Digitigrade leg designs from the past

William Bouguereau, Nymphs and a satyr, 1873

Bouguereau cared for composition above all else. This satyr’s legs here are like transplanted chamois legs and look kind of believable in the still image, even though they would probably look unnatural in motion. The fur coat starting halfway down the thigh also feels contrived to me (nice excuse to show the satyr's fully human buttocks!) but aesthetically it certainly works.

Elihu Vedder, Marsyas Enchanting the Hares, 1899

The legs portrayed in this picture are easy to the eye and give a good outline and balance to the satyr’s figure, but they don’t make sense from the anatomical point of view. The calves look like they have no volume and are somehow fused with the thigh. The shaggy fur isn’t enough to conceal the inconsistency, but the legs aren’t really a focal point of the picture so it’s the overall impression that matters. The figure’s outline is very similar to a common hieroglyph depicting a sitting woman, I suspect it is an actual quotation:

I run into the above problem all the time - I have a nice pose in mind but it would force the legs to bend in impossible ways. I used to try all sort of weird leg designs to keep the anatomy as consistent as I could but ultimately I decided it wasn’t worth the effort. In most cases I’d rather use a leg design which feels balanced and has a good outline even if it’s anatomically impossible.

Even though I study my fair share of anatomy I'm not veterinarian and designing a truly consistent model would require also require the kind of 3D simulations used for big budget games and movies. In fact there are some cool digitigrade designs in games and movies (the minotaurs of Narnia come to mind) but I can't remember any instance of them exploring the full extent of the mobility which an anthro body with digitigrade legs could have, probably beacuse it's not required by the media. For them creating believable walk/run cycles is the most important thing. On the other hand a lot of anthro art is still pictures of nudes and sex scenes in which the nuances of what a limb can or cannot do are more relevant.

Carlos Schwabe, The Afternoon of a Faun, 1923

Same leg design as Bouguereau. Here the weird leg anatomy is exploited well for dramatic effect, they are supposed to look uncanny. A human in the same pose would look awkward rather than impressive.

Unknown artist, Pan and Daphnis, III-II century BC

This is one of the most believable designs I could find in ancient art and it’s similar to what is expected of realistic anthro art today. The thighs actually look like they are part human and part animal, and the legs seem to bend just like goat legs. They are also shorter than human legs compared to the torso, which would give Pan a low center of mass helping with balance (balance being the biggest issue with the idea of bipedal digitigrade legs).

Some thoughts on portraiture of anthro characters

A small oil portrait of the character Tedros, done for Anbessa on Furaffinity.

There used to be a lot of rules on how to do official people's portraits before impressionism and the modern art era. One rule which stuck me as odd is that smiling was strongly discouraged because it could distort facial features and would make the painting tiresome to look at for a long time. It is true that smiling expressions can be annoying to look at unless the smile is very subtle (e.g. the Mona Lisa) or the whole painting is very stylized (e.g. portraits by Boldini, the impressionists, caricatures, etc.). Also telling the model to keep a neutral expression allowed him or her to relax and display their most natural attitude.

I have no idea whether this concept may apply to anthro characters. After all there is no real model to copy from and many animal muzzles don't look very expressive to untrained eyes, especially when they sport a neutral expression. But it's an interesting challenge because neutral expressions are always hard to draw convincingly, it's easy to make them look dull or silly.

Even if you draw a realistic anthro muzzle, the most straightforward way of making it expressive is to map cartoon expressions like these onto its features:

This kind of "standard" expressions can actually be applied to anything (animals, objects, etc.) as long as it has the equivalent of a mouth, two eyes with visible white and a discernible pupil (so we can understand where the gaze is pointing) and very mobile eyebrows. A lot of RL animals look like they have no expression just because the white of their eyes is not visible and they have no facial features similar to eyebrows, while animals like cats and blue-eyed huskies feel very expressive because they have both.

I often abuse eyebrows to exaggerate expressions in my pictures and I know sometimes I'm guilty of stepping into the 'TUDE eyebrows minefield. So in this picture I downplayed eyebrows a lot (though the dark patterns around the eyes work fine as proxies) and I tried to keep a neutral expression. I'm quite pleased with the result. Now I have a few other ideas to test in small portraits like this one...

Human studies (with some intruders)

I've started the great attack against my biggest limit... proper human portraits. These will be mostly based on studies from http://www.female-anatomy-for-artist.com/ since the site has an excellent archive of models of all body types and of different races. I don't want to learn on idealized human bodies since I already tend to idealize too much. I'll post most on my CA sketchbook only like the quick animal studies.

A small treasure from the past

I'm amazed at how little printed information exists about the anatomy of wild animals and how hard it is to find it. Any university bookstore has books with anatomy schemes of cows, dogs and other domesticated animals, but schemes of muscles or bones of any wild species are rare, often only found scattered in specialized reviews. Over the years I've collected on internet several gigabytes of anatomy schemes, skeleton photos, dissection photos etc., and I've bought the best books I could find on the topics. I've found stuff about some really exotic species but not as much as I'd like to. In order to find information on penguin fin muscles I had to dig them out of the Challenger Reports, the logs of an XIX century explorer ship, which also contain a few thylacine dissection drawings - very few of them, possibly the only existing study of this kind, of a species which nobody will be able to study again...

The rest of the Challenger drawings is here (links to chapters are towards the bottom of the pages).

Puffins are regularly eaten in some countries but good luck finding a scheme of their wing muscles. I could only find very crude schemes from a cute study about their swimming motion. I see it's not very useful information for most people, but it's kinda... embarassing? To discover we are so abysmally ignorant about many animals we like a lot, like penguins. We kill them by the milions and treat most of the corpses as trash, yet it seems nobody bothers to take a closer look at the little wonders.

Alex the Mastermind

The initial idea was to send this as a presente to Dr. Pepperberg as a token of appreciation for her efforts to understand animals. Maybe I'll do it anyway, but I'm not sure this version of the image is appropriate for that. As much as I like anthros they are not the answer to everything and this parrot just doesn't feel like an anthro version of Alex.

I like to study anatomy thorougly and so I have always played with the anatomy of the anthros I draw and invented unusual designs like this one. I especially like functional designs which keep key features of the original animal but don't show stuff which is blatantly against physics (such as wings on a human-size creature) or would be very awkward for a real creature. That's why the owl in Harvest Moon has a birdlike body with backward knees, vestigial wings and fake beak over a human mouth. Such designs are cool for original characters like that one, but in this case, the more I look at it the more it feels weird.

When playing with anatomy with a realistic style the risk of falling into the Unbcanny Valley is always high and I think I crossed the line with this parrot. Maybe it's the beak - I left it as in the real bird because it's Alex's "face" I wanted to show after all, but maybe I should have gone all the way and used a vestigial beak like I did for the owl. Maybe it's that the legs bent in that odd way are too promintent (and very different from those of a real perched parrot) and look Exorcist-y. Maybe it's that the coloring is not very good, I did this before learning the details of color temperature and also I didn't pay enough attention to texture, and as a result the feathers look like octopus skin, and are too much detached from the rest of the picture.

I'll let this rest during vacations and give it a freshed look later. If I still have this impression I'll do a remake with either a different anthro parrot or a with a regular parrot.

A hooting dancer

I've been out doing some much needed drawing from life and other things. I'm getting more familiar with birds but I need to pay a lot of attention when I sketch, if I get too carried away I become careless and I begin fucking up proportions and key features of the animals. I sketches this dancing barn owl while keeping an eye on many references to avoid that. (The wing shapes are not very accurate, I just wanted to imagine them as large flat shapes as it is more useful to imagine how they could move. I tried to preserve the shape though. Owls have a quite dinstinctive wing shape when seen from above and below.)

Now they have penguins and camels in the nearby zoo. Sketching penguins with ink is great fun and interesting, I'll need to do many more next time.

Studies from 2009 - Breaking up cetacean profiles

Quadrupeds are hard to draw because their shapes are very complex while birds and cetaceans have the opposite problem: their shapes look very simple, but they are actually full of subtle curves, and without such subtleties they lose a lot.

Their profile may look a like it can be drawn with a single continuous line. Sometimes it's possible, but the result often feels stiff and lifeless to me, it's just too "designy"/oversimplified. This happens specially in dynamic poses, as real cetaceans have muscles and fat tissue shifting all the time and creating small changes in their profile depending on the movement they are doing.

So I've become used to sketch cetaceans and birds trying to break up the profile into many lines, obviously trying to keep them consistent with the actual variations in the curves of their body. Trying to keep them tangent to the curves seems a good strategy to study their bodies and get a feeling of where the subtleties lie. The results seem lively enough in spite of the hard angles.

Studies from 2009 - Quadruped quick sketches

I wrote a previous note about gesture drawing for animals, but that example was about birds which are easy to sketch quickly, though it's very hard to draw them accurately or in unusual poses and the wings are a puzzle on their own. With quadrupeds even getting a basic shape right is more difficult.

At the beginning my biggest problem with quadrupeds was that I saw too many things going on - too many limbs in motion, too many lines to follow, too many masses:

When doing gesture drawing of humans the most important line is usually the backbone, and all other lines follow logically from it. Sometimes it is obvious in quadrupeds too:

But I look on purpose for weird poses where it is hard to figure out which lines are the most important for the overall shapes. Giraffes have unusually long legs, the neck is a unique feature which changes a lot the visual balance of their body, and on top of that their movements are very constrained. So trying to make giraffes look agile and not stiff is an interesting exercise.

(Warning: click for slightly NSFW giraffe)

The little figures on the right are a possible way to sketch quadruped poses considering only a few main lines and masses. I did a lot of them too and it turns out they are also a good exercise in synthesis, to train oneself to remember well the proportions and two or three key features of a species. The can be useful to study features at any level, for example to study small differences between species which look very similar, as in this example with antelope horns:

Owl feathers and facial masks

Studies for the crucified owl painting and the current one.

Some thoughts on sketching feather coats. Even finding good ways of making thumbnails is hard with all the variety of coats, limbs, etc., but I think that certain feelings which are typical of an animal's appearence should be given even in the crudest doodle. Birds often give the impression that they are covered by few broad sheets rather than many individual feathers, so I prefer to sketch them like that.

The facial mask is the most typical feature of a barn owl. From the front is looks almost flat, but from the side it reveals a quite complex 3D shape.

(from Wikipedia)

I don't like drawing anthros with a true beak, so when drawing anthro birds I usually give them a mammal's mouth with the upper part of the beak on top of it. This gives them a very flexible mouth while keeping the appearance of a bird or a gryphon.

But for barn owls the proportions and the shape of the mask are important, they are a big part of the appeal to me. Barn owls also have a few feathers on both sides of the beak which hide away most of the mouth, so that the mask surface is smoother and it can do a better job of focusing sounds. For me small functional details like that one are part of the appeal of animals.


The head shape I've settled on is more or less this:

Studies from last year - More line of action ideas

That's a tern photographed with a high speed camera. She was caught during an extremely complex movement which involves most parts of her body, and that's the kind of motion which makes a tern look like a master of flight and not a chicken. They are probably the most skilled fliers in the world along with hawks and owls.

How do you even begin to study poses like this one?

These were my early attemps to extract LOAs from photos of birds (last part of the batch of sketches with crappy text, I swear!):

When birds are perched the tail is often an important line.

I tried to reduce each pose to a minimum number of lines (in red, the red number is the lines count). But the resulting stick figure has nothing to do with the original, if you look at the red lines alone you could never tell they represent a bird during a wing beat...

Often the most visible lines of birds in flight are the edges of wings and in fact birds are often simplified like this:

That's easy to recognize but also very anonymous. The supreme elegance of the tern's wingbeat is gone.

These are new from yesterday, tried again with the tern photo:

BIRDS EVERYWHERE

The right one is the best I got so far, I'd start from a scheme like this to design a tern in flight. It's very different from a classic LOAs scheme, but it's just three lines and it keeps the lines and proportions I see as most evident in the photo. The left one looks odd because the wing line on the right is too parallel to the body line; both the left one and the center one show the body too large so it doesn't look like a tiny bird at all.

Studies from last year - Line of action for animals

I first learned about the "Line of Action" (shortened "LOA" in my notes) reading John K.'s excellent animation blog ( http://johnkstuff.blogspot.com/ ). The concept is summed up in this picture by Preston Blair:


Arguably the most dynamic-looking poses are those which can be synthesized with just one or two lines of this kind. There is much information about using LOAs to draw human and human-like bodies in motion, but I couldn't find much about using it for animal motion, such as running horses or birds in flight seen from od angles.

So I found it very instructive to study animal photos and decide out which lines stand out in each pose for me. They are the lines which might serve as LOAs for drawing that pose. (Sorry for the unreadable notes in the pictures - these were the last sketches I did like that before sticking to uppercase.)


Solid red lines are the lines I see most prominent in the original. Dashed red lines are the ones I used, which I thought made the pose slightly better looking.

Things I noticed in these first experiments:
- Drawing with LOAs in mind is useful when using photos as references, to avoid the temptation of just copying exactly the photo.
- If at the end of the main line there is a part of the body pointing in a different direction, e.g. a muzzle or a limb, it adds to the feeling of movement. See the bats and dolphin. If the muzzles followed the solid line the bodies would look a bit too simple.
- Making a LOA turn into a spiral is necessary to draw certain animal bodies, especially when odd tails are involved. There is no way to draw a nice chamaleon using just the Blair rule of slightly curved lines...