Category Archives: Analysis
In preparation for the Promotion Art which will show the two characters in combat, I have been looking at existing promotional art for games and movies to get a sense of the techniques they employ. Of course, there are whole courses based on this subject, but lacking the time required for such detailed study I hoped to find the basics of the most prominent techniques used.
The layout of the Protagonist and Antagonist is very important to the image, and whichever is most is seen first often sets the mood. Generally, the Protagonist is positioned to the left and/or infront of the Antagonist to make them more obvious to viewers. Making the Antagonist bigger or above the Protagonist can show their strength and power, and the struggle for the Protagonist to beat them.
Going further, the favoured structures of the artwork seem to be those show below. How strict the image follows these guidelines varies, but they often help to guide the eye to important parts of the image.Other important points I have gathered so far in terms of Promo Art featuring combat are:
- The characters should fill as much of the image as possible, but without being too obscured/obscuring
- Poses should be fast and full of action
- Camera angles can be played with and twisted to match the drama of the scene
- The protagonist should be the main focus, but not so much that they completely overshadow the antagonist.
With my silhouettes decided on I should be able to begin work on some thumbnails, so watch this space.
So after a bit of encouragement I went back to earlier attempt at converting an Ashley Wood character into the Dota 2 style. I didn’t attempt the colour scheme (mostly because this piece is annoying me and the more I spend on it the worse I make it) but I did practice the values techniques I mentioned before. As a reminder here is the Ashley Wood character:
And here is my second attempt:
I approached this new piece keeping in mind the philosophy of using value and detail to draw the eye to the top of the character. Focusing on the values aspect first, I desaturated the original attempt to find my highest and lowest values. This is where I found my first problem.
In my first attempt the overall value range was very narrow, and so I found that the character was actually very dull and close to monotone. Thus I expanded my value range to points which I though suitable and created a little gradient chart for reference.
This worked very well for me as, even as I loosely blocked in the character using the increased value range I found the silhouette becoming much more dynamic and appealing. The reference chart was also very useful for keeping in my boundaries and for finding and comparing new shades (using the colour picker).
One part of the design I struggled with was the armour. Partly because I struggle with drawing armour, but mostly because it was all one colour which made it hard to create a contrast between the upper and lower body as Dota suggests. I will bear this in mind when designing characters in the future and be sure to incorporate more complicated and visually interesting clothing/armour.
I am glad I did this task as it has allowed me to greater understand some of the techniques I though I knew Dota used. Also the comparison to my previous attempt from before I knew these techniques meant that I could more obviously see where I was going wrong and how to correct myself. In the future I will need to think more about how values are placed on the character and how they direct the eye. Contrasting these values is also important, and it is vital to have a range of different areas across the character (something this example sadly lacks) to create visual interest and diversity. Knowing all this I feel much more confidant in my future work now.
So after my attempt at understanding Dota 2’s style through practice, I gave myself time to rest and instead sought to understand it through research. I found the Art Guide released by Valve for public viewing and read through it in the hopes that it would enlighten me to the finer aspects of Dota’s designs. Here is what I found to be of particular importance and interest. (Apart from the colour mixing sheet, all images are pulled directly from the document without alteration).
Use of Value
Dota makes clever use of values to ensure the important parts of their characters draw the player’s attention. Given that it is a top down game, this means they emphasise focus to be on the upper areas of characters, but I feel that this is good practice for an character design as the lower part of a character generally has little need for interest.
As you can see they create a gradient on the character, with darker shades focusing towards the feet and lighter shades towards the head and shoulders. This draws our eye up to the head/torso of the character.
Use of Detail
Another important design point for video game characters which Dota highlights is use of detail. Too much detail on a character will just become distracting noise (especially in game) whereas too little detail renders a character bland and boring. Dota balances this by placing large areas of minor details alongside the areas containing a lot. Again values come into this as creating details using values close to that of the surface they are on mean that they can be moderately complex whilst remaining undistracting.
Another point is that the balance of complex to simple is executed in the same way as the balance in values. Areas of complex detail often focus on the upper areas of the character, whilst the lower areas are left looking plainer. This is even implemented in character’s items.
Use of Colour
The final important aspect to Dota Design is colour schemes. Dota uses techniques similar to Ashley Wood, but on a less exaggerated scale, in order to draw player focus towards the top of characters. They use colours with little saturation all over, but especially in the lower parts of the body. Also small areas are given higher levels of saturation to draw the eye while not overwhelming it.
Dota also makes use of a wide variety of colour techniques (as shown below), generally limiting character palettes to a maximum of three colours and then mixing these to create additional colours for the smaller areas. This works to create a unified colour scheme while maintaining areas of contrast.
Below is an example of how the main colours are desaturated and mixed, and how it differs for the spot colours. I have added my own diagrams for further explanation, showing the general curve of these colours in the colour picker.
Dota makes use of a range of techniques (primarily focusing on colour and values) to draw player focus to the upper areas of its characters. Being a top down game this is especially important, however it is good practice for any character design as it is unlikely that designers or players will want to focus on the characters feet.
So my final attempt at style analysis was to take Ashley Wood’s work and recreate it in the style of Dota 2. I decided to steer away from his work on Spawn as I am becoming too familiar with that and instead looked at another piece of his which caught my eye. It turns out this is actually for the Metal Gear franchise, but given that I am not familiar with the franchise I felt it safe to use for my analysis. The character is a bit too sci-fi for traditional Dota too, but I felt this would allow me to greater appreciate the style and techniques.
I’m not entirely sure how I feel about this piece. I did not enjoy creating it which resulted in me rushing myself so it’s not perfect. As you can see, I took some liberties in the final design due to the roughness of the original piece and how it missed out details. I also tried to add a bit more colour to make it more Dota-esque, while keeping with the original palette.
One thing that I learned to appreciate about the Dota art style is that, while all their characters are muscular, only those that are characterised by brute strength make use of wide and intimidating silhouettes. Those that rely more on speed or intellect have slimmer silhouettes and also have longer proportioned legs, which helps to connote their agile nature. I tried to make use of this in my creation above, however I feel the legs may be too buff and look out of place.
In the end, I don’t think I was in the right frame of mind when creating this piece of work to fully appreciate the style and techniques implemented into Dota’s characters. However, it came to my attention that Valve have released a PDF Visual Style Guide detailing much of what goes into their design process. I plan to read this tomorrow when I am refreshed and I hope that, combined with what vague ideas I had during this tasked, I will have a better understanding of Dota’s art style.
So I was not at all pleased with my previous attempt at combining Bob Rafei and Ashley Wood’s art style together, it was just awful and so much went wrong. So I took a break, went and looked at various tutorials on creating traditional-looking photoshop paintings and got some new brushes, and I sat down and tried again.
So I approached this style differently from the last time and from what I had orginally planned. Rather than start in greyscale and add the colours after I picked the general colours used in Jak to block in all the individual areas (the background was just general colours based off of Wood’s works). Then I decided what my spot colour would be (red) and desaturated the other colours a bit to pull the vibrancy of them. At the same time, I upped the vibrancy of the red to maximum.
When it came to shading I took my new (amazing) brushes designed to imitate oil painting and picked shades sticking mostly to the left hand side of the colour picker (although avoiding going fully grey) for the less saturated colours. I also pushed some of the browns from the background into certain areas to add a little bit more warmth to the character (especially the gun) whilst keeping the vibrancy low). For my spot colour I hugged the right hand side to really make the areas stand out and attract the viewers eye.
While the final piece is not an exact mirror of Ashley Wood’s brush techniques, I am still very pleased with it and believe it will play a part in my final style. Furthermore, his use of colour is a definite must for me as it creates a very unique and eye catching style. I will not however, be using his character silhouette/design as inspiration as I need to take more fro Bob Rafei and Dota 2.
When analysing existing Video Game Antagonists I narrowed my search criteria down a bit more than with my Protagonist analysis. Antagonists several shapes and forms. To put it generally, there are the masterminds who use brain over brawn to achieve their goal, and there are the monsters and brutes who fight their way through any obstacle. Of course there are those that fall somewhere in between (most often being smart but also strong) but these are the two most distinct categories. Since I want to create a large and intimidating Antagonist, I focused mostly on those who are not afraid to wade into the melee and get their hands dirty when searching for existing examples. This does not mean however that my Antagonist will be a dumb ape, but I want his strength and size to be his primary characteristics.
So the most obvious technique present here is one I have been talking about a lot, and that is the use of shoulder width to give characters strength. However, whilst the style of Bob Rafei and Dota 2 often rely purely on the shoulders to give this strength and narrow down towards the hips, many Antagonists make use of this thickness across their entire silhouette to show their intimidating strength and power.
An exaggerated example is Zangief (Street Fighter) with his tree-trunk legs and arms. However the Lich King (World of Warcraft) also demonstrates it on a less comical level with his barrel chest and thickset legs.
Another variation of this technique is to put most of the body mass of the character into the upper body, exaggerating muscle mass. Using Zangief as an example again his mass is evenly distributed, but in Bowser (Mario) we see that his legs make up for only a small fraction of his over all mass. This highlights the idea of the raw strength he has in his upper body.
In terms of colour, Antagonists use palettes just as much as Protagonists. Often the colours we see are angry reds of sinister blacks, but sometimes bold golds or royal purples come in to signify wealth as a form of power. Also we can see cold blues and turquoises pulling any warmth out of the designs.
So what does this all achieve? It acts to alienate the player from the Antagonist. Where as the Protagonist is something the player wants to be but is not, the Antagonist should be something the player never wants to aspire to.
- Their strength and power should be pushed passed the stage of desirable to that of monstrous and inhuman.
- Their colours should be cold and/or violent, not friendly or confidant.
- They should alienate the player.
One of the final stages of my research and analysis was to look at existing Protagonists and Antagonists in the video game industry. My goal for this task was see what similarities could be found in character designs across the various styles and genres. This is my analysis of protagonists. (Skip to the end for a shorter, concise version).
Straight away it can be seen that there are several distinct style of Protagonist, each relying on their on ways of making the player like them and want to be represented by them (the most important role a Video Game Protagonist must fulfill).
In the first category we have characters such as Mario (Mario), Rayman (Rayman), and Meat Boy (Super Meat Boy). These characters sit at the cartoony end of the scale and rely on looking friendly and fun to make player like them.
The second category sits are the more serious end of the scale and is much harder to generalise. Here we have Master Chief (Halo), Ezio (Assassins Creed), and Lara Croft (Tomb Raider). These characters, for lack of a better term, rely on looking bad-ass and/or awesome. They give the player a sense of power and might, use their design to give the idea that they can overcome any obstacle, they make the player feel like the unstoppable force.
The third category is for characters that fall between the first two. Ratchet (Ratchet and Clank), Jak (Jak and Daxter), and Samus (Metroid) are all characters who have cartoony looks to them, but still give off the sense of enduring and powerful characters like Master Chief or Lara Croft. This is done using many of the techniques that came before, just combining them with the cartoony styles.
Protagonists have to offer players new experiences, they have to be something the player is not. The player has to want to play as them. They can do this by being looking like a friend to the player, or by looking like something the player wants to be.
- Strong/bold colours can make the character appear fun and friendly when combined with soft shapes.
- Especially useful for young audiences, though perhaps not for me
- However, they can also make the character appear strong and confidant (depending on the colours, eg red)
- Sci-fi and/or fantasy elements are used to distance the character from real life
- this offers the player to explore new imaginations and escape the mundane real world
- it is an inviting experience to most
- Real world characters should let the player relate to them on a personal level
- the player doesn’t pretend they are the character, they want to imaging they can become the character
My second style mash up was combining Bob Rafei’s characters with Ashley Wood’s style. In this experimentation I was more interested in replicating Wood’s use of colour, although I did attempt to replicate his brush work also.
I struggled more with this than I did with my previous style mash-up. I’m not sure if I was approaching it with the right techniques but I found it hard to replicate Wood’d oil-on-canvas brushwork digitally. I’m also not sure if I got the colour scheme right either. The blue certainly pops like I wanted it too but the painting is also very muddy/messy.
Thus I plan to have a second attempt at this image after researching some better techniques on how to achieve Wood’s brush technique digitally. I feel the colour might work better if I work in greyscale initially and then colour over it.
I’ve been experimenting with different combinations of the main sources of inspiration, trying to find a middle ground somewhere that I am happy with. For now I am focusing on them two at a time rather than all three at once just to keep it simple in my head.
My first attempt was to recreate a DOTA 2 character in a style more similar to Bob Rafei’s. I chose chose a character at random and ended with Lycan:
I broke him down into a base skeleton and twisted his proportions to something more akin to Rafei’s cartoony style. When it came to shading and rendering I chose a style somewhere in the middle ground of DOTA’s and Rafei’s. I felt Rafei’s was too cartoony for my preferences, but I didn’t want to directly imitate DOTA’s techniques.
The end result is not quite as close to Bob Rafei’s style as I would have liked, but this experiment was certainly helpful in seeing how his style works. As you can see below the proportions I have used (red/left) do not quite fit what I, on reflection, believe Rafei would use (cyan/right). The forearms and boots are not quite chunky enough and there isn’t enough difference between the shoulders and waist. However I now know how to improve if need be.
I‘m beginning to appreciate areas of Rafei’s style that I never much noticed before. Whilst I don’t like the end result of his work I’m especially liking the exaggeration he puts into certain proportions (those highlighted by the triangles above). I think this has potential if combined with the other sources in some way. However I am still not a fan to his use of line work, I might try using coloured lines rather than black but I don’t think it will improve my view on it much.