Thursday, 20 October 2016

Arrival at Furaha IV: progress

Well, here is the next video. It shows some further development of the characters. These were painted against a white background, and what you see in the video is that their silhouettes were then painted in a flat dark brown tone. The layer of the silhouettes is underneath the one with the shading of the characters. By setting that shading layer to 'multiply' the colours in that layer will interact with the layer underneath in a specific way. Generally, 'multiply' adds portions of the images that are dark in the shading layer as additional darkness to the silhouette layer, but any light portions will not lighten the silhouette layer.
  Afterwards, I flip on the 'walls' layer as well as the planet layer, that here is simply a render from my Furaha model in Vue. To get a more painterly effect I will later paint it completely anew; actually, at the time of writing I have already done so.
  In the meantime I also reworked the walls. I liked Spugpows remark that the walls had an art nouveau feeling, so I took that literally and decorated them and the floor with Art Nouveau patterns. I will probably show the result of that in more detail next weekend.       

Saturday, 8 October 2016

Painting Arrival at Furaha III: characters


In this video I am trying to get a feel for the characters. I added a cat for some mystery; if you have a very large roomy spaceship, you can have a ship's cat too, I thought. Besides, by having it look at the viewer the painting becomes more interesting.
  In the next stage the feeling will change completely: here, the canvas is very light, but the painting as a whole will be dark. It is just easier to sketch clothing this way. I decided that the temperature aboard ship is fairly high, so people wear loose clothing. 

Monday, 3 October 2016

Painting 'Arrival at Furaha' II: how thick is the terminator?

No, this is not about the waistline of an ageing time-travelling cyborg.

A 'terminator' is also the boundary line dividing the dark and sunlit areas of a planetary surface. The Apollo missions made it very obvious that the terminator, in the case of Earth, is not a thin sharp line but rather a blurred zone in which light fades from full sunlight to total darkness. In the case of the moon it is much sharper, suggesting that our atmosphere has a lot to do with blurring the terminator. Once I had asked myself how wide it should be, I felt it difficult to simply guess its width. I thought it would be easy to find on the internet, but to my surprise it was not. So I did some simple mathematics.
Click to enlarge; copyright Gert van Dijk

In the scheme above the circle at the left is a planet, and the one on the right is the sun. On a point on the day side of the planet a viewer can see the entire disc of the sun. But when the sun sets, part of the sun's disc drops beneath the horizon; in other words, some of the rays of light from the sun cannot reach the surface. So which is the area on the planet where only part of the sun's disc is visible? That area is the terminator.

Point A above is where a tangent line from the 'top' side of the sun just touches Earth, and point B is a similar point where a tangent line from the underside of the sun touches Earth. The zone from A to B is the terminator, and it is not difficult to express that in angles. But that is all without an atmosphere. I reasoned that the same tangent could travel on through the atmosphere, where it could scatter in the atmosphere, casting some light on the surface. So I also calculated point X as the furthest point where light might be scattered.

The trick then was to put in the proper values, in units of one thousand km. The radius of Earth becomes 6.371, that of the sun 693.7, the distance between the sun and Earth is 149600, and the thickness of the atmosphere is 0.1. Mind you, that latter value, 100 km, is the 'official' border of the atmosphere, but a more relevant value would be the height where light is scattered; I have no idea.

Click to enlarge; copyright Gert van Dijk
So here is a close up of Earth with the proper values put in. The thing to remember is how large space is: the distance between sun and Earth makes all the triangles extremely narrow. The angles of point A and B are 179.7 and 180.2 degrees (the angle starts counting at the top of the planet, so 180 degrees is pointing straight down). That is only 0.53 degrees! I thought the effects of the width of the sun's disc would be larger, but mathematics doesn't lie. The angle for point X is 190.3 degrees, so the angle from A to X is 10.6 degrees. That is closer to what space photographs suggest. But how correct is it?

Click to enlarge; copyright NASA

Above is an image from NASA, rotated. I put some orange dots on it, over Gabon, that to my mind define the width of the terminator. I compared that to a map of Africa, and estimate the width to be five degrees. So apparently I overestimated the height where light gets scattered. The lesson is clear: the width of the terminator depends almost entirely on the thickness of the atmosphere. Good; I can now plug in the values for Furaha and paint the terminator at the correct width. Well, a useful estimate, anyway.       

Next time I will show some timelapse video's of the painting as it progresses (as soon as I manage to upload them, that is).

Saturday, 24 September 2016

Painting 'Arrival at Furaha' I

I will try something new: I will document how I produce a painting, meant for The Book, as it develops. I always said I would not show new paintings, as that would probably ruin the chances of ever getting the book published. I decided to make an exception: the painting I have in mind is not a major one showing an animal or plant, but a minor one showing the human side of affairs. As you will see, digital painting uses a variety of digital techniques to help find a good composition, get the perspective correct, etc. I am not alone in this: If you read books on digital painting techniques, you will find that professional illustrators do this all the time. 

I needed a minor painting for the two-page spread introducing the chapter on the Nu Phoenicis solar system and the planetology of Furaha. These chapter introductions show a minor painting and a quote, usually from Souren Nyoroge, whose book 'Furaha and Earth: similar in their differences' elevated him from one from one of The First, to Furaha's foremost historical hero (I may slip from real life comments on the painting to 'in universe' remarks, so pay attention).

The quote was about Nyoroge, the scientist, and Bruyningh, the agitator, looking at the planet after arriving there and realising that from that point on it was not just a planet, but their World. That point in time seemed a good theme for a painting and I already had a scene in mind, with the two of them looking at the planet from their spaceship, silhouetted against the blue sphere of the planet. Over 20 years ago I photographed friends of mine in the Vancouver aquarium, silhouetted against the blue waiter of a large tank with beluga whales, with vertical window frames separating the worlds on either side of the glass. I would like to show that photograph here, but have no idea where it is. No matter: a Google search for 'Vancouver aquarium silhouette' will show you that many people took very similar photographs. However, there are no vertical window frames in the Vancouver aquarium, so I must have added the notion that there were any there to my memories at some time. Memory is  malleable.anyway, the mental image, with frames, looked better than the true view without frames, so I  went with frames. The design is simple: two silhouetted people, window frames, and a planet.

Click to enlarge; copyright Gert van Dijk
I have always illustrated the sizes of Furahan animals by adding the silhouette of a human, and these humans do not just stand there, but usually do something (I have seen others take up the habit of having 'scale-humans' do something, which I like). I either draw these silhouettes by hand or base them on a photograph of someone posing, which is them modified as much as it needs to be. Above are two such recent silhouettes. One was hand-drawn from the start and the other is based on a photograph; guess which is which...

For the 'arrival' painting I could have chosen the same route, but quite by chance I came across the free program 'MakeHuman' that looked like it could be useful. It starts with a generic 3D human of which the age, gender, height, weight etc. can be adapted easily. I made a slender character (Bruyningh) and a stockier one (Nyoroge). So far so good. But then I found that you cannot change postures in the program itself, but have to use Blender instead. Unfortunately, I cannot work with Blender; I find its interface incomprehensible (it is free so is certainly worth having a look). That's the end of MakeHuman for me, then. Luckily there were some example postures included, so I used those to export the two characters. These were not what I wanted, but I could easily redraw their anatomy and pose later; the 3D objects only serve as scaffolding for the drawing anyway.

Click to enlarge
The next job was to import these persons into Vue Infinite and to construct a spaceship. Luckily, all I needed of that spaceship was a floor and window frames. I wanted very large windows to give a luxurious feeling and allow the planet to be shown as a large object. I used 'cubes' to built the frames, imported the characters, hung up a big blue sphere and added some lighting. After that, I moved the 'camera' around to play with the composition. Above you see a view of the Vue programme showing the characters on their tiny section of spaceship.

Click to enlarge; copyright Gert van Dijk
And here are a few images of the results of this early stage. I decided to go along with something like the third one. At this point, I haven't started to paint yet, and I will keep that for the next instalment. But there are quite a few things to think about:

  • I have to change the blue sphere into a proper Furaha model (I already have one). I will have to work out where the 'terminator', the line dividing night and day on a planet, should run. It has to look attractive but also has to be correct. This also includes working out how wide the blurred part of the terminator should look. The mathematics of calculating that should be easy, but I will check astronomical sources anyway.
  • The window frames look fairly boring; perhaps I should change their shape
  • The humans should get clothing; what does one wear during protracted space travel? Is it hot on a spaceship?
There are other implications: the image tells a lot about the Furaha universe. If people stand on a floor in a spaceship, they either have shoes that stick to the floor, or they have artificial gravity. I'll go with the last choice. So do almost all science fiction movies, so I doubt many people will even notice. 
  I already had a concept in mind of how people get around on Furaha: they use large lumbering vehicles that float in the air much in the same way that bricks don't (yes, Douglas Adams said that). They use a gravity-repulsing mechanism as a reverse zeppelin. A true zeppelin uses a great volume with very little mass to lift a few kg of mass; well, the repulsor zeppelin also lifts just a few kg, but using a great mass. Don't ask for blueprints. Not yet, anyway. This 'Leyden Mass Repulsor Net' (TM) works for spaceships too. Quite well, in fact, as its efficacy increases more than linearly with its own mass (I just made that up). Basically, you wrap the Repulsor Net (TM) around some mass, such as rocks, lots of water or a concrete-filled submarine, apply energy to the net and there you are.
  So all this explains why there is artificial gravity on the spaceship, why spaceship designers do not care about saving weight (the opposite, in fact), and in turn it explains why they can have large windows.

Tuesday, 26 July 2016

What do Roman soldiers, trilobites and the Furahan Droodle have in common?

Click to enlarge; copyright Gert van Dijk
Before answering that, you are probably wondering what a Furahan 'droodle' is. Well, one description found in the 'Annals of IFB Field Expeditions' reads as follows: 'The droodle is a slow an silly creature, behaving in its snug little world like the Lord of Creation, notwithstanding its utter insignificance'. Apparently the Annals did not require any semblance of scientific impartiality before accepting contributions, but that is not the point. Further on the author continues: "It can behave this way because it is well protected because of its foul taste and because of the overlapping armoured plates covering nearly its entire body."

An early version of the droodle is shown above. And there we are: the droodle has an armour consisting of overlapping plates, and so did trilobites, and so did Roman soldiers. I came across the subject when I was preparing to paint the droodle anew for The Book. Most of my new paintings have very little to do with the old ones, but I like some old designs enough to go over them again, taking the opportunity to improve them in as many ways as I can. I started wondering how animals manage to move while covered with what seem like very stiff plates. How are these plates attached to one another? Obviously, in arthropod legs the exoskeleton of adjacent parts of the leg form joints that often have just one axis of movement, much like our own knees: we can bend it stretch a knee but it does not move sideways nor can we rotate the leg and foot backwards. I started thinking about whether that also applies to the plates covering the droodle.

Click to enlarge; copyright Gert van Dijk

Above you see the result of simple experiment: I wanted to form successive hoops curving around the animal's back while widening at the sides. I imagined a hinge between two hoops with the axis of rotation about halfway up the animal. Of course, such an axis of rotation would make sideways movement impossible, but so be it. I assumed that the plate in front would slide over the plate in back. I could imagine that in my mind's eye for half-circular hoops, but felt I needed some visual help with hoops that widened at the side: could they in fact slide over one another over their entire length, or would they intersect, making the movement impossible? So I made a rough shape like that in Vue, of which the top surface represents the plate. As you can see, at the centre one hoop can easily slide under the next one while it wants to move over it at the sides. This does depend on the site of the hinge and some other aspects, but it does show that you cannot assume any angle or shape to work. I could of course still paint it the intended way and no-one would be the wiser. But science has preference over art in such matters. So what was wrong?
Click to enlarge; Manton, The Arthropods 1977
I then thought of trilobites: their name indicates that their bodies had three lobes lengthwise, with a thick part in the middle and much narrower side flanges, much like the droodle. And trilobites could roll up their bodies, so they solved the problem how to slide one hoop under the other better than I had (weel, they had more time...). I obviously needed expert guidance, perhaps a book called 'Biomechanics of trilobite intertergite movement' (the hoops are called 'tergites', just so you know). I found something close: 'The Arthropoda' by S.M. Manton, 1977. I am not new to reading scientific papers, but this book is as intricate as it is condensed: the reader is assumed to be rather well-versed in arthropod classification and anatomy. The book contains sentences like 'There were no coxal endites or gnathobases.', in its own way as wonderful as 'It was a dark and stormy night'.

Click to enlarge; Manton, The Arthropods 1977
So here is Figures 1.5 from that book: have a good look at the top right in particular: the tergites are connected by a fold of skin, doubling back on itself. There is no hinge to be seen anywhere. If the tergites are indeed connected only by such a fold over their entire length, they would have much greater freedom of movement than if there were just one axis of rotation: this is a good idea.

I then wondered if his is how all armour segments are connected in arthopods, and browsed through the book. As you would expect, there are a myriad adaptations of tergite movement. In species that burrow, successive tergites are kept from sliding over one another and have a built in 'door stop', allowing the animal to push the soil out of its way. In other species there are additional small tergites normally hidden between larger ones. When the body is flexed, the gap that would otherwise appear between the large ones is filled by the small ones. In many cases tergites only cover the back of the beast. There may be other bits of hoops at the belly (sternites) or the sides (pleurites), all of which are connected to one another by the folds of skin.

Click to enlarge; from The Arthopods, SM Manton 1977
Here is an example of the intricacy of the internal anatomy of a millipede. The large top image shows the poor millipede cut lengthwise with its body flexed (the back is at the top). The lower left image is a horizontal section of the body rotated sideways. Complex, aren't they? The feature I would like to call your attention to is the folding of the skin between tergites: I found that in all such plates.

So I learned from all this that arthropod tergites can in fact be connected by 'hard points', but in many cases the skin folds allow flexibility and freedom of movement. So that was one problem solved, but all this did not answer the question how to ensure that the tergites do not 'intersect' while rotating, as they did in my simple model? Or how do you avoid having large gaps form when the animal moves? There may be several answers to these questions. Perhaps the tergites should be flexible.


( video does not seem to work; I will check later...)
Have a look at the YouTube video above, of a millipede flexing its body in all directions: the tergites do not show any gaps at all, and yet they slide over one another in at least two directions of rotation: they must be flexible. But would that work for a big animal, in which you would expect the tergites to be stiff? (but never brittle: the armour must be capable of withstanding blows, and allowing it to deform it a bit should absorb the energy of a blow). 

Another solution would be to forgo tergites that run from one side of the animal to the other; split them up in separate parts instead. These smaller plates could each be tough, and be connected with skin folds. At the top of this post you'll find a simple model I made to see what a droodle designed in this manner might look like. Mind you, this is not what the painting look like: the droodle has already evolved some more and no longer looks like this, but it does still have multiple overlapping tergites.

Click to enlarge; source here
So what about the Romans? Everyone who has ever seen a film with Roman soldiers in it, or who has read an Asterix book, knows that their armour consisted of metal hoops circling the soldier's body. This seemed very similar to the tergites of trilobites or other artropods, so I wondered how the legionnaire's hoops were connected, That was easy to find out: there are books describing actual archeological finds. The image above is from tha reconstruction based on such finds. The Latin word for this particular type of armour is 'lorica segmentata' ( I also leaned that all these films might be wrong: this plate armour type may not have been the standard type of armour; chain mail may well have been more common.) And here is how the loops are connected: not by hinges, but by leather straps on the inside of the hoops.

( video does not seem to work; I will check later...) 
The video above shows the Roman lorica in action (the inside is well visible two minutes into the video). The hoops could slide and rotate a bit with this arrangement, in exactly the same way that the tergites of trilobite could slide over one another thanks to being connected by folds of skin. Only the trilobites had their armour some 520 millions of years before the Romans invented their lorica.

So this is how the droodle came by its armour and by its scientific name of 'Lorica segmentata'. There is a long list of items the Romans did for us, to which I would humbly like to add that they can make you think how exoskeletons work. Not a bad thing at all. 

Saturday, 11 June 2016

The anatomy of giants in 'Game of Thrones': did they get it right?

Every now and then a giant shows up in the television series Game of Thrones. Obviously, GoT is not a nature documentary so there is no reason to be difficult and analyse it scientifically. After all, it has dragons that cannot fly but do. But as their eggs only hatch after having roasted in a funeral pyre there is magic involved, and that should do the trick. Still, every time a giant ambles across the screen, the science routines in my mind spring into action and start wondering about its anatomy.

Click to enlarge
The giants do not just look tall but broad as well, and that makes sense. Their legs seem extremely broad and columnar, as wide as the feet. The feet do not even stick out much in front, unlike ours. The outer shape of the clothing suggests that the giant has feet like an elephant, with toes and a big elastic pad in the sole, all encased in a cylinder. But all those rags and furs hide the giant's shape so it is hard to tell. Perhaps the actor is standing on high heels or stilts to add some height, and the wide trouser legs are meant to hide that. The leg/foot assembly does look rather long.

So every time I see a giant I think: 'That looks convincing; I wonder whether they asked some biologist how to make a giant'. And then of course my attention is diverted away from that by someone being murdered unexpectedly in a gruesome manner; it's GoT, after all. In this series people die all the time, and with some recent story developments they might even do so more than once. Recently (series 6, episode 7) a giant stood still in one shot, together with some puny humans. His name is apparently 'Wun Weg Wun Dar Wun', in case you wonder, or 'Wun Wun' for short. I thought I should get that frame to have a closer look, after the killing would be over (temporarily, that is).

Click to enlarge
So here is the shot in question. The giant appears twice as tall as the man on the left, who I suppose is normal height. I estimated that man's height to be 1.8 meters, so the giant would be 3.6 m. tall. Some internet searching revealed that the actor playing Wun Wun is over 2.1 m tall. What the producers  did was to use low camera angles to simulate his size. Here is a video of that, but be warned: Wun Wun is visible for only a few seconds.

Click to enlarge
I copied the giant, and reduced his size to what should be about 2.1 meter, compared to the 1.8 m man. I stuck him in again to get a feeling of what the actor may have looked like without special effects. Here is the result of that, including doing away with the blue overcast. It does not look as if he is walking on stilts, so the cylindrical shape of his legs is probably there to suggest just that: big legs. He looks very broad; should he? What should the proportions of a 3.6 m tall giant be like?

To find out, please read the first and second post in this blog on why body size matters. In short this is what happens if you double the length, width and height of an object or animal: its mass and hence weight will not become twice the original amount, but eight times as much. But the strength of a bone is given by its cross section, and if you only double the radius of a bone, its cross section becomes four times as large. But the bone needs to be able to support eight times the weight. This means its radius has to increase disproportionally: if the weight increases eight-fold, then the cross section needs to increase by a factor eight too. The radius needs to increase by the square root of eight, which is 2.83 times. In case you are dazzled, it boils down to this: increasing bone length by a factor of 2 means that bone diameter has to increase by a factor 2.8.

Mind you, in real life that is probably not enough ('real life'?; what am I thinking here?). Muscle strength also depends on its cross section, and to keep the same relative strength means muscle cross section has to increase by a factor 2.8 as well, meaning more muscle mass. All these extra increases in bone and muscle will add mass, so the bones have to be even thicker and... You see where this is going. At some point there is no mass left for lungs, guts or brain (judging from Wun Wun's speech patterns, some savings were indeed made in the latter department). Say we use a factor 3.0 to accommodate for all that.

Click to enlarge
Here is a nice schematic human anatomy image found on the internet. I cut it up in sections and increased the width of the sections appropriately. Well, mostly: I reasoned that the increase should certainly apply to the thickness of weight-bearing bones, such as legs and the vertebral column. I did the same for the arms as well, or the result would look silly. But in principle pelvic and thoracic width do not need an additional increase beyond the factor two we started with. I did increase them some more, if only to give thigh muscles some room.

Click to enlarge
Here is the result of this simple attempt. Do not forget that he is twice the size of the man we started with. The resulting giant certainly looks large as well as stocky, and so he should. Note that I increased the feet a bit more, to allow all the thicker foot bones to lie next to one another, as they must. But I did not alter foot anatomy any more. Before we discuss that further, let's approximate the giant a bit more by using the body outline and adding a suggestion of clothing.

Click to enlarge

Here it is. This is what a morphologically reasonably sound 640 kg giant might look like.

Click to enlarge
And to get a feeling for size, here is the giant next to a normal-looking human on the left, half the size. The giant is not that different from Wun Wun, confirming my intuitive guess that the designers got it right. The hands look quite good. The actor must have been covered in very thick layers of padding to get the stocky look. It is not often I get to write that the television or film industry got their biomechanics right. Regular readers will know that there was little reason to be happy before; see the posts on Avatar and John Carter of Mars.

Click to enlarge

The one remaining matter is whether the giant should have his feet examined: can he stand on human-type feet or must he have elephantine feet, as shown above? If a 1.8 meter m. tall man weights 80 kg, a 3.6 m. giant will weigh eight times that, so 640 kg. Being bipedal, one leg has to be able to withstand all that weight. Mind you, we already have taken that into account as far as bone strength is concerned. Still, that is a lot. But Wikipedia tells me that draft horses weigh up to 1000 kg and giraffes weigh up to 1930kg, and these animals do not have elephantine feet. While running just one of their legs may be on the ground, with a lot of dynamic forces acting on the bone as well. So I do not think a 640 kg humanoid needs elephantine feet, but that does not mean he cannot have them. All this brings up the topic of whether very large bipedal animals should have elephantine feet with embedded toes, or whether they should have free toes, sticking out. I am in favour of the latter, but that is something for another post, on toes.    

Unless the GoT designers tell us what Wun Wun's feet look like, we will never know.  Unless... unless of course we get to see a naked giant. GoT is not afraid of nudity and there have been calls for nudity to be more equally divided among male and female cast members. Perhaps that equality should include not just sexes but species, too. So let's have a naked giant; purely for scientific reasons, obviously. Actually, just the feet would be enough, thank you so much.      

Tuesday, 24 May 2016

Painting a Cthulhuoid carapax (Digitally painting Furahan lifeforms II)

As blog titles go, this one is not likely to win the prize for 'succinct clarity'. Actually it could, provided readers already knew what a Cthulhuoid was, what its carapax was, and why you would want to paint said carapax. To make matters worse, there formally never was a 'Digitally painting Furahan lifeforms I'. But one earlier post would in retrospect deserve that title.

What does the title mean? Well, the word 'Cthulhoid' describes a clade of marine Furahan animals that do not seem to be able to make up their minds whether they should be pelagic of benthic ('pelagic' refers to the 'just water' part of a sea or lake, not close to the bottom nor to a shore, while 'benthic' refers to the bottom of a sea or lake). Some Cthulhuoids use the tentacles close to their face -hence the name- to move around on the bottom or even create their own hiding places, while others use their fins to propel themselves through the sea. A 'carapax' (a term I prefer over 'carapace') is an animal's armour, say its shell. The cthulhuoid carapax covers the head and usually part of the back. Finally, why would you want to paint one? You, the reader, might in fact not want to do this at all, but I wanted to, to create an image for The Book.

I am not going to spoil The Book by showing major paintings here or anywhere else, but I can make an exception for part of a minor illustration. The illustration in question shows a few examples of the riotous array of colours and structures of cthulhuoid carapaces. The problem with 'riotous' colours, in stripes and spots, is that every spot must be painted in the correct shade for where it is on the object, and that includes different shades within each stripe or spot. With oil paints this proved to be a painstaking job, requiring small pointy brushes, a very steady hand and lots of patience. Digital painting has made painting such complicated objects much easier, as I will illustrate here. I will assume some familiarity with 'layers' (in digital painting, a layer is like a sheet of glass: what you paint on it covers things on underlying layers, but parts unpainted on a layer let you see underlying layers. You can paint on a layer under another layer. I use Corel Painter because it can mimic real brushes quite well.

Click to enlarge; copyright Gert van Dijk

Let's start with a suitable carapax shape. I modelled one in Vue Infinite and made a simple render in which the 3D shape is overlaid with simple lines that define contours of the shape. These help get the perspective right, in a fraction of the time that a conventional perspective construction would require.  On a separate layer I drew lines with a 'brown pencil' to outline some interesting spots, aided by the lines that help keep the 3D shape in mind, and also help ensure symmetry.

Click to enlarge; copyright Gert van Dijk

The next stage uses a layer under the brown pencil one. This new layer contains the basic colour of the beast, which in this case means dark blue sides with a lighter colour down the middle. Note that I made absolutely no effort to represent shading here: the colours are supposed to be completely flat.

Click to enlarge; copyright Gert van Dijk
I then added another layer, again just under the pencil layer. On that one I painted the spots an even deep yellow colour, to contrast with the blue underpainting. Again, this is completely flat. Note that the result contain three layers: the pencil lines, the yellow spots and the blue basic colour.  We will leave these layers for now and hide them from view.

Click to enlarge; copyright Gert van Dijk

Using the same Vue render as before I then painted the carapax again, but this time without colour, using just shades of grey to convey a sense of depth as well as a surface texture with same plates on it. I rather like the way the shininess turned out: the surface is shiny, but more like a pearl than like chrome. The shininess should allow the colours to remain well visible.

Click to enlarge; copyright Gert van Dijk
The trick now is to combine the flat colours with the grey layer defining the shape. There are at least two ways to do this. In the one shown above, the grey 3D layer changes the aspect of the solid colours below. There are many ways in Corel Painter or in Photoshop of making one layer affect an underlying one. It is often hard to predict what they do as their names often make limited sense. The result shown above was obtained by applying the grey '3D' layer to the underlying flat colours as 'hard light'. Not bad, is it? You may note that part of the 3D structure indicated by the grey layer is obscured by the strong colours. That is very often the case with strongly contrasting patterns.

Click to enlarge; copyright Gert van Dijk

This particular image is based on the opposite approach: the grey 3D layer was used as the underlying basis and the colour layers were moved on top of it, where they affected the grey layer through an option labelled 'colorize'. As you can see the result is not the same, which is part of the fun of digital painting: there are new options to discover daily. Of course, it may be better to stop discovering them and get to work at some point, or you will never get any work done.

Click to enlarge; copyright Gert van Dijk
Finally, I went back to the earlier version and decided to change the colours on the flat colour layers, which only takes an instant. The blue basic colour became solid yellow, and the yellow spots turned black. The grey layer is again used to provide a 3D aspect to the carapax, but this time I turned down the 'hard light' effect so the highlights are less conspicuous. I then added two shiny spots with fairly sharp edges to get a surface effect like porcelain. On yet another layer I painted flat white regions at the edges of the carapax. These were then made almost entirely transparent to represent reflections of lighter objects in the vicinity. I present this version here to show that separating colour and structure allows for some quick experiments. It is not the way I paint most often though: usually I paint shadows directly, using appropriate colours.

To paint other shells I did not use this method, as I thought that using the same outline every time would make the result boring. Instead, I designed and painted a new shell from scratch each time.  

And there you are; a painted cthulhuoid carapax. The illustration should end up as probably about two by two cm, so it will be small. This particular carapax belongs to the species Myrmillo testudiformis, or in common speech the 'turtleback snigel'. Such shells are collector's items, by the way.