The Danger of Drawing Caricatures from Photos

July 31st, 2008 | Posted in Freelancing

Yesterday I posted this drawing of Michael Cera and mentioned how it was a “cold” drawing… meaning I had one photo reference to work from and no other familiarity with the subject either from seeing him in a film or on TV, or any other forms of reference.

Artwork ©2008 Tom Richmond

Here’s the photo reference I used:

I got a few comments and some private e-mails asking for more details about what it means to draw “cold” and what makes it different from using multiple references or drawing from a live model. I thought it would make a good topic for a more involved post.

Drawing caricatures from nothing but photo references is tough. Anybody who has ever done it will tell you that photographs lie. Tricks of light, focus, angle, perspective, etc. will provide you with false information about your subject. A given photo will suggest some features are ripe for exaggeration while other photos will tell you something different entirely. Basically it all boils down to this: you are basing your observations of a living, three dimensional subject on a two dimensional reference, and are therefore restricted in your observations to the superficial information that photo imparts. The best caricaturists in the world can do a terrific caricature from a single photo that looks just like the photo, but that doesn’t mean it will be a great caricature of the actual subject.

Take a look at the photo of Cera above. Based on that, I drew his eyes large and close together, exaggerated the width of his face and the roundness of it, gave him a small chin and exaggerated that by making the space between his upper lip and nose larger, and gave him a tiny mouth. I had nothing to go on but this single picture. A few people wrote me saying they had a hard time recognizing him in the sketch without first being told who is was. That does not surprise me.

When I draw caricatures for a job, I get as much photo reference together as I can. In fact, I make up a 13″ x 19″ sheet of pictures of each subject comprised of different angles and expressions that I use as my main reference. Here’s one I used recently of Alec Baldwin for the “30 Rock” parody I did this spring:

Using this kind of collection of photos I can get a much better feel for his features, expressions and personality, and can do more effective and CONSISTENT caricatures of him. There is a big difference between doing a caricature based on the superficial and doing one based on the foundations of a face. Drawing a “superficial caricature” is dependent solely on the exact physical traits of a given reference, and there are lots of potential pitfalls there. The decisions an artist makes with respect to that reference may not apply to the same face from a different angle or with a different expression. The artist may find trying to turn the head in a second drawing requires a different set of decisions to make it work. Each drawing by itself may be effective, but as a pair they are mismatched… they do not read as the same person. However if the artist bases his or her caricature on observations of the underlying foundations of the face… making decisions based on the continuity of the face between multiple angles and viewpoints rather than the superficial nature of a single photo, then the caricature becomes more solid and far more consistent. The artist is caricaturing the person and not the photo.

In the rather unique case of MAD parody work, the consistent part of that equation is crucial. I cannot base my caricatures entirely on one photo or even a series of photos. I have examine the subject as a whole… head shape, feature relationships, posture, etc. and figure out what the foundations are and then exaggerate them (hopefully) consistently. Only in that way does the viewer believe they are looking at the same character in panel after panel, with various viewpoints and expressions. It’s a little like developing an animated character design… you have to be able to draw the character consistently from many angles to make it believable. There’s a lot more room for error in something like a MAD parody, but the concept is similar. Here are a few of the caricatures of Alec Baldwin I did in that 30 Rock parody. While each leans somewhat on a certain photo, none are done dependent on that single reference:

You might notice in certain pictures in the reference page above Baldwin’s forehead looks fairly big, but overall observation showed me he needs a short, wide forehead. His pursed lips became an important part of his caricature when looking at the entire sheet whereas in certain pictures it isn’t as prevalent. Those are just some examples of how individual pictures can deceive you.

Better yet is to see the subject in motion on TV or in a film. That gives you a feel for their real persona. In the case of doing a parody for MAD this is absolutely essential. In the few times I’ve been required to do a parody in advance of a film’s release I have sought out DVDs of films the actors have recently appeared in to help give me the insights that only that kind of source can provide.

That is why drawing live caricatures isn’t drawing “cold”. Of course you do not know the person sitting in front of you so you can only draw quick inferences as to their personalities (although those quick inferences can be surprisingly accurate) but you can see their face in three dimensions. You can see them move and speak and make various expressions as you draw. Your observations as far deeper and more accurate from live than from any single photo reference, or even multiple ones like my reference sheet.

A lot of live caricaturists suffer from “superficial caricature syndrome”, where they rely too much on the surface features and do not pay attention to the underlying foundations when they switch to using references and working from pictures. Several very talented ones I know can do incredible single caricatures of individuals, but when asked to string several together consistently they struggle. That is something that needs a lot of thought and patience to accomplish.


  1. Great post. One thing you didn’t mention – and I wonder if you think it plays a part in single photo references – the distortion of the camera lens itself? Especially for a point-and-shoot camera, I know things that are square will have a slight spheroid-ization to them.

    The real question is: where’s the follow-up caricature of Cera? 😉

  2. wjapple says:

    it’s interesting to hear you say this stuff about using photo reference.
    I’ve been doing a lot of photo referenced celebrity caricatures and running into a lot of the issues you talk about. and i think that lens distortion does play a pretty big role, because i have done a few that were a bit off because of it.

    anyways, its and interesting insight, and I’ll have to make sure i come up with more than one reference photo from now on.

  3. yondaime_kazekage says:

    thanks for the information tom. no wonder i had difficulties to draw from a photo.


New profile pic courtesy of my self-caricature for the Scott Maiko penned article “Gotcha! Mug Shots of Common (but Despicable) Criminals” from MAD 550

Workshops Ad

Dracula ad

Doctor Who Ad

Superman Ad

%d bloggers like this: