4) Converter tests
4.1) List of devices, video sources and signals tested
Video converters tested
Video sources used
Video signals used
4.2) The test procedure
4.3) Results

4) Converter tests

As the development of DVA Fidelity was kicked off by one of our video A/D converters returning unpleasant results, this section covers more detail about our hands-on experience with the DVA Fidelity Analyzer tool in practice. NOTE: As we consider the discovered test results to be valuable information for others dealing with A/D video converter equipment, but do not want to discredit any manufacturer or any device, we have anonymized this section in this regard.

4.1) List of devices, video sources and signals tested

Video converters tested

Video sources used

Video signals used

~137 hours of captured test-video (3,23 TB video data)

4.2) The test procedure

Overview: NOTE: For video that came from a physical carrier (e.g. VHS), we did two recordings over the exact same signal chain in order to distinguish tape (carrier) errors from timing/signal issues during the evaluation of the analysis results.

4.3) Results

During our intensive testing period one of our first findings was that in order to make a serious evaluation of an A/D-converter it is crucial to record a testvideo with a duration of several hours. It is simply not possible to observe a converter's behavior in total in just a few minutes - or even an hour. This is because - as we experienced it - some undesired behavior may appear frequently, others in bursts, and some issues only happen rarely, even over a long time period (e.g: 1-2 times in 4 hours).

If you want to monitor the way the converter in question performs a timing correction (dropped/inserted field/frame) you need a recording long enough to not only show an incident once but also to verify this behavior as a characteristic of the converter. For example, one issue with one converter was that every hour a field insert occurred. That issue was repeating in an almost perfect frequency with up to 2-3 seconds precision. We would have never found this behavior if we would have performed our tests with a shorter duration. Our experience showed that a duration of 4h of each recording seems to be sufficient to monitor a converter's characteristics in field/frame-handling and produce comparable results.

Probably the most important impact of these tests is that we have gained knowledge about the average amount and different kinds of timing-correction done by A/D converters. Before commencing these tests, for example, there was no information available whether to consider dropping 2-3 frames over a period of 4 hours was to be considered as a malfunction-indicator, normal behavior with an average amount of frames - or not. Testing 5 different converter models from different manufacturers and different epochs was enough to get a good impression about the average timing interference and corrections of A/D converters.

This information is very useful when testing and comparing other converters, as their behavior can now be put in reference to each other.

Back to index