|

Do you find flat, unattractive grayscale images to be the scum of the earth? Do you want to ensure your own photos don't undergo a lack-luster conversion? This guide will help you turn your color images into contrasty, appealing black and white images that show depth. It will discuss a few of the more effective methods for creating monochrome conversions.
The first thing you must have is a well-exposed image. Digitally, expose in color--in this world of awesome post-processing power, there is no reason to throw away color information in the camera. Both having a RAW file and 16-bit color depth are important, and will go a long way in increasing the quality of your final product.
Process your color photo to be properly white balanced. Correct colors go a long way in making digital filter application effective and worthwhile.
To demonstrate these conversion techniques, I contacted my friend jauderho, and asked him if I could use his beautiful color exposure of the J. S. Cain Residence.
The original image, correctly exposed and white balanced by Jauder:

Method 1: Two Hue/Saturation Balance Adjustment Layers
Layer › New Adjustment Layer › Hue/Saturation
Call it HSB1, and set the blending mode to Color, and hit OK.

Add a second Hue/Saturation Balance layer, call it HSB2 and leave the blending mode on Normal. Hit OK. When the HSB window comes up, move the saturation slider to all the way to the left (-100). Desaturating the image will make it look like this:

With a histogram that looks like this:

Now, Jauder did an excellent job on the original image, so simply desaturating it gives it a great feel of depth, and it is fairly contrasty. The histogram shows decent separation--it won't take a lot of work to really make this B&W shine.
We can play with the HSB1 layer now. Double check that the blending mode is set to Color. Click on the Hue slider and move it around until you get something you like.
The Hue slider, set to Color blending mode, acts as a universally-adjustable photo filter. Any hue in the color space can be your color filter, it is the ultimate set of Cokin color filters.

The application of that hue gives us a result that looks like:

With this histogram:

I added the saturation, so the results could be compared, on a separate frame. The saturation on the "filter" gives:

Histogram:

Overall, the effect of the filter is the midtones moving down and the highlights getting a bit brighter. The sky got richer (deeper), which adds more contrast. Since the sky got darker, and the clouds got brighter, the filter I added was some shade of red.
Now, a little trick I call "backwards USM" or "reverse USM" (just to give it a name) which adds more local contrast, in turn boosting perceived contrast even more.
Filter › Sharpen › Unsharp Mask
You will see three values: Amount, Radius, and Threshold.

Typically, when using Unsharp Mask (USM) to sharpen, a fairly high amount, a small radius, and small threshold work best. When doing "backwards USM", a mild amount, a large radius, and a mild threshold value will give the desired contrast-boosting effect: To get started, try 30/150/15. Usually a mild amount of reverse USM does the trick--it is something that you want to use to enhance an effect (local contrast), and used heavy-handedly, it can destroy the image. Keep an eye on the histogram when adjusting the USM levels, that will show when the shadows or highlights are clipping. After backwards USM application:

With a histogram that is a bit more separated, and the shadows and highlights are clipping slightly. In defense of the highlights, they were blown a bit in the sky before backwards USM, but it is not detrimental to the image.

Finally, I added a high-pass filter to increase the contrast and add a bit of 3D effect. A curves layer, or just leaving it after the USM would be fine. The B&W conversion left at the USM stage has an HDR feel to it, with great depth and luminosity fidelity.

Hisotgram of the final, filtered B&W image:


Method 2: Channel Mixer Adjustment Layers
I will do Channel Mixer next, because it is very similar to the HSB method. I do not use this method, but don't be afraid of giving it a try--it is a very popular B&W conversion method.
Layer › New Adjustment Layer › Channel Mixer

Set the first Channel Mixer layer to Color blending mode again, and hit OK. Leave CM2 on Normal blending mode, and click the monochrome box in the lower left.
Then go back to the CM1 layer. The RGB percentages work much in the same way in selecting hues as the HSB method, but this is channel specific. It's possible to adjust the RGB channels individually as well, from the drop down menu at the top. The "Constant" increases or decreases the overall luminosity effect.
It is typically a good idea (to start) with your RGB values summing 100. So 100 0 0, or 50, 50, 0, or 61, 153, -114. The constant slider can be used to make up the difference, if required.

It's not necessary to create a separate layer just to force the image monochrome--I do it for more versatility, and the ability to see what sort of color effect the filter has. The final product at R+200, G-50, B-50.

I enjoy the Channel Mixer method due to the intensities of the filters--it is possible to get a very luminous, very powerful filter of any RGB combination. Notice how the sky is nearly pitch black in the CM method with a very strong red filter. The artifacting in both comes from these being 8-bit images, which don't have the information capabilities to keep a decent gradient under extreme conditions.

Method 3: Lab Alpha › Grayscale › RGB Method
The last method which well get explained in detail. This method is becoming one of my favorites, but I have to explore it more to figure out the "ins-and-outs".
Image › Mode › Lab Color

Discard the a and b channels, leaving only the Lightness (alpha, luminosity) channel.

That will desaturate your image to the usual, flat one-click B&W:

Next, move into Image › Mode › Grayscale:

Select the luminosity values of the Gray channel by CTRL + clicking on the Gray channel;

Now that you have marching ants, invert the selection;

And move into RGB mode (Image › Mode › RGB).

Create a Solid Color Adjustment Layer:

Pick a gray near 50%; you can tweak it later.

Set the blending mode to Multiply:

That will richen and deepen the luminosities:

After a treatment of reverse USM and a high pass filter:

With a little more tweaking, this image would pop even more. Given my familiarity with the HSB method, I would have to say that version is the most pleasing to me. But, with a little time and some playing, any of these could be excellent monochrome images, and they would each have their own strengths and personalities from the different conversion methods.
Why go through the trouble of all these steps to take color out of an image? Because all of the one-click methods (excluding plug-ins) give flat results. Notoriously bad are simply desaturating, or just converting to grayscale. These lose a lot of information, and rarely have a decent appearance.
Better than either of those methods is using a Gradient Map Adjustment Layer. This will allow you to desaturate the image, and coarsely "set" your black and white points, and the mid point of the gradient map.

When attempting to create a successful monochromatic image two goals will help you achieve the proper results: Subject detail, and subject separation. The more detail in your subject, the more interesting it can become. Flat images don't have a lot of detail, so local contrast is important. Separating the subject from the background, if possible, will go a long way towards making your subject "pop!" Images lacking depth and tonal range have a difficult time with subject separation.

Once again, thank you to Jauder for allowing me to use his image. If you would like to use his images, please contact him personally; if you would like to use my images or reproduce this tutorial, please contact me.
My HDR tutorial.
Posted at 9:46PM, 17 September 2006 PDT
(permalink)
|