Warholized Graphic Rendering

Andy Warhol created iconic high-contrast re-renderings of Jackie Onassis, Elvis, Mao and others. I’ll show how to do this in Photoshop.

The Miley Cyrus VMA spectacle provided a golden opportunity for this rendering style; which can be as tight or loose as you want.  I located the assets, used the original Kraftwerk album cover as a guide. Then it was the high-contrast Warhol-serigraph technique to punch it out of contemporary event photography, same for the Donald Duck repeats which mimic the Jonas Bros, who are subbing in for the original band. The type treatment was a loose-fit on the original.

building the file

This technique uses:

• Threshold
• Masks
• Solid Color layers that clip to whatever area you mask.

Masks can be added or subtracted to without destroying original pixel data. Previous tutorials on this focused on duping layers, and then hacking out each area before applying Threshold to render continuous tones into a high contrast-tonal statement.

Working with masks mimics Dave Lefner’s reduction linoleum print strategy. Or, if you’re ancient enough to remember cutting rubyliths, you’ll be right at home.

How to clip layers

1. Open or create an image that has several layers.
2. Hold down Option (Alt on the PC) and position your mouse cursor over the line dividing two layers in the Layers panel.
3. Your cursor changes to two overlapping circles with a small arrow icon. You can also choose Layer→Create Clipping Mask or choose the same command from the Layers panel menu.

Working with Masks

1. Start masks big and work toward small
2. Keep the clipping tight, 1px edge
3. Punch holes in masks by either filling with white or black depending if you want a color to reveal or conceal
4. Dupe the master layer and bury it. You’ll make a mistake, and its good to have a backup
5. Clipped layers are layers that are linked directly to the layer beneath it. Here its whatever solid color/gradient to the mask in question.
6. Set color to multiply to get the color/black combo you need.
7. Control saturation with layer opacity

I started with a PS6 doc 907 x 960px, a Facebook album size. My source was Kraftwerk’s “Robots” album. It had the graphic elements necessary to pull this lampoon off. Yes, its wildly low-rez, all it needs to be is a guide.

background and color layers

Miley and Thick masked, but still in their native full color.

Threshold applied, Adjust to preferences.

Threshold applied, Adjust to preferences. Masks with colors not yet activated above, but waiting.

Close up of masks showing conceals/reveals and clipped color layers.

The starter mask in all its glory.

Miley Cyrus Kraptwerk, finished.

Add type layers and graphic elements as needed. The added Donald Ducks, are each on a separate layer. Same for the Jonas Brothers on the right. The original source album below:

Kraftwerk “Robots” album cover.

Generating Black And White Photo Prints From Color Labs

Sample image in RGB, profiled Adobe 1998.

You can get quality chromagenic black and white images from an RGB photo-printer, in this instance a Costco Noritsu 3411. In some locales its  a Noritsu 34-Pro, Noritsu 3111,  Fuji Frontier 590, etc. You get the idea.

 These are not the same as  the  Epson 7880 inkjet printer, using their 8-color Epson UltraChrome K3TM inks. The RGB process requires a tighter control on how you profile the image to the specific printer at the location you choose.

Go ahead, hate on Costco. I did too until I found out that Art Center students were printing their portfolio pieces at Costco. A 12×18 print for $8.99? Hit me. I bought a membership, figuring it was a $0.96 weekly lab fee.

The Noritsu printer in its various configurations is used by Costco and higher-end labs alike. What you get out of it depends on what you put into it. This means determining printer profiles. For Costco you have your choice of “Automatic” or custom profiles. Automatic is a series of probable averages based on color outputs. And if you’ve noticed indifferent images coming out of the printers, odds are very good they were sent in that way—minus color-balancing, corrections, and on automatic.

Once you select the profiles you need, download them, and move them into your Library/ColorSync/Profiles folder:

Selected color profiles, as loaded into the Library/ColorSync/Profiles folder.

STAY NEUTRAL ACROSS THE BOARD

When you start outputting RGB grayscale, things get interesting. The object is to keep all three colors neutral at whatever point you sample them, ie a neutral 50% gray is 128/128/128. In real life things aren’t always so neat. I’ll sample 4 points (why 4 Photoshop, why?) to look at shadow, 3/4 tone, 1/4 tone and highlight info.

Art, with Threshold and Point Samples [3×3 average]. Points 1 & 2 are available black and white, with 3 & 4 representing 3/4 and 1/4 tones.

Periodically, the profiles are updated, and/or you may have just upgraded yourself from Photoshop CSx to 5, 6, whatever.

I’d recently upgraded to CS6, which I’ll discuss later. Bottom line was I’d sent in prints, which I thought were correct, but were coming back with a pronounced magenta cast. What the hell? The lab techs at the Los Feliz/Atwater here inLos Angeles were very helpful in running a spot correction. But the problem was on my end. Now to solve it.

The Once And Future Prints

The first problem was to break down all the likely places where a color shift would occur. I made screenshots of each step, and imported them into each target, so when the prints were output, I could review all the information in one place. Surprises awaited.

• Assign Profile
• Color Settings
• Convert to Profile
• Save As (with embedded profile)
• additional curve corrections

The outputs would be both tif and jpg in two iterations:

• “0” [sourced as Adobe 1988]
•  “1” [sourced as sRGB]

Additional checks would be provided by the Gretag Color check pallette.

Gretag Color Checker, sRGB profile

image 0 workflow

Here were the checkpoints to “0”, the Adobe 1998 native file:

This image was saved as a flattened jpg and tif.

“0” tif, flattened with info.

0 jpg with info

image 1 workflow

I repeated the process again with the sRGB tif and jpg, Image 1. The sRGB profile is much narrower than Adobe 1998. Skip down to the end of the post for specific references.

This time I added a curve-correction layer, with a window in the mask to contrast native and corrected states (dainty red square and red highlighted oval on the Info pallette).

reviewing the results

The four prints showed interesting variations. Clearly the sRGB source profile that I converted to the Costco printer file had a more pronounced magenta cast.

• After saving the jpg, you must open it up again, reassign the color profile, even if you’ve already done it. If need be, apply the correction curve, then flatten image. This is a Photoshop tic that must be addressed.

Other observations:

• Magenta was also more noticeable in the jpg rather than the tif saves in both “0” and “1”.
• The screenshots, which were imported as both jpg and png formats. These “emotional-neutral” images (screenshots)  skewed big-time.

conclusions

But people don’t care about screenshots. They’re looking at faces and skin tones. Are they warm, or cool? Believable or not?

Further info on color spaces

I’ve used Adobe 1998 for my native color profile for my photography. Only after I’ve made all adjustments, etc, will I convert it to sRGB for the web, and hammer it some more.

Now, a chart!

Adobe 1998 has a wider gamut than sRGB. Unfortunately,  as Zach Sutton said “RGB came first, and almost everything on a computer is built around sRGB.”

Since we’re talking about printing here, you can correct and balance an sRGB file, just remember that you’ve clipped a lot of information.

Further insights on Adobe 1998 and sRGB

AdobeRGB vs. sRGB

Comparison of Adobe RGB and sRGB Colors

Why You Should Probably Use sRGB