Q: How did you execute your Holmes graphic (see above) ? Did you scan the material for the clothing and hat then skew it in Photoshop.  It looks very cool ! Maybe you don’t want to share your secret. I hope you do.

A: It’s not really any kind of secret. It’s just some simple Photoshop trickery. I didn’t save a “step by step” nor did I save the original pattern files, so I will just have to explain it. I do have the original unflattened file so you can see the various elements.

First, I inked the image and scanned it as I always do, then moved the line work to it’s own layer.

Then I painted the background using the paintbrush tool and some airbrush type brushes:

Then I painted the face using my techniques explained HERE:

Finally, the part you asked about… the clothing. Holmes is famous for wearing a type of pattern called “houndstooth”. A simple google of “Houndstooth Pattern” yielded several nice samples. Sorry I didn’t;t save the exact file and I couldn’t find what looked like the same one again, but any decent one would do. I created a single rectangle of the pattern, and then using the “Image>Adjustment>Hue/Saturation” feature I changed the color to a nice green/red for the hat. Then I simply copied the rectangle and moved it “under” each section of panel of the hat, between the seams. Then I used the “Edit>Transform>Distort” feature on the rectangle to create the illusion of perspective and direction of the pattern for each panel. I also would rotate or skew the rectangle to get it to fit in the bill or back of the hat. Then I would erase away the parts of the rectangle not needed to fill the particular panel I was working on. Then I did the same thing for the next panel of the hat.

I again used “Image>Adjustment>Hue/Saturation” to change the colors to a more brown/red for the jacket. I used the same method to alter the rectangle to fill in the various sections of the jacket. Here’s what that layer looks like with the rest hidden:

The shading was done using the “Burn” tool primarily, with some additional painting using a brush on “multiply” mode so it would not have any opacity effect on the pattern.

It’s a pretty simple way to approximate a complex pattern on clothing, but you have to use it warily as it makes for a shocking and out-of-place realism in an otherwise cartoon illustration. I generally dislike the use of realistic textures or patterns in cartoon illustration, and do not use it much. I did use the same method when I did some Spider-Man and Superman illustrations for MAD KIDS, but I created the patterns myself that time:

In this case I was able to use the magic wand tool to isolate the inside ovals or diamonds from the background material and create some shiny light effects that gave the costumes a little more volume and texture.

Thanks to Michael Garisek for the question. If you have a question you want answered for the mailbag about cartooning, illustration, MAD Magazine, caricature or similar, e-mail me and I’ll try and answer it here!

For years I’ve been using an easy trick to create a layer in PhotoShop containing my scanned, inked line art that remains intact as I color “beneath it” sort of like an animation cell. It’s a simple thing to do:

1. Scan line art as grayscale image
2. Duplicate background layer containing inked art, rename “Inks”
3. Set layer mode to “Multiply”
4. On background layer, press “Command” + “A” to select and then “Delete” to delete line art on that layer
5. Convert to RGB or CMYK

Finished. Because “Multiply” mode means that whatever is on that layer in “multiplied” with what is below it, all the black lines stay intact and all the subtle gray lines become transparent and overlay the color I place beneath it, while all the white becomes transparent. Neat, easy trick.

Except nothing is ever easy, is it?

There are two difficulties with that technique. First, the white areas on the multiply mode are not gone, they are merely inert when in Multiply mode. This means that once you take that final step and “flatten” the image for sending to the client, all the white areas combine with the lower color layers. Since it’s in Multiply mode when flattened that just means any color below it takes over and the white in effect is gone. That works great IF you have only one layer of lines in Multiply mode. But what if you need to have different layers of objects in a given illustration for some reason? Then it does not work, because if you merge a multiply layer with another layer, any areas on the other layer that have no color in them become opaque white, and no transparency is transferred. In other words, if you want to have a single figure, inked and colored, on it’s own layer on top of a background illustration you cannot do that with the “multiply trick, because once you merge the multiplied inked figure layer with it’s separate colored layer, the “white” comes back all around the figure:

The linework for the boat and the color beneath the boat are their own layers
in this image, with lines for background beneath and color for background beneath
that. It looks like this if flattened to all together at once.

This is what happens if I just merge the boat line layer with it’s colored
underlayer. The white on the inked layer comes back.

You can select the white areas with the magic wand tool and delete them to create the transparency, but that is problematic as the wand tool doesn’t do a very good job of making good edges and you end up with a kind of “halo” effect that necessitates a lot of clean up around your image. I’m working on a job right now that requires a multiple layered final file, and this is a real headache.

There is a larger problem with this technique, though, and it applies to the process of four color printing. I just learned about this from MAD after I noticed that the blacks in my “30 Rock” parody seemed dull and washed out compared to other MAD jobs.

“Multiply” mode doesn’t just drop the black linework “on top” of the color… it literally multiplies it with the color below. That means that your black areas aren’t just 100% black, but they are black plus the cyan, magenta and yellow inks of the color beneath it. All blacks in a CMYK printed image are more than just 100% black ink… they have the other inks in there as well (in fact, PhotoShop has a setting for “rich black” in CMYK mode that is a specific combination of the four inks in percentages), but the density of the inks easily becomes very heavy when using the Multiply trick.

The problem with this in printing terms is that ink density (the percentage of each of the four colors) has it’s limits for the printer, yet PhotoShop literally dictates the ink density based on absolute percentages. You have 4 different inks in CMYK printing: cyan, magenta, yellow and black. Each can be up to 100% coverage. Therefore the max ink density is 400%, meaning 100% of all four colors. Printing a 400% ink density is impossible… it will never dry. PhotoShop’s default color setting profile calls for a max of 300% ink density, but even that is a little strong and those settings do not apply in PhotoShop to an image you are working on, but only to those that have been converted to that color profile. So, you may be working in a profile like CMYK SWOP v2 (default North American printing setting) but you can easily exceed that 300% ink density when working, especially using the multiply line trick. Printer’s want lower ink densities. MAD‘s printer, by example, wants a max ink density of 280%. Working in RGB and then converting to CYMK will limit you to a 300% ink density (or whatever the profile calls for), but I don’t trust conversion like that to keep the colors right.

So, in an effort to figure out a better way, I discussed it with several knowledgeable PhotoShop gurus and found a different line art trick that works around these issues. It’s called the “Channel” line art trick, and it works just as well and almost as easily, but results in a layer of line art where the white is literally not there and yet the black and gray lines are merely transparent as opposed to being in multiply mode, which results in a lesser ink density.

Here’s the process:

1. Scan line art as grayscale image
2. Create a new blank layer, rename it “Inks”
3. Go to the “Channels” palette, there is only one channel called “Gray”
4. At the bottom of the channels palette, click the “dashed circle” icon entitled “Load Channel as Selection”
5. In “Select” drop down menu, select “Inverse”
6. Go to your “Inks” layer
7. Press “D” on your keyboard to reset swathes so full black in active color
8. Press “Option” + “”Delete” to fill selection with black
9. On background layer, press “Command” + “A” to select and then “Delete” to delete line art on that layer
10. Convert to RGB or CMYK

Using this technique, your line art layer will contain all your lines but the white will be gone, rather than just inert due to the multiply mode. So instead of this:

You get this:

The great thing is that the channels trick also preserves the subtle gray lines and any washes or values you had in the original inks, as the selection of the channel is smart enough to not just select the absolutes but also the transparencies of the image. You can use this trick to create as many layers of line and colored objects as you want and merge them at will to create layered images. best of all, the transparent black reacts differently to merging than the “multiplied” black, resulting in lower ink densities.

The one caveat here is that you should scan your lines in at a higher resolution for this technique to make sure you do not lose any linework. I do most of my inks at 200% of print size, so that is plenty large if I scan at 300 dpi. If I was inking at 150% or closer to print size, I’d bump up the resolution of my scan to twice print resolution, or say 600 dpi as opposed to 300 dpi.

I am sure this technique has been used by many people, is all over the internet and I am hardly the originator of it, but it was cool nonetheless to “figure it out”.

Isn’t shop talk fun?

Still under the gun with work before the big Reubens trip, but rather than post the Dreaded Deadline Demon I thought I’d share a link to a website I recently discovered that is both a lot of fun and very sad at the same time.

PhotoShop Disasters is a website that’s a little like Jay Leno‘s “Stupid Headlines” skits on “The Tonight Show“… except funny. Contributors send in bad PhotoShop jobs and the authors poke fun at the results. Here’s one from a recent post at PsD:

Ooops. One too many hands! That guys must have trouble with “hand” penalties.

There are many examples, some subtle enough you have to look for them and some so horrible you wonder how in the world some of these people got jobs. Sure, there are thousands of community art colleges out there cranking out tens of thousands of graduates with degrees in “graphic design”, but are there really people out there putting together print ads where they forget to place both legs on a model?:

These are fun to laugh at, although I would be lying if I said I hadn’t screwed up a few illustrations in my time with dumb mistakes… including recently where I literally gave one of the figures in a poster image two left feet. D’OH!