Bit of a Con Really - Follow-up ...

Its more akin to saying that if you want to play handball with eggs, don't do it on a concrete patio.

No, when I say "poor studio standards", I'm talking about such things as the failure to set up cameras correct, keep a close eye on burst phase, etc, etc, etc. Garbage in, garbage out.

Many years ago I read about the work at National Geographic that was put into making color separations and printing plates to produce extremely high-quality images in the magazine. It was not simple.

At that time they only accepted kodachrome slides.

Geoff.

There is a little more to it than that, Bill. The points in the CIE diagram are only part of what makes for a proper image. If you are referring to the triangle of points on the colorimetry plot, you are only seeing the color of the points, not the luminance. The resulting image is a matter of saturation, hue, and luminance. You only see the first two with the CIE chart that shows gamut.

The other aspect of getting the points on the colorimetry chart right is that it tells you nothing about the colors in between and at different levels. All it tells you is the color of the points you measure. How they are mixed and create intermediate colors depends on the spectrum of each of the primaries and the color decoding scheme in the display. The underlying assumption in video is that we have a spectrum for each primary that is similar to the CIE standard observer curves, which are approximations of how we perceive color. When you deviate from those spectra you have to compensate or you will be intermediate colors that have to much or too little energy in a particular primary. There are no standards for how this is done, because there are so many variations in backlighting and filters in the displays. There are not even good metrics for getting to the bottom of the problem yet.

The result is that the LED backlit displays can look very good, but sometimes have a little strange color reproduction.

As for the black level and contrast ratios, they are only an improvement to the degree that they can control backlighting locally. As the number of controlled areas increase, the useful contrast ratio in an actual image may begin to approach the on/off numbers that they advertise, but with less zones of control, those numbers are simply meaningless to real video.

Leonard

This is true only if you have custom LUTs or decoding algorithms for a display based on the relationship between the spectra of the lighting and the CIE standard observer functions that cameras are generally aligned to approximate. The other thing that no one mentions is that trying to make up for spectral shortcomings with different filters and decoding reduces the efficiency of the lighting system.

There is usually a "rest of the story" beyond the naive assumptions that get thrown around about reproducing color. This thread is full of examples.

Leonard

VIR was introduced decades ago. It inserted reference signals into the vertical interval, near the start of each field of video. That allowed automatic adjustment of the equipment, and eliminated the video gain, black level, chroma gain, and phase controls that each operator could adjust, to 'their' preference. NTSC wasn't the problem, it was that everyone along the signal path could play with it. A system that had VIR from the cameras to the transmitter had no problems. Of course, that doesn't stop opinionated people from bashing a system they don't understand.

Only by idiots.

The cross country network feeds, that were owned & operated by AT&T. Those were replaced by C & KU band satellite feeds in the '80s. Some TV stations now feed CATV headends via fiber optic. They maintain the off air equipment as a backup, in case of a failure in the F-O path.

I was a TV Broadcast Engineer in the '70s - '90s in the US.

In a very broad sense you are correct, but in terms of understanding what is going on with color reproduction in LCD displays ( and others) you are making a point that is the equivalent of trying to make D65 with an incandescent lamp.

White is a rather useless term. All "white" has a color and is a mix of other colors.

Primaries do not get diluted with white. They get desaturated by adding the other colors. What you want is a spectrum that is correct, not "white," nor "full spectrum," nor narrow band RGB. Correct depends upon the assumptions that are made in recording the image, as well as upon the filters and color decoding that you implement in the display.

As I have said many times, the underlying assumption that video has used is that RGB spectral densities should follow the standard observer curves. When you violate that assumption on the display end, you get some unusual results with some colors and you have to compensate in your color decoder. The degree to which, and the techniques used are unclear in these sets. The results are mixed. Given the sloppy nature of color decoding and color management in consumer displays in general over the years, however, these sets are likely to be perfectly acceptable to most consumers. They are likely better than much of what they have been viewing in the past by quite a margin. That does not mean that they are going to accurately reproduce color. Most consumers and most of the posters here would not likely know ( and may not prefer ) accurate color reproduction in a display if they were to happen to see it.

Leonard

I'd be most surprised if all but the very smallest station with only one camera made mistakes like this.

There is no such standard as D6500. One standard, the one used for most video for the color of white, is D65. D65 specifies NOTHING about the spectrum, only the x,y coordinates of the COLOR of light. It happens to be approximately 6504K. The term D6500 is slang and sloppy use that confuses the issues of colorimetry and coordinated color temperature.

There are other standards for the color of white that are used for purposes other than video. Some specific purposes in film and cinema, as well as in video use other standards, but for the most part D65 is accepted as the color of white for modern video. The truth is that virtually no consumer displays come out of the box set anywhere near D65, nor producing correct color for any color, including white. What you see in showrooms and when you take a set out of the box is likely a color temp for white that is nearly twice what it should be.

Leonard

You have not looked at many modern displays carefully. Many, actually most of the better sets, have these controls in the user menus now. Some even have color management that goes far beyond gray scale and let you adjust the colorimetry (and perhaps luma) of the primaries and secondaries. Most professional calibrations these days never involve going into a service menu.

Leonard

While this is true, it would be virtually impossible to get all colors right with some arbitrary color backlight. You could get a subset right and get all the others completely wrong.

Leonard

This is not a simple subject. I have Mees' "The Reproduction of Color" (which is, what, 40+ years old?) and it's tough sledding. I had less trouble with integral calculus.

You would be very surprised.

Leonard

What I was implying was that you could reproduce a wider range of colors if the primaries weren't as close to the center of the chart. Radial movement represents changes in saturation -- dilution with white.

As for not knowing accurate color reproduction when you see it... What sorts of preferences does the average viewer have? If you don't have the original for comparison, it can be difficult to judge.

I never considered myself an expert in color reproduction, but your comments have encouraged me to dig out Mees and give him another try. (I'm not promising anything.)

My Pioneer does, but heck if I'm touching them without instrumentation.

The pioneer has six adustments, for RGB and CMY.

If you look at the first season of "Barney Miller", you'll see poor camera convergence, and slight color shifts between the cameras. And this was in the 1970s, and at ABC's studios.