« Question All | Main | A Marriage From Agnosticism to Atheism to Christianity »
Preparing a Photo for Press in Black and White Using Photoshop
By Marc | March 4, 2008
Intro Notes
I love using Photoshop. It offers so many tools and doohickeys that no other imaging software currently can. But, not many people can afford the $650 for it. A more economical alternative is Photoshop Elements ($99). Then, there are the open source Gimp, Paint.net, and a variety of others. But none of them offer all of the items needed to effectively manipulate a photo for press quite like Photoshop. I guess you could use Corel PHOTO-PAINT, but you’d have to buy CorelDRAW ($429) to get it.
The free applications (at least the major ones) currently don’t offer an information window that gives you a numerical readout of color samples that you need for this. Also, all of them are too automated for tuning. (Auto-contrast, auto-color, etc.)
But, unfortunately for the pocketbooks of many, I will be using Photoshop for this tutorial. I will create a tutorial for Photoshop Elements if possible for similar results, but we will see what happens. I have a very old version (3.0), but the techniques I use are the same for all versions. So, if you don’t have the most current version, you don’t have to worry about missing out.
If you are putting together a newsletter, term paper, yearbook, self-published book, or anything that you wish to insert photographs to print in black and white, getting a descent print is somewhat of an art. So many times I see a bad printed photo that looks more like a smudge than a person’s face. Or how about a person wearing a white shirt against a white background and it looks like a floating head.
A press-printed photo is not the same as a photographic print. If you look very closely at a newspaper photo or any photo in a book, you will notice the image is made up of dots. With different sized dots, an image can have the illusion of a full gray spectrum. Something to remember in printing a photo is not to have large areas of solid black or solid white. A solid white highlight of an eye’s shine is okay, but anything more can make the picture uneven and funny looking. A general rule I personally follow is a low range of 4% gray and a high of 92%.
Two main goals I go for are: 1. gray levels within the range and 2. contrast. If you would like to use the auto-contrast and auto-colors to achieve good contrast, then by all means go for it. On some photos, especially older ones, these auto filters can do worlds of good. But, in the end your grays should be no lower than 4% and no higher than around 92-95%.
I’m going to go in to some explanatory detail throughout, but I’ll finish with a simple list of steps with an example. The links from here on out go to pictures of dialog boxes.
Scanning
Okay, if your image is already in your computer, skip this step, if not you might have to scan. First, set your scan resolution. A general rule for printing is to set the dpi (dots (or pixels) per inch) to twice the lpi (lines per inch of the printing) at the size you print. This can be confusing if you don’t know what the printer’s resolution is, but I go by a common lpi of 133. So, scan at 266 dpi, but I generally bump it up to 300 to be on the safe side.
But you also have to figure out the size of the original versus the size of the finished product. Let’s say my original is small, like 2 inches across, but you want it to print it at 4 inches. You should then double the scan resolution to 532 (or a safe 600). It doesn’t have to be an exact science, but it’s always best to scan high and reduce later when you resize. If you’re printing smaller than your original, just scan at 300 and don’t worry about doing hard math.
Another thing I should stress: Scan in color (RGB (or red green blue)). Even if your original is black and white to begin with, scan in color. When you scan in gray, you are more limited than color. You might not notice at first, but fine details can be lost in gray. If your photo includes someone in dark clothing, scanning in gray can limit that clothing to a solid mass of dark gray. You can also lose information in a white shirt, losing details. But scanning in color allows you to see folds and variations in shade. It doesn’t always make a difference, but it’s best not to lose that potentially valuable info.
If You’re Using an Already Existing Picture…
Watch out! A Jpeg that is saved in very low quality can be dangerously ugly when you start giving it contrast, so be on the lookout for what looks like a poorly played Tetris game all over your photo. Be careful when adjusting or perhaps use a different picture.
Convert to Gray
When changing to Gray, first convert to Lab color (Mode – Lab color)(recent versions: Image – Mode – Lab color). Then convert to Gray (Mode – Grayscale)(recent versions: Image – Mode – Grayscale). So, the file should go: Original Color to Lab to Grayscale. I do this to aid in preserving information in the image.
Follow this same technique of making a stop at Lab color if you are converting a file from RGB to CMYK (for 4 color printing). It keeps your colors from changing too much when making the transition.
Find Your Darkest and Lightest Areas of the Picture
In most cases you won’t have to do this, because you can sometimes easily tell.
Open up the Brightness/Contrast window (Image – Adjust – Brightness/Contrast). Slide the Contrast all the way to the right and the Brightness to the left. The picture should look completely or almost completely black. If you see white at this point, notice where the white is in the picture. If you don’t, slowly move the slider to the right until white appears. Where this white show up is the lightest area of the picture, so make a mental note.
Now, slide the Brightness all the way to the right and do the opposite, noticing any black against a mostly or all white background. Adjust the slider slowly to the left until black appears and that black will be the darkest area of the picture; make another mental note.
Cancel the Brightness/Contrast window (don’t save the adjustments).
Adjust Your Curves
First of all, make sure your Info Tab is visible (it’s by default in the upper right hand corner of the screen and looks similar to the figure at the right.)
Open up the Curves window (Image – Adjust – Curves). Pay attention to the buttons in the lower right corner. These are the main ones you will use. Double click the black dropper. This will set your high black level. But, since there is no black level, type the number 20 into the R, G, and B boxes. This color translates to around 92% in grayscale. Click OK. Then, double click the white dropper and change the RGB values to 245 in all three boxes, which translates to 4% gray. Click OK, then click Auto.
Click the black dropper and move it to the darkest area of the picture (remember the mental note). Look at the Info tab, at the place marked K (for Black). You should have 2 numbers. The first number is the gray value before you opened up Curves. The second is the current value. If the current value for darkest is above 92% (sometimes the computer likes to make it 100), then click and it should change to adjust your picture.
Do the same thing for the white dropper and move to the lightest area. Make sure that second number is 4%. If the number is lower, click. If it’s higher, there is an area that’s lighter than that. Find it and click. Now click OK.
Adjust Your Levels
You don’t always have to do this, but most scanned photos need it. It adds some contrast and makes the midtones a little more varied.
Open the Levels window (Image – Adjust – Levels). Change the middle space above the diagram from 1.0 to 1.25 or 1.5, depending on how much contrast you want. Click OK.
Adjust Your Curves Again
Follow the same steps from before.
Sharpen It Real Good
Open the Unsharp Mask filter (Filters – Sharpen – Unsharp Mask). Use the default values of 50%/1 pixel/0 levels. Click OK. If you want it sharper, do it again.
You should now have a fairly decent photo ready for printing. Sometimes it still doesn’t look right, so play around; making sure that in the end you have good contrast and correct high and low gray levels. It doesn’t have to look pretty on the screen but it should look distinct and clear when printed.
The only thing to do now is resize to your desired specs.
Okay, now for a step-by-step (the picture has already been scanned and cropped):
(If you want to use auto adjustment filters, this is the best time. My example used no such auto ajustment.)
1. Convert to Grayscale (original color to Lab to Grayscale).
2. Find darkest and lightest values. (click Cancel after done) (the boot at the bottom right and the paper the girl is opening were my two extremes)
3. Adjust Curves.
4. Adjust Levels. (this step forward isn’t always necessary, but it helps to flesh out the midtones and shady parts).
5. Adjust Curves again.
6. Sharpen.
Topics: Crafts, Tutorials | 4 Comments »
March 4th, 2008 at 9:04 pm
Very nice. You should do more of those.
March 6th, 2008 at 12:47 am
Thanks, man! I’ve got a few ideas.
September 4th, 2008 at 6:42 am
Thanks for the help! Should I follow the same steps for a color photo: convert to Lab then CMYK and follow the rest of your steps?
September 4th, 2008 at 9:57 am
Your welcome! Actually, for color: yes convert from Lab to CMYK (the colors remain much truer to the original) but the rest of the steps are pretty much for b/w. Color doesn’t have the restrictions of needing contrast the way b/w printing does. I think definitely use the “unsharp mask” filter to help it “pop” a little.
The newer versions of Photoshop have some excellent filters for automatically fixing things like skin tone and old photos in image->adjustments. For me, color is a matter of eyeballing it.
Try this link for color corrections.
I wish I were more help on the subject.