Abstract

 Here we will analyze the color distortion that occurs to the color image data when we produce an LRGB image, and demonstrate a technique to rebalance the LRGB back to the original precise G2V calibrated RGB image.

Introduction

As astro-imagers, we constantly push our cameras and gear to their limits to record highly colorful and detailed images of faint deep sky objects. There are two primary techniques used today for producing color images but is missing the connection between them which must exist to produce the most accurate and color correct photographs. The first of these is what is called "G2V calibration". While you can read about the details of doing such a calibration here on Al Kelly's web site, for this discussion we will first concern ourselves with why we use this, and second, how to apply this accurately produced color image to our second most popular technique in the imagers toolbox: The LRGB color composite.

Why G2V Calibrate?

Anytime we produce an image with a digital sensor, we take three images, each one filtered with a different standardized color: Red, Blue and Green. These images are then combined in software to produce the final color image. But here's the catch - unless we standardize on what we consider to be a white point, the three images can be combined in any number of mixes and the colors will not be right. Astrophotographers have chosen the color of our sun for the white point, since it is central in our eyes color vision response, and it appears as pure white to us in the daytime, when at the zenith. (The sun is not yellow, take a sheet of white paper and put it in the sun. What color is it?) Rather than go into a lengthy discussion on why the G2V star such as our Sun was chosen as white, I will say that each color gamut available such as Adobe RGB, or LAB color space uses a different white point and we must compensate. By using a standard candle so to speak for color, we can now compare our images taken over large areas of the sky and the spectrum of colors recorded by our cameras will look similar in the final images. So my image of the Trifid nebula will look just like yours, except for size. Direct comparison with a common standard.

Why LRGB?

Soon after the widespread use of color filters in CCD imaging was started back in the 1980's, it became painfully apparent that long exposures were needed through the dense color filters to get dim nebulosities to show up. Producing a direct RGB image was not only extremely time consuming, but also the long exposures made for tracking errors, and focus shifts with amateur grade equipment. The solution that emerged, and is now commonly used today for many deep and detailed amateur shots is the technique of LRGB. Here we combine a deep black and white full spectrum image with color frames taken with the camera binned 2x2 or more to drastically shorten the total exposure time for one subject. There are many sources to read about producing LRGB images, for example one of the best is Ron Wodaski's Advanced CCD imaging book. But there is a dilemma: The LRGB image is convenient to make, but not color accurate. So bad are the color distortions produced by introducing a new Luminance image into the LAB color space or by combining with layers as Luminosity in Photoshop, that they no longer resemble the original calibrated G2V data we so painstakingly produced.

A typical LRGB series

Below is an example of a typical LRGB composite. On the left is a deep Luminance image, that was binned at full resolution and has been processed to bring out the maximum detail. In the middle is the G2V calibrated color RGB frame, that was shot originally at 2x2 binning for maximum color signal. Crisp reds and electric blues define the beauty of this nebula clearly - but something happens to all LRGB composites as illustrated on the right. The saturation is obviously less, and the reds shift to the all to commonly seen (this is all too common, unfortunately) salmon pink hue. This is where most imagers stop, and consider the image done. Obviously this does not resemble the right colors we started with!

+

 Understanding and Rectifying the Problem

At the end of this article John has put together the details of the mathematics of producing the LRGB "Lab" composites and the color shift issues. Here I will summarize the details of both the color and saturation shift, and during our analysis a simple solution will be presented.

Where the Color Shift Comes From

When you combine a Luminance frame into a color RGB data set, you will pass through Lab color space. A direct method is to simply convert the RGB data that has been scaled up in size to Lab color space, then replace the "L" channel with a new highly detailed deep frame. Alternately you can paste the new Luminance frame right over the RGB data, and combine in layers with Luminance. Or the third way is to paste the RGB image over the new Luminance frame, and combine in layers using Color. They all come out the same because Photoshop does the mathematics internally using the Lab color space. In Lab mathematics, it can be shown that the three channels are not completely separate. The "a" channel is tied loosely with the L channel, but the "b" channel is not significantly connected. The result is that if you modify the L channel only, you will still change the "a & b" channels proportionately. Now the "a" channel controls the reds and oranges, and the "b" channel the greens and blues. Since the "b" channel is not significantly altered, it does not change hue as the brightness of the L channel goes up. The "a" channel is a different story! As you increase the brightness of the L channel, the "a" channel shifts its hue. The result is that reds become more orange resulting in the infamous "salmon pink" coloration, while blues stay pretty much the same. This is what the math predicts, and we will see in the examples below, that it bears out well in an actual astronomical image.

Where the Color Saturation is Lost.

Unlike the hue shift which plagues only the "a" channel, as the Luminosity in the "L" channel increases, both the "a" and "b" channels will loose their saturation. Because this is a non linear function, its is quite complex to predict a correction factor. Just turning up the saturation in a standard image processing software will only do a partial job. You can visualize the loss in saturation as the whites start replacing the colors at high brightness. The brighter the L channel gets (or any RGB image), the greater the loss of color saturation and contrasts. Many imagers try to compensate the salmon pink shift in hue by simply turning up the saturation. You will see below that this is quite unsatisfactory as well.

Demonstrating the changes with CM

First I will introduce you to a new imagers tool, created by professional photographers to better understand and change the color tonalities in their photographs. The version I am using comes standard with the full Picture Window Pro software, but the Color Mechanic is available separately as well for a lot less money. You can download a free evaluation copy from www.dl-c.com and try it for yourself.

On the left is the color analysis and correction diagram of Digital Light and Color's Color Mechanic. I have never found anything like it that does color adjustments and changes such as this amazing tool. Here's how it works, since you will need to understand it in the next steps. The very center is pure white - the sum of all colors equally. Three of the corners are the main RGB colors we are familiar with in normal color work, and the other three corners are exact even mixes of those three producing Cyan, Yellow and Magenta. Full saturation is in the corners, and as you go toward the center, the saturation decreases linearly. As you move in a radius about the center white point, you shift the Hue, or color frequency. When you click on a part of an image, a small circle appears on the diagram corresponding to its hue and saturation, allowing easy comparison of changes in our images.

Now that you have a good feel for this tool, lets move on to analyze what happens when you make an LRGB astronomical image.

The final acid test - the original G2V calibrated RGB and the new LRGB with correct color and much more structural detail.

You can see here the pair of images now have overlapping samples when comparing the original calibrated RGB image on the left to our finished LRGB on the right. Both the reds and blues now line up perfectly. The final LRGB now has stunning high resolution detail AND correct colors. It is our hope that we can change your thinking on processing LRGB images to a whole new level now. Don't drop the ball at the finish line - go all the way and get that color right on target !