My biggest complaint about PQ

[David Kennett]

First of all, thanks to all those who have taken the time and effort to provide the sample clips! I have observed many of the picture defects mentioned, while some see things which I do not. I suspect a difference in decoders or native monitor resolution accounts for this.

My biggest complaint is the clipping (compression) of blacks. It appears consistently in all the clips I have seen, and can be seen in the test pattern clips from Joseph. The four darker grey scale blocks are merged into one. This is evident in all the clips I have seen, as the picture takes on a contrasty, over-saturated color appearance. The details in the darker part of the picture are GONE. The "capped lens" clip provided by Joseph supports this. I would expect to see some noise in the black, but it appears that the black level has been pulled too low, and is indeed clipping the blackest blacks.

It appears also that the red "black" level is too high, since the darker parts of many shots takes on a reddish tinge. This could be confirmed if absolute black (capped lens) could be seen, but again, it appears to have been clipped.

To assist others in evaluating this, here is a quick summary of the color balance process. It works for both cameras and display devices as well:

A color camera must first create a good B&W image of a B&W subject. If it creates a color "tinge", then that "tinge" will be superimposed on the color image. It's as if you had a separate brightness control and contrast control for each of the three primary colors, red, green, and blue. The brightness is adjusted so that the darkest part of the picture has no color, and the contrast adjusted so the brightest part of the picture has no color. On cameras, it is typical to also have a "gamma" adjustment for each color to also remove any color from the middle grey areas. In the old days, TV cameras had all these adjustments available to the operator, which provided me (and many others) with some job security. Today, it is most likely set with firmware by the manufacturer. The white balance optioins on cameras compensate for different color temperatures of light.

I have a pdf file of the EIA resolution chart (which I sent to Joseph) used for years to evaluate resolution, as well as grey scale. It contains wedges of converging lines to measure resolutiion, and grey scale blocks to evaluate color balance and to see if any clipping or compression of blacks or whites occurs. I will be glad to send it to anyone, along with instructions. Just let me know!

I would like to see more tests done to confirm (or not) my observatiions.