Making Of 'The Lightning Mage'
In this Making Of I am going to talk about the steps it took to go from initial idea to final concept for my character, The Lightning Mage. This Making Of will cover all steps leading up to the final 2D design and will not talk about the creation of the final 3D model. However 3D will be used to help us conceptualize our ideas more quickly and easily.
The goal is to create a piece of concept art that you can use to create a 3D model. It is kind of a pre-visualization tool on how you want the final render to look. Important focus points are the shapes of the character, the implied textures and the overall mood dictated by the lighting, not the level of polish.
To contrast my usual preference of making hulking, brute, warrior-type characters, knights and other physically strong characters, I decided to make a character that was physically not so strong but had very specific magical abilities. My goal was to have lots of Asian influences in the character because normally wizards are associated with the picture of an old man complete with the pointy hat and the white beard.
The initial idea was to make a character wearing bright robes with intricate oriental patterns on it. The ability to cast lightning out of his palms was a given because it is a nice visual effect and would require some aggressive kung-fu type movement to make look convincing. The idea that by casting this lightning he would inflict burn wounds upon his own hands lead to the final design and the bandaged hands you can see in almost all the images.
When designing a new character it helps to ask some basic questions, this will help you to pinpoint certain aspects of the character. At all times this profile can be modified. But these are the initial questions and answers I came up with for this particular character:
Q: Target audience and medium?
A: The character is part of a mature role-playing game, in which you play with a group of heroes, either in co-op mode or CPU controlled.
Q: Who is your character? Is he/she important? Good or bad? Profession? Background?
A: The character is a trained lightning wizard. He has completed a kind of Buddhist training that allows him to be very calm and objective. The character lives alone in a temple that harbours a sacred artefact. His duty is to protect this artefact. When the playable character of the game convinces him he needs this artefact he joins as an NPC to make sure the artefact will not be miss-used.
Q: What kind of world does he/she live in?
A: The world resembles a medieval world with cultures similar to the Asian, African and European ones we know. However there is a strong fantasy element in this world that allows for magic and wizards and exotic wildlife to exist. This world is under threat by demons that want to plunge it into eternal chaos.
Q: Any special abilities?
A: He is a wizard who can control energy; this gives him the ability to shoot lighting out of his hands. When using such powerful spells he is prone to injuring himself, which is why his hands are covered in bandages. Extensive use of his powers has given him brittle, pale skin, and many burn wounds. His veins also clearly show; they represent the paths the energy must go through within his body.
Q: What drives your character? Goals? Objectives?
A: The character is highly sensitive to the energy fluctuations in the world, and always strives to keep a balance. The evil beings attacking the lands disturb this energy balance. To restore this balance he is willing to aid the player to get rid of this disturbance.
Q: What attributes does he/she use? Weaponry? Tools? What does he/she need those attributes for?
A: The lightning he casts out of hands is his main weapon. The bandages he wears protect him from the damaging his hands too much when using these spells and cover his burned hands.
The lantern hanging from his belt has a special candle in it; the flame of this candle starts to flicker when there is danger. The glowing crystal fragments hanging from his chest piece contain the souls of great demons who have tried to take the artefact. These crystals capture the demons preventing them from to possessing new hosts.
Q: Does he/she have relatives/companions?
A: The guru who trained him has trained more talented individuals, some of which he knows and some of which he doesn't. Most of them use their powers to keep the balance and are likely to help each other. Others use their powers for their own good. After he completed his training he was sent to this temple to protect the artefact. He has lived there for years now in solitude.
Q: Is there any kind of side character?
A: During the progression of the game he leaves his temple to help the main characters. During these quests he is likely to run into some other mages trained by his guru and might even find his old master...
The Design Process
The design process is a chronological description of the steps taken to get from the initial idea to the final design. Designing a character is very iterative, that is why I present this overview in distinct sections per iteration.
The first sketches made revolved around the face of the character. The priority was to find the personality of this character as quickly as possible. A 3D head was used as a template; this was printed out and faded. The heads where then drawn on top of these grayed out templates on paper. The coloring was done in Photoshop (Fig.01).
After a head was found that was fitting for the design idea, I attempted to create a matching body. This proved rather difficult since the style employed for the heads was hard to use for materials like cloth and metal. Too much reference was used in the design of the clothing, giving him more of samurai look instead of the magical look I was after. You can see my design on the left of Fig.02, next to two references images from the internet.
After some feedback from peers I decided to restart the process and throw everything out. This time around the starting image would be created with a lot less references; instead the focus would lie on first finding an interesting shape and silhouette. Many quick sketches where produced in this process. I then picked the most successful sketches and developed them into slightly more detailed sketches, slowly reducing 12 potential designs down to 1 (Fig.03).
The final design was very energetic, but it was also very fantastical and lost some of the believability of the first iteration. This happened because references had been disregarded in favour of finding the most interesting shapes possible (Fig.04).
So once again, I started the design process from scratch, in search of a new look for this character. Learning from previous mistakes, I used the best of the two previous approaches for this iteration. During the previous design iterations the amount of references I'd gathered had grown. With more appropriate forms memorised it would be easier to do a new take on this character.
I came to the conclusion that I needed to increase the character's level of torture to make him more interesting. All the previous designs had this notion that he hurts himself while using his lighting casting abilities. If this would be taken a step further, and I showed the mage after years of using this spell, it would definitely give a more interesting design.
Physically he is severely damaged by his powers, with burned hands and lower arms covered by bandages, popping veins carrying the magical energy under his brittle and pale cold skin, dry lips and blackened eyes being some of his notable physical characteristics. Besides his physical torture I decided to sow his leather cap onto his face, integrating the costume and making for some really interesting details (Fig.05).
After the initial sketches it was evident the process was finally on the right track. The shoulder plates allowed for some great detailing, using a lot of Asian elements to make his origin clear. At the same time the overall shape was still something unique and original. A lot of elements were based upon references, but then heavily transformed. The cap, for example, was based on the hat shown in many Confucius images (Fig.06).
Next I began to search for the forms of the details that would define the character. This was about taking all the elements out of ambiguity and giving them a final shape. During this process the 3D base mesh was used. The male base mesh was adjusted to look like an Asian male and was used as guide to keep the proportions consistent between sketches (Fig.07).
The idea to use colorful robes from the original concept was abandoned, instead more natural and desaturated colors where used. Many materials have their natural color and things like the robes are worn down heavily. For the drawings done on paper the tool of choice switched from pencil to a combination of fine-liner and COPIC markers (Fig.08).
In preparation for the final image, the adjusted 3D base mesh was used extensively to find the best pose. Cloth studies were done based on lots of reference photos taken from a life model. This was done to study the movement of the robe and bandages. After understanding the movement and shapes of cloth it is easier to recreate it (Fig.09).
Creating the Final Image
The final image was started by choosing an earlier sketch which had a pose that was meaningful to the character. The chosen sketch was based on one of the 3D poses. Instead of choosing an action pose where he uses his lighting skill, I chose a more serene pose because it expressed his personality more. He doesn't really enjoy fighting or using his powers, he uses them only when necessary.
In his normal state he is a very serene man who is always deep in thought, and this pose expresses that serenity. The sketch was used as a starting point for the final image; since the tonal values in the sketch suggested the coloring of the cloth and not the lighting, the 3D model was used for experimenting with the lighting.
Some elements such as the crownpiece were modeled in 3D, instead of being adapted from the sketch; this was to emphasize their geometrical nature. The 3D render was used as guide for the painting and lighting process (Fig.10).
Since the entire final image was done at 600dpi I encountered some problems during rendering, Maya crashed repeatedly attempting to render in such high resolution. In the end the image was divided into smaller 4k renders and assembled in Photoshop.
The main lighting setup used only grayscale lights to indicate the tonal range of the image, then a second lighting setup was used to suggest the color of the lights. The main tonal value goes from up to down, but the color value goes across the image, yellow from the left side and blue from the right side (Fig.11).
These 3D elements were then pasted on top of the sketch. Besides that, many details on the sketch were replaced by sketches made for that specific detail, as you can see in Fig.12. Some elements such as the 3D hands and face completely replaced the sketch. At this stage the lighting was painted on the sketch using the 3D models as reference.
After the tonal values were in the ballpark, I began coloring the image using the Colorize brush and really saturated colors. In the end the lighting would be quite blue and desaturated and the cloths old, but it helped to start out with vibrant saturated colors and gradually wear them down during the creation of the image.
After the basic colors were laid down, it was time to apply some texture and detail. For this a lot of photos were gathered from the weirdest sources. A good example is the metal on the chest; the texture is a photo taken from a used tea egg. All the fabrics also used heavy photo detail. Sometimes it helps to put things out of scale; for the leather cap, photographs of worn leather close-ups were used. By using the pattern much bigger than it actually is, it will create a more textural worn effect (Fig.13).
These photographs were used in conjunction with all kinds of blending modes inside of Photoshop. Fig.14 shows some "before and after" shots to demonstrate the amount of impact the texture detail had on the final image.
After the forms and color were described properly, it was time for the lighting. The goal was to emulate the lighting used on a head study I'd done earlier, which I felt was very successful. As you can see in Fig.15, I experimented a lot with smoke, light, hue and overall lightness and contrast. In addition to that there were also experiments with other things such as the glow of the crystals, light emitting from the lamp and even some lightning to show off his magical capabilities. In the end I realized that subtlety was the key.
After the lighting was finished, I used the Liquefy tool to fix any weird proportions I noticed during the creation of the image. It was not efficient to do this earlier because it would have made my entire library of layer masks useless. So in order to keep working efficiently this step was postponed until the end.
Now the power of 600dpi really came into play. The Liquefy tool created some distortion, but because the image is never viewed bigger than 50%, this distortion will never be noticed. And if downscaled, the image's distortion will be free. The only downside of working with such a large file is the big memory consumption and the occasional lag.
Something that needed to be addressed earlier was the feet of the character, as they seemed a bit awkward. I used the Liquefy tool to correct a lot of the problems, which meant existing work could be reused instead of thrown away. A render of the leg and foot with a special texture and various paint-overs was used to find the correct shape. After that it was just a matter of liquefying the image and repainting the shading (Fig.16).
Fig.17 shows the resulting final image. This image will be the main image used as reference during the modeling of this character. As I said at the start, the goal was not to create a beautiful piece of art, but to create something that can be referenced throughout the modeling process as a form of reference and inspiration.
Overview of Sketches and Other Images
The sketches and images made during this process are presented in chronological order. Notice that after the final image, many more sketches were made. These were sketches of small details that were quickly drawn-out on paper while working on the 3D model. It is a lot more efficient to try out variations on paper, then to model multiple variations and choose in 3D (Fig.18).