Here’s a little tutorial on how to make reference photos work in your art, by making them subject to the idea and an overall composition, not the other way round.
All my paintings begin with an idea of a scene I want to paint. Reference always comes in a second or even third step. I’d decided to paint the scene in the Silmarillion where Fingon’s mounted archers take on a young Glaurung. I had wanted to do this scene for a while.
Working from a photo from the get go never works for me, and certainly not in a complex, multi-character scene. Photos tend to be much less dynamic than the compositions I want to achieve (since I’m an artist, not a photographer), so working from photo without a lot of change often results in boring paintings.
Once I have a picture idea, my mind starts thumbnailing. I try out different compositions in my head, and find where the snags are.
Whenever I do a battle scene, I want to portray both combatants in some way, so it’s very obvious from the start that my camera angle needs some thought. If I want to show Fingon’s face, I can’t have Glaurung (left). If I want to show Glaurung, I’ll have a few Elves from behind, and Fingon somewhere off in the distance (centre). So I decided to show Fingon charging past and shooting behind him (right) – quite obviously, that meant some extremely good reference.
Refining an idea and finding (or shooting) specific reference
At this stage, the scene is still only in my head. I know Glaurung will be in the middle ground, so if Fingon is galloping towards us, he’s closer to us, and needs to shoot behind him. Riders further back will need to shoot straight to the side. This is the reference I need.
Very often, for complicated poses, I shoot reference myself. I know exactly what I want, and I avoid copyright issues. For riders, I often take photos of my daughter at her riding lessons. In this case, I came up blank. None of the hundreds of reenactment photos I’ve shot at events had any mounted archers, and I needed photos of people who knew what they were doing – because I don’t know a lot about archery. So that ruled out family members posing with a bow while sitting astride a sofa.
A Patreon supporter of mine then pointed me to several great mounted archers with Instagram accounts – and there were such an incredible lot of great photos! I immediately reached out to Erin Jardine and Freja Trulsdotter, who gave me permission to use their photos.
I started filing away those photos that were the angle and poses I needed. At this point, the photos and my idea start bouncing off each other in my head, as my idea is defined.
Making your reference work in perspective
A lot can go wrong when you combine several reference photos in one image. Ask yourself: Were the photos all shot from roughly the same height? Look at the horizon line for clues – it’s where the camera was, and your viewer’s eye will be. If the photographer was standing, chances are that the photos will work together. If you have one extreme bird’s or worm’s eye view in there, it won’t work with the others.
Here’s a trick to keep several people standing at different distances to the viewer in perspective: Assume the viewer is standing in the same room. If he is the same height as the characters, align all the eyes at the same level. Unless they’re a king on a dais. Or Maedhros. You can indicate different heights that way, too.
(I haven’t seen this lineart in years, and I apparently changed Celegorm’s expression before finalising this – he cracks me up!)
If you want to put characters into a bird’s eye view, you need to find another converging point by adding invisible heads to them. Or by handing each one a seven-foot lance and having their tips converge. The horizon line has to move up there, too.
If you align the eyes of people on horseback, it will look as if the viewer is also sitting on a horse. (Plus, eye-alignment can sometimes look a bit like differently sized people all dangling from a washing line.) Here, I wanted to put the viewer in the middle of the action, but not on a horse, to add a sense of “Gosh, I’ll be trampled!” to the scene. So what you do is align not the characters’ eyes, but a point that is level with the (standing) viewer’s eye. When you stand next to a charger, your eyes are barely above its rump. So this is where I aligned my riders. I chose their saddlebows, allowing for some unevenness for the movement and uneven terrain.
Plus, the fact that the characters’ heads are all at different levels adds a lot of movement again, forcing the viewer’s eye into an up and down movement, like a gallop.
Instead of choosing three poses that were roughly the same, I decided to use three slightly different ones for the main riders, two slightly from the right, one slightly from the left side. Fingon would be passing us on the right; the one to his left would thunder past us on our left side, while the one to his right is already swerving to cut right across us and vanish off to the right. That way, pressing a mental “play” button on the scene, we see ourselves standing right there as the cavalry passes us left and right. This is a very effective way to thrust the viewer into the middle of the action. It works with any movement.
Compositional rules – distance, crumping, overlap, line of action
I did my best here to adhere to the rules of composition. Putting figures at several distances is a great way to add a sense of space. Having elements or character overlapping each other adds to the sense of space. “Crumping” means clumping several elements together in uneven numbers – here. the three main riders. The line of action is very much defined by the horses’ movement, and I added several elements to the piece that lead the viewers eye – the dragon’s tail, the eyes of the riders and the dragon, all converging in the middle of the piece; the tree trunks that lead the viewer’s eye back into the image in places where other elements threaten to lead it out.