In the training classes these days, the role of the SMPTE color bars or beasts of a similar nature appear to have fallen into disfavor. Primarily used as the pattern to apply the blue color filter to when adjusting for color and tint, even this has fallen to the side as weaknesses in the blue filter methodology are pointed out to the end user and the method is almost discounted completely as a result. (The blue filter method is actually still considered to be better than nothing at all.) The color bar patterns are also used for adjusting the color decoder in the display, but there too, it has fallen into disfavor as color decoder controls have not really made an appearance on any digital display for five to six years now. I think I last saw them on the Sony SXRD rear projection sets.
These days, it is all about color management systems and using colorimeters and spectrometers to do the color and tint part of things. We are all taught how we can’t always trust our eyes and to look at real images of people before we sign off that this color bar pattern actually looks right.
So has the usefulness of this pattern been lost? Well I am going to throw a curve ball into this and go the other way. I never thought I would be the one saying it, but I encountered another display just recently (60″ Sharp Quattron) where the built in color management system was not working properly at all and just leaving the color and tint controls at the default positions also resulted in an ugly looking image. The blue filter method did not work, and instruments completely failed in this situation since they resulted in images even more garish, although they did provide for pretty graphs that said there was no problem at all.
This is yet another abject lesson in how we should never rely on just a pretty graph to tell us when an image actually looks right or not. So what is a person to do? The test gear fails us and the blue filter fall back fails us and people just don’t look right. I know that the picture is terribly wrong, but it is not clear what I need to do next to fix this. The flesh tones of people are not enough to tell me what direction to go in. Where do we proceed from this point in time? How can I help my client out here?
So I go back to what I call first principles. I bring up the SMPTE color bar pattern and I go from memory, knowing what this pattern needs to look like and what each of the colors needs to look like. No one color in this pattern is supposed to appear brighter than the other colors. With this information in mind, I begin to play with the color brightness controls on the Sharp to make one color brighter and another one darker until things look balanced to my eyes. I make changes to the tint of particular colors as needed when the cyan looks a little too blue for example. When all is said and done, and I am happy that this color bar pattern looks the way I expect it to look, I go back to my flesh tone reference images and all the people look right. The image is saved for the client and he cannot believe the difference he is seeing.
This is the third or forth time I have had to do something like this for a display in recent years and it is good to know that this pattern still has a use in helping us to salvage a lost cause of a picture into something amazingly watchable. Just don’t try to take any measurements after this. The instruments will still lie to you.
Here are some test discs that you may want to look at for good sources of SMPTE patterns and a whole bunch of other patterns:
- Digital Video Essentials: HD Basics [Blu-ray]
- Spears & Munsil HD Benchmark and Calibration Disc 2nd Edition
- Disney Wow: World of Wonder [Blu-ray]