Paint a scene of awe and beauty

Craig Mullins paints a huge in­te­rior basil­ica.

ImagineFX - - Issue 130 | January 2016 - Craig Mullins

Wel­come to my take on a Re­nais­sance basil­ica, a bit like St Peters’ with some steam­punk in­flu­ence thrown in. To do some­thing like this you should be up on your ba­sic lin­ear per­spec­tive. I’m us­ing sim­ple 3D to get the some of the el­lipses cor­rect, be­cause the lens is very short (wide an­gle) and things can dis­tort in strange ways. In ad­di­tion, the con­cen­tric el­lipses in the main dome are sub­tle and have to be done pre­cisely (but not tight!) to look cor­rect. I didn’t use 3D to es­tab­lish val­ues, but that’s up to you.

I’ve painted the girl in the dress and the boy sev­eral times now. I think they im­ply a larger world that I’ve been adding to over the years. It’s in­ter­est­ing to see how much I’ve learned over time. Some­times it’s en­cour­ag­ing to redo a sub­ject or way of work­ing, to see if new knowl­edge has any light to shed on things. Paint a self­por­trait ev­ery year and keep them. In 20 years you might see some im­prove­ment. Ha, just kid­ding, let’s hope you do!

One thing you might see is sev­eral years of non-im­prove­ment fol­lowed by a break­through year. Im­prove­ment in art is like that some­times: just grind­ing it out hop­ing for a light­ning bolt.

1 Sim­ple sketch

The first comp sent for ap­proval is a mono­chrome, low-con­trast sketch. I know that the con­trast will be­come higher and the colours more var­ied, but some­times a sim­ple sketch is all you need to get the go-ahead. I want to have the feel­ing of a space that’s filled with light, and the eas­i­est way to show this is to ex­ag­ger­ate the light and colour bounce of all that sun­light.

2 Ap­ply­ing 3D to the sketch

I want to stay true to the un­usual per­spec­tive. I’m us­ing a very wide-an­gle lens, so the dis­tor­tion is dif­fi­cult to do by eye. But the 3D is sim­ple: I take a sphere and four cylin­ders, se­lect all the ver­tices be­low the mid­point and drag them down. I model the steps out of du­pli­cated boxes. There’s a cheat in the per­spec­tive. Can you find it? I’ve done it for com­po­si­tion rea­sons.

3 Con­trol the con­trast

Now I glaze over my cool colours with trans­par­ent warm ones. The en­tire ar­chi­tec­tural por­tion has a nar­row value range, or low con­trast, con­sist­ing of mid­dle to high val­ues. A lot of get­ting a paint­ing to read prop­erly in­volves controlling your con­trast. Some of the tex­tures are quite ‘ac­tive’ at this stage and I’d have a sim­i­lar prob­lem if I were to make a dark mark.

One of the keys to at­mo­spheric per­spec­tive is pre­cise con­trol over con­trast, the trick be­ing to achieve va­ri­ety in colour and value, but in a very nar­row range. As I glaze (and scum­ble) over the im­age, I “quiet down” some ar­eas, making the colour and value range more nar­row. Make a se­lec­tion on your im­age and hit Cmd+L. The his­togram tells you how much con­trast an area has.

4 Treat val­ues with care

I con­tinue to work mostly down in value, and take note of ar­eas that are of a darker lo­cal value (mean­ing that they’re just dark ob­jects to be­gin with). How dark th­ese ob­jects are in your paint­ing de­pends on how far away they are, the den­sity of the at­mos­phere, how many par­tic­u­lates there are float­ing about, and so on. It’s a mat­ter of sneak­ing up on th­ese val­ues mov­ing around the im­age, be­cause all your val­ues are so in­ter­de­pen­dent.

If you look at the fi­nal im­age and how dark that fountain is and tried to paint it that way at first, then you would have the ‘ hole in space’ prob­lem. So I would rec­om­mend grad­u­ally glaz­ing down all the ar­eas you think might be dark, so that they sup­port each other and make sense over­all.

5 Do it right first time

Here the darks con­tinue and some of the shapes are be­com­ing more de­fined. It’s a di­a­logue be­tween the artist and the paint­ing. The fountain has more con­trast than it needs, but I need those ex­tra val­ues to get those fig­ures to read. I can al­ways glaze back over the top of them to lower the con­trast.

This is an in­ter­est­ing point, though. Best prac­tice would be to paint things to their proper value at once, with no wash or glaze later. A lot of dig­i­tal art abuses this, though. An ex­am­ple would be if I wanted a shaft of sun­light com­ing through one of the win­dows: should I paint all the in­for­ma­tion that’s be­hind the sun­beam at a higher key, or paint it with no sun­beam and blat some light yel­low at the end? Given that we’ve all done the lat­ter, try to paint that sun­beam into the paint­ing it­self. It takes a lot of work, but the re­sults could be worth it.

6 Prac­ti­cal con­sid­er­a­tions

Hav­ing said this, the time and lack of flex­i­bil­ity of­ten makes cool stuff like this im­prac­ti­cal when cre­at­ing com­mer­cial art. What if the client doesn’t like the sun­beam? The time that it takes to paint it into the im­age is con­sid­er­able.

An­other con­cern is that, as I ex­plained in the pre­vi­ous step, I don’t know what my fi­nal val­ues are go­ing to be. It’s dif­fi­cult to re­paint that sun­beam at a dif­fer­ent key sev­eral times. I usu­ally take the mid­dle road: paint it as close to the fi­nal value as you can, and then you can be del­i­cate with an at­mo­spheric layer at the end.

7 A fresh ap­proach to the struc­ture

Here I’m con­cerned with two is­sues. First, the im­age has too lit­tle vari­a­tion in ma­te­rial, like it’s a toy made out of plas­tic. Big com­plex struc­tures like this have a mil­lion dif­fer­ent ma­te­ri­als, and suggest­ing them all means con­trast, and that de­stroys the feel­ing of scale. The sec­ond prob­lem is the de­sign doesn’t feel steam­punk enough. It looks too much like St Peter’s, which is the main in­spi­ra­tion for the space.

So to solve both prob­lems, I stop wor­ry­ing about glaz­ing val­ues from light to dark slowly, and start throw­ing in darks with aban­don. The ra­tio­nale is that there are many ex­posed struc­tural el­e­ments made out of iron. I put all th­ese darks on a sep­a­rate layer and paint them in quickly (but not so quickly that the crude­ness might af­fect my eval­u­a­tion of them) to get an idea. In the end, I do keep some of them, as you will see as we move along.

8 Study HDR im­agery

So now I’ve backed off a lot of the struc­tural pig iron, be­cause it seems to have the space-punch prob­lem. What I do in­stead is paint the fore­ground a lit­tle darker. The idea is that the tun­nel the viewer’s stand­ing in has very lit­tle light com­ing into it.

This is a con­ven­tion that I wanted to avoid, but time made it the best course. Imag­ine that you really were stand­ing here, in that space: you wouldn’t see this fore­ground as dark. If you took a HDR (high dy­namic range) pho­to­graph you could ex­pose the fore­ground just fine. Would the im­age be­come flat? Per­haps, but I think it would have an in­ter­est­ing look. In­stead of show­ing vol­ume through light and dark masses, it’s shown through pre­cise high-fre­quency de­tail. The dark fore­ground is a well-known paint­ing con­ven­tion. If you have the time, look at HDR im­agery and an­a­lyse how it works: it’s a dif­fer­ent way to think about things. You now have one more pos­si­bil­ity in your tool­box on how to show form.

9 Don’t just push lights and darks

The ba­sic space is read­ing okay, so it’s time to start adding de­tail and fig­ur­ing out what some of the ar­chi­tec­ture will be. I also set the bot­tom of my value range. Once you have that ‘ floor’ you can reckon up in there. The op­po­site end of the value range is easy; the win­dows in the dome are al­ways the top end. I fig­ure that the paint­ings up on the walls would prob­a­bly be the dark­est dark. The lo­cal value is black and is re­ceiv­ing not much light, which means it will be dark. You can al­most make an equa­tion out of it. Lo­cal value x in­ci­dent light x dis­tance = value. As the dis­tance goes up, value goes up too, de­pend­ing on how much haze there is.

In ad­di­tion, no­tice that I’ve de­crease the value range of the fig­ures in the fountain. Some of the light val­ues are very over­stated. This is an easy trap to fall into if some­thing isn’t read­ing: just push the lights and darks. Slow down, make bet­ter shapes and bet­ter draw­ings, and you’ll im­prove as an artist.

10 Pol­ish­ing the scene

At this stage in the paint­ing process I’m just adding de­tail and defin­ing ma­te­ri­als through­out the scene. I’m mostly do­ing this by drop­ping down in value.

11 Don’t trust your eyes

Some­times it’s im­por­tant to really mea­sure the scale of things. The girl in the black dress was gi­gan­tic: I didn’t see this right away, and wasn’t even sure that it was true. But mea­sur­ing things in per­spec­tive is al­ways a good thing to do. This is es­pe­cially true if there’s any type of wide-an­gle dis­tor­tion or other un­usual per­spec­tive, be­cause your eye isn’t as re­li­able in such cases where the eye doesn’t have as much ex­pe­ri­ence.

Newspapers in English

Newspapers from Australia

© PressReader. All rights reserved.