Each photographer or designer needs to do color correction from time to time as part of their job. Sometimes the task is very complicated. Sometimes it is simple but tedious, and takes a lot of time. Sometimes you know what you need to do, but don’t know how to do it. And sometimes you are so overwhelmed by correction layers, layering modes, and third-party software, that you forget what you wanted to do in the first place. FireGrade, our reliable and user-friendly plugin for Adobe Photoshop, will help you do color correction in a nice and easy way. Now you can leave all that annoying stuff behind and quickly color-correct images within Photoshop.
FireGrade internally works in the Lab/Lch color space, which closely matches human color sensation. However, keep in mind that the match is very good but not perfect. For example, Lab does not provide perceptual uniformity. The Munsell Color System most closely matches human color sensation. It provides perceptual uniformity for lightness, chroma, and hue. So far, there are no color correction tools for the Munsell color space, but we will do our best to add that functionality soon enough. Stay tuned!
What is Lch, and in which way is it better than Lab? Actually, Lch is essentially the same thing as Lab, only represented cylindrically. But why is it better? That’s because it is easier to work with colors if you split them into chroma (с) and hue (h) instead of using the ab coordinates. For example, it is much easier to grasp "Chroma = 60, Hue = 135" than “a = -42, b = 42” (at least, for humans). In this case, the combination of values corresponds to a “greenish” color. Hue shows the “greenishness” (color tone), and Chroma shows chromaticity (that is, how “colorful” the color is, what is its “color power,” the intensity of “greenishness”).
It is always easier to manipulate only one component rather than two, let alone three. Working in Lch, you can quickly and easily “think” in these metrics, so you can manipulate each color component separately.
Both in Lab and Lch, L stands for Lightness.
Lch color space looks like this:
Before installing FireGrade, please verify that your hardware and software meet or exceed the minimum system requirements.
Windows 7 x64 or higher
1 GB free RAM
100 MB free disk space
1920x1080 screen resolution
Adobe Photoshop CS4 (x64) or higher
Mac OS X:
Mac OS X 10.8.5 or higher
1 GB free RAM
100 MB free disk space
1920x1080 screen resolution
Adobe Photoshop CS5 (x64) or higher
Installing FireGrade on Windows. Run FireGrade.exe and follow the installation wizard:
Installing FireGrade on Mac OS X. Run FireGrade.mpkg and follow the installation wizard:
Important Note! If you are installing FireGrade on Mac OS X, do not change the location of the plugin’s folder. Otherwise, the licensing system will be unable to work, which will prevent you from using FireGrade. The "/Nuclear Light" folder containing the plugin must be located in the standard applications folder ("/Applications").
Also check the installed plugin’s properties so that the required access privileges are provided.
Put your license file into the installation folder. If you have Windows, the default folder is "\Program Files\Nuclear Light\FireGrade" (unless you’ve specified a different folder when installing the plugin). If you have Mac OS X, the folder should be "/Applications/Nuclear Light". If the license is valid and you did everything right, you can start using FireGrade.
You can find the plugin in the Adobe Photoshop filters menu: Filter -> Nuclear Light -> FireGrade
As we said above, FireGrade works in the Lab color space. So your image must be in that color space, too. But what if you have a complex document with lots of mask layers and other stuff, it is in RGB or CMYK, and you do not want to switch to the Lab mode? Smart objects will come to help!
Convert the layer of interest to a smart object, open the object, convert it to Lab, use FireGrade, and save the changes. If you need to do “nondestructive” editing (that is, you may want to do corrections at any moment), convert the layer inside the smart object into another smart object, and apply FireGrade. Now you can open the smart object again to make changes.
Smart objects are a standard feature of Adobe Photoshop, and here we just showed you some tips and tricks.
Keep in mind that because of the limited and varied gamut of displaying devices in the RGB/CMYK color spaces, the final result may differ from what you expected. That is especially true if after correction, the resulting color lies outside the document’s profile. The result may also differ from what you see in the FireGrade preview when using the plugin if your monitor’s profile doesn’t match the document’s profile. FireGrade always uses Adobe Photoshop to convert previews to the monitor’s profile. Therefore, we strongly recommend working in Lab, and converting the image to the RGB/CMYK target profile only at the final stage. This way, you can verify that the result of conversion is good, and possibly do necessary corrections.
We also recommend working in the 16 bits per channel mode, which will prevent unwanted posterization effect in some cases. Even though FireGrade internally uses floating point numbers, the 16 bits per channel mode will let you get better results compared to 8 bits per channel. Particularly, if you use the 8 bits per channel mode, quality may be compromised when you finally convert the image to the RGB/CMYK target profile. All in all, 16 bits per channel is better.
FireGrade does not impose any limits on image size. It can handle very big images. The only limit is imposed by Adobe Photoshop itself. At the moment, it is 300,000 by 300,000 pixels. Of course, you will need a powerful computer to process such a huge image.
FireGrade fully supports multi-core CPUs and multiple CPUs.
FireGrade’s user interface consists of a few panels (windows).
The first one is the Preview Panel:
The second one is the Tools panel:
The third one is the Info panel:
Let’s take a closer look at these panels.
At the bottom, there are a few control buttons:
The buttons have the following functions:
In the square brackets, you can see the shortcut keys and mouse actions for quick access to functions. The first value of the modifying key (before the slash) is for Windows, and the second one (after the slash) is for Mac OS. Below you can find similar quick-access functions.
You can drag the image in the preview with [Mouse Left Button] as indicated by the open-hand mouse pointer.
Now take a look at the Toolbar. At the top, you can see some status indicators and information icons:
At the bottom of the FireGrade window, there are the following image analysis panels:
The first histogram shows pixel distribution by lightness, from minimum (left) to maximum (right). The second histogram shows pixel distribution by chroma, from zero (left) to maximum (right). The third histogram shows pixel distribution by hue, from 0° (left) to 360° (right).
If the maximum allowed values of lightness or chroma are exceeded, circle-like warning signs will appear. The maximum allowed lightness (L) is 100. The maximum allowed chroma (c) is set at 127. There are no information signs for hue, because it does not have any allowed values due to its circularity: After 360°, comes 0°, and vice versa. Chroma cannot be less than zero for mathematical reasons, so an information sign appears only at the right.
Though it is possible that an image has a chroma exceeding 127 (provided that it fits into the Lab color space used in Adobe Photoshop), we selected the histogram length and set the chroma excess indicator based on the value 127, which is good enough for statistical reasons and makes it easier to use the tool.
The gray color is used in histograms and other image analysis panels for pre-correction image analysis data, and the light-blue and orange colors are used for post-correction data.
On the next tab, there are a few Vectroscopes:
The first Vectroscope shows pixel distribution on the Hue/Chroma plane (which is similar to the a/b plane in the Lab color space). The second Vectroscope shows pixel distribution on the Chroma/Lightness plane. The third Vectroscope shows pixel distribution on the Hue/Lightness plane.
Now we need to tell you something about the projections. In the first case, you are looking at the Lab cube (or at the Lch cylinder) from above; you can see all pixels in the projection’s color shape, which is like a shadow on a surface. In the second case, the Lch cylinder has been cut by the semiplane along the 0° Hue line and “flattened” to form a cube, and now you are looking at the cube’s frontal projection. In the third case, you are looking at the side of the same “flattened” cube; you can see pixel distribution by Hue/Lightness.
On the third tab, there are a few waveforms:
Waveforms show pixel distribution in the image, from the top-left corner to the bottom-right corner, using the scanline method.
Like histograms, waveforms allow you to see pixel distribution by Lightness, Chroma, and Hue.
Take a look at the FireGrade window’s top-right part, which contains the plugin’s main tools:
The orange color in a tool’s tab title indicates that the tool is in use, and that it has been used to do some corrections.
Now let’s take a closer look at each tool.
This tool consists of three components:
Each component of the balancing tool has the same controls:
The three shortcuts for the color picker and the reset button are for three balance ranges: Shadows, Overall, and Highlights.
To change chromatic balance, drag the control with [Left Mouse Button]. To reset chromatic balance to its initial state, use the combination [Ctrl/Command] + [Left Mouse Button Click]. To change Lightness, manipulate the respective control with [Left Mouse Button]. Just like in case of chromatic balance, you can reset Lightness to its initial state by using the combination [Ctrl/Command] + [Left Mouse Button Click] for the respective control.
The input fields show the current values of corrections applied. Of course, you can simply enter a correction value in the input field. Besides, you can hover the mouse over the input field, and then use [Mouse Wheel Up] or [Mouse Wheel Down] to increase or decrease the value. This trick will work with any numeric input field in this plugin.
If you click the reset button, all chromatic balance and lightness values in the component will return to their initial state.
If you need to “neutralize” an image area, select it with the color picker in the preview. When you select a color with the color picker, you basically tell the chromatic balance tool to shift balance by a color that is an opposite of the selected color: Hue will be rotated by 180°, and chroma will be the same. It means that the selected color will be effectively neutralized, that is, its chroma will become equal to zero.
The color picker mode has its own indicator. To exit from the color picker mode without applying the changes, click the preview with [Right Mouse Button].
To increase the accuracy of changes, hold [Shift] while dragging the chromatic balance or lightness controls. This way, you can change the value just a little bit. You can use [Shift] for the same effect with the plugin’s other controls, too.
The nonlinear balance controls work just like linear balance controls, except for Midtones:
The impact of nonlinear balance differs from that of linear balance both for chromatic balance and lightness. Generally, that impact may be described as strong at the impact point and weakening toward the margins. For example, if you change chromatic balance, the most marginal colors (the ones with the highest chroma) will be changed less than the less chromatic, more neutral ones. The same is true for lightness. For example, if you change the lightness of shadows, midtones will change only slightly, and the highest tones will not change at all. The same applies to other ranges: All changes in the selected subrange have a lesser impact on the neighboring subranges.
Therefore, in nonlinear balance the color picker works not like in linear balance. The degree of color neutralization would be rather low. The higher the chroma of the selected color, the lesser the impact. So neutralization will be only partial.
On the third tools tab, there are color equalizers.
Before proceeding to describe the tools, let’s consider how a color equalizer works:
input values -> output values + correction values. That is, there is an input value, for which the color equalizer produces an output value, and also adds (or subtracts) a correction value. The graph of output values as a function of input values and correction values looks like a horizontal line (or like a curve, if correction values are added). So the graph is as follows: y = f(x).
There are 9 color equalizers:
For example, in case of using the equalizer L->L (Lightness -> Lightness), if you have a correction point with the input value L20 and the output value L10, then for all pixels whose lightness value is L20, the final value after correction will be 20 + 10 = 30. Pixels in the vicinity of L20 will get a lesser correction value based on the correction curve.
Equalizers for other combinations of input and output values works in the same way. The only difference is that there may be other input and output color components, with different values. Say, you use the equalizer C->L (Chroma -> Lightness) with the input value C20 and the output value L10. It means that all pixels whose chroma value is C20 will get +10 in lightness. Pixels in the vicinity of C20 will get a lesser correction value based on the correction curve.
Now take a closer look at the color equalizer’s panel:
The color equalizer provides the following controls:
You can work with correction points just like you do in case of color balance. The “precision” modifier [Shift] for dragging objects and the combination [Ctrl/Command] + [Left Mouse Button Click] for resetting a correction point work in the same way, too. Moreover, there are additional modifiers to make your work even easier. Here are some of them:
The Q-factor, or the quality factor of the correction curve, indicates how sharp or how smooth the curve is in the vicinity of the correction point. The higher the Q-factor, the sharper the curve; and vice versa, the lower the Q-factor, the smoother the curve. By default, the Q-factor equals 1.
You can use the color picker to add a new correction point for the input value based on the pixel area captured in the preview. The color picker captures an area of 7 by 7 pixels. The values in the captured area are averaged. The same principle of averaging values when using the color picker applies to all tools of the plugin.
Using this tool, you can directly manipulate any color. Basically, you work with 3 projections of the 3-dimensional Lab cube. You can switch from one projection to another by clicking the respective buttons on the panel. The schematic representation of the Lab cube (see the top-right corner of the panel) shows the selected projection; the arrow shows the observer’s view direction, which is perpendicular to the projection.
You can manipulate colors by adding correction points consisting of two parts: a source point and a destination point. The correction’s direction is indicated by an arrow that goes from the source point to the destination point. Whey you initially add these two points, they coincide. It means that there is no correction. But as soon as you move the destination point, the color corresponding to the source point moves to the location of the destination point. Just like with color equalizers, the Q-factor defines the smoothness of impact in the vicinity of the correction point.
Direct Color Control looks as follows:
It consists of the following elements:
You can work with correction points just like in case of color equalizers, except for additional functions and modifiers:
The farther the correction point is from the current plane, the bleaker and more transparent that point looks. And vice versa, the closer the correction point is to the current plane, the brighter and less transparent it looks. A similar principle is applied to the color space shifting grid: The farther nodes and lines are from the observer, the bleaker they look; and the closer they are, the more saturated they look.