Composition and thumbnail readability
Getting your image to "read" immediately with no effort at a small size is a good way of analysing the success of your efforts. When painting you can use a navigator window or a smaller duplicated instance of your canvas on screen that allows you to look at the image on the whole without getting bogged down in the detail.
Overall I think Narholt's concept was fairly well defined as it is pretty clear what is going on. However in general I felt it fell fairly flat as an image for a few reasons.
By analysing the thumbnail and then the full image the main issues I identified were:
- The composition whilst heading in the right direction, needed to be more dynamic and have a better flow to it.
- The main focal points needed adjustment in terms of placement and relative value.
- The scale of the environment needed to be clearer.
- There were lighting inconsistencies, contrast issues in the main rock formation and the sky was way too dark. The value range overall was making the image look quite flat.
- The figure's proportions and pose needed work.
- The colour palette was lacking vibrancy.
- The weather effects needed to be more dynamic.
Composition and Dynamic focal points
The rule of thirds is incredibly useful because it just works. Narholt clearly applied the idea of the rule of thirds to this piece with the cave entrance and warrior being placed on thirds points but I felt that tweaks could be made to make it more balanced and flow better. The temptation with rule of thirds is to just pick all your focal points from a canvas that has been divided into thirds, once. This can work but it can also end up leading to a pretty static composition. The image below shows how I divided up the canvas to determine the new focal point arrangement.
It is a method I use often for full scene illustrations like this one. Once the first focal point is determined (in Green) you don't stop there but keep dividing the canvas up into smaller successive blocks using the rule of thirds to pick secondary and tertiary focal points. This may be clearer by looking at the different coloured grids which show how I picked each point (numbered 1, 2, 3). The reason for doing this is that by picking more than one focal point you create movement around the image and get more variety in the positioning of these points. By using a rule of thirds grid each time, you can be sure that the subdivisions will work as mini compositions in themselves.
The white dashed line shows the general movement that will be created between the three focal points. Note that they are all at differing depths in the image as well. After this was set up the whole composition was painted to aid these points and the flow between them.
I didn't get a clear sense of scale from the original image especially around the cave. Was it something she would have to crawl into on hands and knees or a giant cavern? I decided to go for a cave entrance that was a bit larger than a person and push it into the distance a bit more. I made the rocks more imposing as a feature by extending them off into the distance and off the canvas so they looked like part of a larger ridge. Don't be scared to go off canvas with your subject matter. It can help to add a sense of scale and imply there is more to the environment than just what is shown.
The level of detail around the cave was low in the original so it was hard to get scale cues. With some focused texture detailing at the cavern entrance the scale becomes a bit clearer. Note the little tick marks indicating blobs of snow or ice. In focal areas it helps to use a smaller brush size and add indications of detail like these, because small marks really help draw focus and can imply scale. This could have been pushed even further given time
Values are the most important thing to get right as they alone determine all the forms, the depth and overall read of any image. The first thing I noticed in Narholt's image was the value structure of the piece was making it seem flat. It seemed overly dark especially in the sky which was a very oppressive block in the composition. It was also blending in too much with the distant ground to create an undefined depth for the background.
The large rock formations in Narholt's image were very high in contrast. A bit too high given the lighting conditions. Higher contrast tends to draw the most focus which is what I think he intended, but it also tends to bring things forward in space more.
The image below shows the actual numbers of the darkest and lightest values in various spots on the canvas. The half circles show the actual colour picked. The black numbers are the darkest values in the area, the white numbers the lightest values. You can see that the warrior shares similar levels of dark value with the entire mid-ground mountains which in turn are only slightly darker than the background. if you look at the thumbnail again they almost read as if they are all at the same depth which was the main reason for the image appearing flat.
In "real life" things tend to get much lighter and much lower in contrast the further away they are due to atmospheric perspective. Furthermore the darkest object with the most separation between high and low values (ie contrast) will gain prominence and come closer to the viewer. The mountain peaks which have the highest contrast clearly draw the most focus, but they weren't in my mind the best primary focal point.
In my paintover I tried to create a gradual but clear separation in contrast and value range between the foreground the mid-ground and the background. For the mid-ground rocks I reduced the separation between the darks and lights to make sure they fit in the middle depth of the painting. I lightened the values of the sky substantially and reduced the contrast in relation to both the mid-ground and foreground elements.
I'm not going to explain in detail every difference but you can study the two images and values closer to figure out more about what's going on between the two. Basically in the paintover the further away the spot, the greater both the light and dark values get, and the less the difference between them. The focal points tend to get amongst the darkest and highest values overall, but also the greatest separation.
The key thing to remember is that it is always a balance between the dark and light values and the separation between those values which determines where an object will sit in the painting and how much focus it gets.
Doing studies really helps your understanding of values but sometimes just using the colour picker tool to sample various areas in reference photos or master paintings like I've done here is a really useful thing to do on its own. It helps you understand how value (and saturation) works across an image without painting a thing.
Counterchange and overlap
Besides control of values you can use the ideas of counterchange
to create the impression of receding depth. Narholt has used this to a degree but I felt it could be pushed. Counterchange is when something light is put in front of something dark or vice versa. By repeatedly overlapping layers of light areas with dark areas you can quickly create the impression of depth. You do have to keep in mind the contrast and values of objects at the depth you want them to be in as they recede but it is a very simple and effective technique.
The two panels below are vertical slices of the original image next to the equivalent slice on the paintover to show how counterchange and overlap can go a long way to creating depth.
Design, shape language and repeating elements.
Design is often not given enough love in the race to gain rendering skill level ups. Technical ability is all well and good, but design is just as, if not more important for concept work. Shape language is a design term for the basic shapes (circular, triangular, wavey, spikey etc) that predominantly make up anything. It's a simplification tool but it is important in conveying feeling too. Things that are spikey probably aren't going to be viewed as friendly, things that are rigid and cube-like like will imply structure and logic. There's more to a shape than meets the eye.
So anyway, Narholt's shape language used large triangular spurs of rock that were fairly aggressive and unfriendly but perhaps a little bit too much so. This is a style choice of course so there was nothing wrong per se, but it was only around the cave area. In the paintover I refined his shape language a bit and also scattered it throughout the image in various scales, from the distant rocks and mountains to the smaller rocks in the foreground. This helps add interest and lend a sense of consistency to the design of a piece.
Shape language applies to anything, objects, creatures and characters as well. You can use contrasting shape language to draw focus and suggest some fundamental difference as well. So think about the shape language before starting your next piece, repeat it around, use it at various scales, think about how to use opposing shapes, what emotions they convey. Don't go overboard but definitely be mindful of it.
Figure pose and proportions
Figures are always a good go-to to imply scale and add interest to environments. The downside to using figures is that everybody inherently know figures inside and out so if you don't nail them viewers of your work will know something is wrong. I thought Narholt's figure didn't have very accurate proportions, was awkwardly and stiffly posed and didn't have a sense of motion to it. Cutting her off at the knees didn't work for me in this situation either.
I adjusted the proportions so her torso and limbs weren't so stretched and gave her a neck so it didn't look like she was shrugging. I made her pose more balanced and tried to imply she was in a still, but ready-for-movement stance. I also tried to use values to show depth changes and form within the figure itself. The same that applies for values overall, applies to within any object as well, so her far arm is lighter to show more atmosphere between it and us. Nothing but lots of figure drawing and anatomy study will help you improve your figure skills.
The cape wasn't very realistic in terms of its folds and movement so I tried to fix this. Reference is very useful for things like drapery because it is hard, without a lot of practice, to get this right straight out of one's head. I didn't go wild with detailing as I felt this was more about the environment than the character, but as she was one of the main focal points it certainly wouldn't go astray if that is what is wanted.
The original image I felt lacked in colour vibrancy. I can't go into detail about colour theory here, but using complementary colours in your work can help achieve this. A quick way of tweaking images and getting a better palette to work with if things are looking a bit drab is to use the colour balance tool.
The image shows an idea of what you can do by just slider pulling: (Note this was done in Gimp but the tool is pretty similar in Photoshop)
You can easily make your shadows warmer, or bring more cools into your highlights, more warmth into the mid-tones or whatever you want. Understanding colour theory will help you get what you want to achieve, so if you've been putting it off, start studying. I recommend James Gurney's Color and Light
book. It's cheap, easy to read and a one-stop shop for understanding, well, colour and light.
Movement and dynamic effects.
I felt the use of atmospheric effects in Narholt's image were a good idea to bring life to the environment but I wasn't really feeling the lightning. It wasn't adding a lot to the composition and actually added clutter to the image. I didn't think that you would even be able to see distant lightning in a blizzard so I left it out. For the snow I felt more movement would be better.
I did the snow in two layers. The same technique can be used for rain.
- For the small flakes:
- Fill a layer with black.
- Run a noise filter on it.
- Run a light gaussian blur over it all.
- Play with the curves tool to compress the value range so you get more distinct white points on a field of black.
- I put the layer on "Screen" but you can experiment with other modes.
- Motion blur the entire shebang or select portions to blur in various directions to get that sense of movement. You can also use the smudge tool at a large brush size to paint quick and dirty blurs over large portions of the snowflakes at once.
- Then simply use a mask or erase out where you want remove the snow where it interferes with the image too much.
The larger flakes were just a soft round brush with scatter applied to it which I dabbed a few strokes on a new layer with. Motion blurred, deleted, and value adjusted individual flakes until I got an arrangement that helped the composition.
I hope that this critique will help Narholt and others to push their next piece even further. I will be inviting the artists whose work I crit to submit any post-crit changes to me so I can add their "finals" to each post. It's ok if you don't want to work on a piece after the fact, but if you do it will help bring each critique full circle to post it here.
Annnd done! Phew...
I realise I'm definitely going to have to do shorter crits from here on as I doubt I can keep up this amount of in depth exploration each week! To help me make these posts leaner, meaner and more useful, please holler in the comments below on what you thought. What bits did you learn from the most (if any?!?) What would you prefer to see less of or more of? Were the technique tips more useful or the fundamental principles explanation?