried swapping cables and input channels and the problem stays with the particular camera.

The problem remains if I power the camera from a 12 V car battery rather than from its own mains adapter. So I don’t think it can be a ground loop problem.

The problem remains if I connect the camera directly to the card (cable length = 10 cm!). In any case, the usual cables are only 3 m long (BNC, coaxial). Moving the cables around makes no difference. So if it is EMI interference, it is happening inside the computer box, perhaps between cards in adjacent slots, but only with the signal from this one camera.

The problem remains visible if I put a cap over the camera lens and turn up the gain (so that it is a grey rather than a black image), so it is not a consequence of fluorescent lighting, and anyway the camera has an option to fix the shutter speed at 50 Hz (this is the mains frequency in Germany).

Changing any of the numerous settings of the camera (e.g. sharpness, gamma, etc.) does not improve the problem.

If I connect up the problem camera to a video-cassette recorder (Sony Hi8 EV-C2000E) and connect to a TV, the problem is not apparent. So if it is something wrong with the camera, there is apparently a way to correct it.

So: I think the problem is some interaction between this camera and the card. Maybe there is some filter in the video-cassette recorder that removed the problem. Or could it be some sort of impedance mismatching such as one gets in audio systems? Or, this is my hunch, is it an issue that the camera is higher resolution than the card (1305*1049 vs 768*576)? The older problem-free Samsung cameras are 752*582. Any suggestions please? Even if the problem can’t be solved, your answers might help me decide with what further equipment to improve the system in future. Very many thanks.