Tuesday, May 24, 2016

Making Sausage: An occasional discussion about populating the world's most obscure webcomic.

Figure1: The Masters Sisters 5.0

And now for something completely different*  To satisfy that little bit of art geek that's inside all of us, from time to time,  I'm going to talk a little about how the material on this site gets created.

It's safe to say that the artwork has evolved a bit over time.  Figure 1 shows the current iterations of two of the main characters.  I've finally gotten them to the point where they appear as I've always imagined them.  It hasn't exactly been a smooth journey.  Image 2 gives a pretty good indication of the evolution of the artwork using the eldest of the sisters as the example.

Figure 2

Since this thing puts no money into my pocket, I've had to make due with the drawing tools on hand, not the ones I'd like to use. So, when I started this site, Windows Journal for the tablet version of Windows NT was the only graphics package I had to work with.   I'd draw a crude image in Journal with a stylus and import it into any program where I could add color.

That's how the first images of Venn above were drawn.  (The characters of Venn and her siblings are a homage to five amazing women I know.  They've sort of taken over the strip - much like their meat space inspirations.)

Enter Keynote:   The Journal program didn't last long, fortunately.  While working on a slide show for work, I needed to draw a blob for a background.   Hmmm, said I, when it turned out that Keynote, no doubt reflecting the early artsy nature of Apple, had a vectored graphics element.  Now, it's the only program I use to create the panels along with a MacBook and the built-in finger track pad on the laptop.

Vectored drawing programs:  There's a lot more to it than what I'm about to relate but a huge difference between vectored and non-vectored tools has to do with the ability of the later to dynamically maintain image fidelity when you expand or contract the size of the picture.  With a non-vectored image you're dealing with a fixed image like a photographic negative.  When you print the image you're stuck with the fidelity and clarity of the source.  It might look fine as a 5x7 but look blurry blown up to 8x10 and beyond.  A vectored image, however, would have the same clarity at 8x10 as at the smaller sizes which is something you can appreciate from the image below.

Figure 3

Detail remains crisp regardless of the image size - to a point.  Keynote is no Adobe Illustrator.  There are limits to how compressed or expanded an image can be dependent on the complexity of the source image.  Venn 2.0 only had 28 image layers so it tended to be more extensible that Venn 5.0 with 20 layers just in the right eye.

Keynote was a great improvement on the Journal program.  The images were crisp, edges clean, and colors much richer.  The middle image from figure 2 was one of the first images I created with it.  (Unfortunately as is often the case with software, some engineer thought they knew better and subsequent versions of Keynote are less useful for my purposes.)

Mix and Match: Since I have to fit my artwork into my schedule (including times of insomnia) and I wanted to publish fairly often I went with a 'paper dolls' approach that's part way between the 'constrained comics' fixed artwork approach of 'Dinosaur Comics' and fresh custom art for every installment.  So I guess you could call it semi-constrained. I've built a series of image libraries (heads, bodies, parts and accessories, backgrounds, etc.) for each character that I (usually) use to construct the panel art.  

A whole supporting cast of foils and henchmen await their turn to be the butt of the snark. The henchmen kind of get the 'Cloud Atlas' treatment with changes in clothes, hair color, facial hair, etc., allowing me to use them in various guises.

With a good set of dialogue in mind I can usually pull one together in 15 or so minutes.  Sometimes I'll just pound out a new styled image for a panel like the one from 5/12/16 if I need something different but the new stuff gets added to the libraries for later reuse.

Since the comic is made from image libraries I have to be careful of the same problem that corporate logo designers face - a full screen image needs to render well as a 1 cm image too.  

Rendering the main characters:  For reasons I'll get into sometime, the main characters need to be pretty realistic but this isn't photo-realistic portraiture so economy is important particularly since I need to produce a lot of images.  Fortunately there are a couple of shortcuts that allow me to have my cake and eat it too.  I operate on the assumption that when judging intelligence, emotion, beauty, etc. we tend to first notice the facial triangle of eyes, nose and mouth and the general outlines of the face and apply less emphasis on other features (Assuming they are fairly typical.  Put a horn on someone's forehead and that's what they'll notice.) 

Figure 4

Referencing Figure 4 you can see that with the first versions of Venn using Keynote, outline and the facial triangle was about all I worked into the image.   The later image adds the hint of a lot more detail but that image also lives or dies on the quality of the facial triangle.

Drawing order: Eyes, mouth, nose, facial outline, hair, body and posture.  Each area is slightly less detailed than the one that preceded it.  The eyes nose and the mouth define the face, but I spend 3x more time defining the eyes than I do the mouth and 2x more time defining the mouth than the nose.  The remainder of the face takes less time than one eye.  Take a look at Figure 1 again because it illustrates this perfectly.  The individual women play realistically because their faces do.  And their faces do because their eyes do.  But the further you get from the face the less detailed the image becomes.  The arms and hands appear realistic but only because the outlines are correct.    The hand next to the one character's face is more detailed than the one at her waist but it doesn't diminish the effect and I doubt many would notice on their own. The viewer's brain fills in the missing details.

Figure 1 Redux

One thing I wasn't happy with about the early images of Venn was that the face looked too young.  Venn was supposed to be middle-aged not in her 20's.  It took me a while to master Keynote as a drawing program enough to start creating image maps complex enough to age the characters without looking odd at the sizes used in the comics.

As I said,  the new main character image sets are supposed to be middle aged for the most part, with no plastic surgery or Botox - like the amazing women who inspired her.  I'm not drawing for movie studios so natural aging isn't a bad thing.   Figure 5 shows the pretty simple highlights I added to achieve the appearance of age I was after.

Figure 5

That's the thing with faces - and why facial plastic surgery often doesn't work out - the subtlest of features can have a tremendous impact on how a face looks.  These additions to the image are really mild but you can see how they 'age' the face.  By the way, there's another thing about this image that adds to its realism - do you know what it is?  Did you consciously notice?  It's asymmetry. One eye is slightly more droopy than the other.  The nasal contours aren't exactly mirror images across the midline - like a real face.  We don't have a perfectly symmetrical body plan.  It's just MOSTLY symmetrical.

Art is telling a story without words.  Telling that story is easier if you key into existing perceptions/prejudices that the viewer brings to the party.  Sadly, these perceptions are often gross generalizations that reflect something shallow in our collective thinking but there are only so many battles a single panel can take on.  Making the images of Venn take on a personality and convey convey intelligence is a great example.  Figure 6 points out a few of the primary attributes added to the image to make her look intelligent (the viewer obviously may have differing, albeit erroneous, opinions)

Figure 6

The eyes: There needs to be a baseline level of tension in the eyes (lids).  Glassy-eyed, Botoxed or doe-eyed doesn't really work here.  If there's no expression in the eyes, it's best to trash the image and start over.

Brow line:  Bangs don't help.

Strong nose:  A strong handsome (even somewhat equine) nose has always been associated with strength.

Svelte:  Our culture does not automatically look favorably at those with a heftier girth.  A single panel cartoon by its very nature is an exercise in dealing with first impression limitations.

That Mona Lisa smile:  With the right smile and the proper sparkle in the eyes, you can just see those gears turning. It's that shit eating grin people get when you know they are trying to politely choke down a smart-assed thought about something you just said.  You know that you might regret it, but you just have to know what they're thinking...

Ok, that's probably more than enough for now.   Let me know if this is in any way interesting.


* with gracious thanks to Monty Python