Converting A Color Image to Black & White in Gimp
I know there are tons of tutorials out there on how to convert from color to B/W. However, most all of them are for Photoshop. This is for the folks out there who, like me, use Gimp for most day to day editing.
I could go on and on about how great Gimp is, but I’ll save that for some other time. The simple fact is that you can do just about anything in Gimp that you can in Photoshop, and Gimp is totally free. It’s just a matter of learning how Gimp works.
Just like in Photoshop, there are numerous ways to convert an image to B/W in Gimp. I’m going to cover a few of the more well known ways here. I’m sure there are more, but these are the methods I regularly use. This tutorial was written using Gimp 2.2.17. Other versions may be somewhat different in menu options, but all functions should work the same. Click pictures for a larger view.
Lets start by looking at a couple of the more popular ways to convert. I have chosen a base image that is heavily saturated and is strong in all RGB channels. This will produce the best conversion to B/W. You will have different results on images that have a different RGB balance. These are the most popular because they are simple and easy to do. The first way, and way that I probably use the most is a simple desaturate. It can be performed easily by going to Layers -> Colors -> Desaturate.
This method produces a very clean Black and White image with flat contrast and little if any added noise. The caveat here is that this method offers no ways to adjust the image. It simply removes all the color data from the image and leaves you with whatever luminance you had to begin with.
Here is a 100% crop of the original image, followed by a 100% crop of the desaturated image:
As you can see, this is a very efficient method of making an image Black and White.
Another very common method is a simple grayscale conversion. In this particular instance, the results are almost identical to the desaturation. Again, this removes all the color data from the image, but leaves you with no way to adjust the output. Grayscale can be found by loading the image and going to Image -> Mode -> Grayscale.
A 100% crop shows that there isn’t much difference in this and the desaturation method. If anything, the contrast is a little weaker, but that can easily be fixed by manually bumping the contrast up after the conversion
Another method which is a little more obscure and probably used less is the Hue/Saturation method. This method AGAIN removes all color data and is pretty much identical to the desaturation method. The difference here is that you get a “lightness” slider that can help adjust the output of your B/W image. This method is accessed by loading the image and going to Layer -> Colors -> Hue-Saturation.
The Hue-Saturation dialog box is opened and the conversion is done by moving the Saturation slider all the way to the left, as shown. The Lightness slider is just above the Hue slider and can be adjusted for overall image brightness. Another benefit of this method is that by adjusting the Saturation slider but not moving it completely to the left, you can remove some color, but not go totally B/W for different effect.
A 100% crop here shows results very similar to the first two methods:
A method that produces some stunning results and really shows how the conversion works is the decompose RGB method. This method takes each channel, Red Green and Blue, and converts them each to grayscale individually. They are then merged back together, but can be stripped out with a little digging. First lets look at how to perform this conversion, then we’ll analyze how it works. Load the picture and go to Image -> Mode -> Decompose. This brings up a dialog of radio buttons wanting to know what to decompose. Select RGB, make sure “Decompose to Layers” is checked and click OK
The software does its thing and it opens the results in a new window:
As you can see, the finished image isn’t much to look at. This is because it has merged all three channels back together into one image. To see the individual results, go to Dialogs -> Layers. What this does, is open a way to view each color channel individually. You’ll notice that each color layer has an eye icon next to it. This means that layer is turned on. Click the eye icons next to Green and Red layers, leaving only the Blue channel visible.
As you can see, this layer is the one that contains most of the shadows for this image. The resulting B/W image is very dark. Most all images will have more shadow and noise on the blue channel than on any other channel. This image is no exception.
Now this is where it gets interesting… Click the eye to turn off the Blue layer and click the eye to turn on the Green layer. You get a very nice looking B/W image from this color channel.
The green color channel generally contains the data which converts into the best looking Black and White images. This photo proves that generalization is correct. This is a very usable image with good luminance and overall balance.
Now lets look at the last layer, the Red layer: Turn off the Green layer and turn the Red layer on:
As you can see, there is a lot of luminance in the Red layer and most of the highlights of the photo are here. If this is the image you are looking for, feel free to use it, but most people will go with the Green layer. To use any of these layer views, simply leave the other layers off, and save the image.
You may also choose to decompose by other methods in that radio box dialog. Feel free to experiment with it. All of the other selections give interesting and different results. Just be sure to decompose to layers so that you can view each portion of the decomposition. Most of these are not usable for stand alone images, but they can produce some very interesting results when used as layers merged back with the original image, or another B/W conversion.
The last method I’m going to cover here is the Channel Mixer. This is a unique method of conversion because it lets you adjust the final image before conversion. The actual conversion is similar to decomposing the RGB channels, but you do not get the individual layers in the end. What you get is a single image which consists of all three RGB channels merged. The way to do this conversion is by loading the image and going to Filters -> Colors -> Channel Mixer.
Make sure the Preview and Monochrome boxes are checked, then you can play with the sliders. If you want to keep the overall brightness of the picture, check Preserve Luminosity as well. What these sliders are doing are combining the results of the individual layers you saw when Decomposing by RGB. You have the same three colors here. This is simply letting you blend all the channels into 1 image that has the good characteristics of each. This method is generally the best method for getting a B&W image with excellent contrast and overall balance.
It may be hard to notice the subtle differences in these conversions in the tutorial, but try them out for yourself and you’ll see what each one does, and you’ll quickly develop a favorite way of converting to Black and White in Gimp.