The UK freelance artist discusses the artistic principles that unify his architectural photography with his CG visualizations, and the lessons that 3D professionals could learn from photographers.
James Lawley's work straddles the boundary between architectural photography and architectural visualization. A highly sought-after freelancer with a client list including both architects and developers, the artist creates images that balance minimalist rigour with human warmth, whether created as 3D renders, photographs of real spaces, or a mixture of the two.
Below, James discusses the relationship between photography and digital art, the lessons that 3D professionals could learn from photographers, and the crucial role that AMD hardware plays in his workflow – no matter which side of his business he is working on.
AMD: How did you get started in the architecture industry?
James Lawley: I studied product design at university, and through that I ended up designing things that interacted with spaces in interesting ways: benches that were designed around the environments they were placed in; that kind of thing. Naturally, that got me interested in visualising spaces.
I was fortunate that where I was at university, there was a good architectural visualization studio who I did a bit of freelance work for. I learned the ropes there for a couple of years, applying the things I knew from modelling furniture digitally.
After a few years, I wanted to move away from where I was living, so rather than go to work for someone else, I decided to set up on my own. It made sense to blend the two sides of my work: the digital side and the photography.
AMD: Had you previously worked professionally as a photographer?
JL: No. I'd done some photography at university – doing product design, I was photographing things [I was creating] – but I didn't train as a photographer. But I'd done a lot of travel photography, some of which has been published on stock photography sites like iStock, Getty Images and Shutterstock.
AMD: Did you know of anyone else blending visualization and architectural photography?
JL: A couple of people. Peter Guthrie, a big hero of mine when I was getting started, did a mixture of architectural photography and visualization before he started The Boundary, so I knew it was something that could be done.
My decision to blend the two was 100% artistic. Before setting up the studio, my passion for photography had always informed my CG work, but it seemed logical that my retouching skills and knowledge of architecture could serve photography projects just as well.
AMD: How does your work break down between the two fields?
JL: It's probably about 70% architectural visualization. A lot of my photography is for architects; the visualization tends to be for developers wanting to promote build before completion.
Most architects are now doing some degree of visualization in-house, and the proportion has been rising over the past three or four years. It tends to be developers who want something a bit more polished for sales and marketing.
AMD: So how do your architectural clients use your photography?
JL: I work with several architects in London who do one-off houses, and that's been the majority of my photography work in the past couple of years. They're generally high-end residences, and the architects often do the entire renovation project. Some of them create a project book, both for their own portfolios and to provide to the client at the end, and that's where they've got me involved.
A typical architectural photography project would involve 100-150 photos, usually shot over a one- or two-day period.
AMD: How does your experience in one discipline feed into to another?
JL: If you're good at one, you're not necessarily good at the other. Some people I've worked with in the past are very technical within 3D software, but don't necessarily have a photographic eye. Equally, I've got friends who have a great eye for composition and lighting when taking photos, but aren't technically skilled.
I try to think about my visualization as a photographer. When you're doing a photo shoot, particularly of interiors, sometimes the best thing you can do is take as many pictures as possible. On site, your eye can be drawn to one thing, and then you're not looking at the other parts of the frame. When I get home, sometimes I'll notice something horribly out of place, or that doesn't fit with the rest of the image, whereas if I'd taken the photo two steps to the right, I'd have avoided it.
I take a similar approach to my 3D work. I don't just model up a space, put in a camera and decide that's the angle I need. For each different composition, I create at least four or five different options.
Pre-planning can sometimes be a curse. There are so many standard compositions that you think are going to work well – flat-on, or a three-quarter view, for example – and sometimes you end up sticking to them too rigidly throughout the project. If I go a bit over the top and render more views than I need, sometimes I find something more creative between the two.
AMD: Are there any other photographic workflows you replicate in CG?
JL: Most of the time, my cameras are set to shoot black-and-white JPEGs, which also gives me the colour RAW data. By shooting in black and white, my composition is entirely driven by light and shape; then, when editing the shots, I can choose between colour or black-and-white images. I’ve found that the shoots I’ve worked in black and white almost always come out better compositionally – even when I use the colour images at the end.
This approach has begun to feed into my CG work: I now use a black-and-white film LUT in the renderer's frame buffer, driving me towards a more complete composition. I’ll obviously render the image out in full colour, but it challenges my thinking when setting up cameras.
AMD: What principles of composition do you follow? Most people only mention the rule of thirds.
JL: The rule of thirds is the most rigid. You've also got the golden ratio. And where the horizon and vanishing points fall. The horizon is often put smack in the middle of the image, but it can help to apply the rule of thirds there too. I use Adobe® Lightroom® to edit photos, and it has a whole set of readymade grids and guidelines. I often cycle through them to see how the images conform to them.
I also try to be intentional with framing. For example, in an interior, where the angles of the walls and ceiling meet the edges of the image can be very important for the composition. In photography, if there's a wall there, you can't move it, but in CG, you can change things around more freely, allowing for greater creative freedom, although that can make imagery misleading.
AMD: But if you're visualising something that's going to be built, you can't move walls at will.
JL: No, but most 3D software will allow you to use a ‘clipping mask’ on the camera view. That means it can be slightly behind a wall, or where you would have part of the door frame peeking into the shot [and the foreground objects won't show up in the render].
In the real world, a great technique for photographing smaller spaces is to raise or lower the camera height, allowing the space to look larger. That's something you can do in 3D as well.
AMD: Beams of light often play an important role in your compositions. Which part of your background does that come from?
JL: I think it probably comes from photography. I'm quite low-tech with shoots. I prefer not to use any artificial lighting, and I rarely take any additional camera flashes. I try to plan out a photo shoot based on the natural light. I'll get the architect to send me the plans, so I can work out where the light will be coming from at different times of day, and where and when to get the best-lit shots.
The architects I work with very much value light as part of the work, and some of the commissions have been to illustrate how outdoor and indoor spaces interconnect: a living space to a dining space to a kitchen to the outdoors, for example. In reality, there is no single image that will do that, but as a photographer I’ve learnt you can use light and reflections to indicate those things without actually having to show them. This has rolled into my computer visuals too.
AMD: A lot of your interior images use soft diffuse light. Is that a conscious choice?
JL: Part of it is just wanting to keep things natural, I think. If you look at 3D tutorials online, quite often they only model three sides of a room, because that makes it really well lit and saves modelling time, but I don't think that gives you an accurate portrayal of what the space will be like.
I've also started taking the same approach to exteriors, modelling the entire site outside the building. Although you only see what's outside the windows, if there are things outside that are going to be casting shadows, or to be reflected [on surfaces in the interior] having a full 3D model helps with the project authenticity.
For example, on the Little Hadham project [a set of visualizations of a development of seven new-build houses], I created the entire site map in 3D. Because I'd modelled all the vegetation out into the distance, I was able to provide more options for each of the views. I'm currently working on interiors for it, and they're quite easy to do because you can show real views across the site.
Many studios say that this is overkill and cuts into the budget, but I feel that it is the value I add, and keeps architects coming back to me.
AMD: The interiors you photograph often have minimal colour palettes. Does that make it hard to compose a striking image?
JL: I almost think it makes it easier. It forces you to make the composition compelling rather than being interested in the contents of the room.
I try to carry the same approach over into CG to ensure that all elements in a render match correctly. In Adobe® Photoshop®, I sometimes put a Hue/Saturation adjustment layer right on the top of the image and push it right to the max so everything is really saturated. When you do that, you often find that the things your mind is telling you are wrong: that the greens actually have a lot of red in them, or that the white cladding on a house is really quite yellow. When you do that, then work back through [each material] you can be more intentional about the colour temperatures you want. It allows me to focus on what matters, and not be distracted by materials.
AMD: What are the key qualities of a successful architectural image?
JL: A solid understanding of what the client wants to communicate. As a creative person, it can be easy to want to run with your own ideas. With residential jobs, you can think it would be nice to show the space with a particular type of furniture, say, when the client may already have a specific goal in mind.
Then there's lighting. You can have the most beautifully designed building, but if it isn't lit well, it isn't going to look good – and for me, that means lighting it as it's actually going to look. There are companies producing amazing conceptual work, but that's definitely not my field. I go for very natural, very inviting, homely spaces: places I can picture myself living in. Lighting images so that you feel drawn into the space, that you feel a connection with it, is really important.
AMD: What commercial trends do you see in the industry?
JL: I think CG artists will need to work a lot more closely with the marketing department from the beginning of a project, and we will see more focus put on creating a set of vignettes rather than a single striking image.
One example is Chequers House. [A building conversion project for which James produced interior visualizations.] The housing developer had been involving interior designers as well as architects, and they had a large marketing department. There were a lot of different people involved, with a lot of different opinions about how the apartments should look. With CG, they didn't need to decide immediately: I could show all of the options.
We were able to use all [of the variant images] for different parts of the marketing, to appeal to different demographics. Even for smaller developments, being able to give a client images for different target markets can be very beneficial – and not something you can do with a real space without huge set-up costs.
I believe we will continue to see more teams getting involved in final visuals, whereas before I had more freedom. Return on investment is getting more focus at every stage of the process.
AMD: What about artistic trends?
JL: Recently, I've been toying with cinemagraphs [still images with subtle looping animations, popularised by Kevin Burg and Jamie Beck]. If you look at my AD + N Loft images – it's a full-CGI personal project, recreating a real apartment space in Brussels – you will see that there's a bit of movement to the light, to the dust.
That kind of subtle movement – which you can easily achieve on top of a still render using Apple® Final Cut Pro® or Adobe® After Effects® – is really interesting. Instagram and Twitter are so relevant when marketing to consumers, and with Instagram, people don't want videos a minute long. In a five-second clip, having the fire flickering or steam rising from coffee cups heightens engagement.
AMD: What camera gear do you use?
JL: I've been using Canon gear, but I'm in the process of switching over to Fuji. A couple of years ago I got a point-and-shoot X100S, which I fell in love with, so I've just bought the XT3. I've been so impressed with the quality of the sensor in the Fuji. The amount of post-processing you can do with the images – everything goes through Adobe Lightroom – is amazing. You don't have to capture everything in the moment.
I tend use a wide-angle lens (around 20-28mm, usually with an aperture like f/16) for ‘whole room’ images, and then switch to 50-70mm for details or vignettes, with a much larger aperture like f/2.0.
AMD: What CG software do you use?
JL: Autodesk® 3ds Max® and Corona Renderer, plus a number of 3ds Max plugins: [object scattering tool] Forest Pack Pro is incredible. For retouching renders, I use the Nik Collection [of plugins for Adobe Photoshop]. After compositing, the last step is always to run the image through its analogue effects function. I find it adds a lot more realism than, say, manually creating a vignette.
AMD: Have you used any other renderers?
JL: Before Corona, I was almost entirely using V-Ray. I'd also dabbled a bit with Maxwell. But I switched to Corona a year ago and haven't looked back.
What appealed to me was how easy it made to do very photographic things, like creating volumetric lights and fog. Rather than having to set them up manually, you use Corona very much like you would a camera. The bloom and glare it creates are also very nice.
AMD: What workstation do you use?
JL: I've got quite an unusual setup, because of my background as a photographer, I'm on a iMac Pro®, but running Windows® as well as macOS®. It has a 10-core processor, an AMD Radeon™ Pro Vega 56 graphics card, and 64GB of RAM, and I've been blown away by it. The AMD graphics card in it runs a 5K screen flawlessly.
I can have Windows running in Parallels rendering something in 3ds Max and Corona, while I'm working in Photoshop in the Mac operating system. I can also have my email open, music playing, all the rest of it. The iMac Pro is completely happy.
AMD: How does your Adobe software perform on the iMac Pro's AMD graphics card?
JL: The iMac Pro has handled everything flawlessly so far. I’ve yet to find anything in any Adobe package which stresses it much. Running the latest Photoshop CC, it’s able to process really hefty files – upwards of 200 layers – without skipping a beat. Likewise, Lightroom processes my RAW files in an instant with excellent colour reproduction on the 5K display.
The Radeon Vega has definitely been beneficial with all of the retouching work as Adobe has begun to utilise GPU hardware more in recent years: both through Apple’s graphics API, Metal, and the open standards OpenGL and OpenCL, used in Lightroom and Photoshop.
In the small amount of video work that I do, I’ve been really impressed with the iMac Pro’s handling of Final Cut Pro [which also uses OpenCL] and After Effects when working with 4K or 5K video. I often use layers of video with different blend modes and bloom effects, and everything remains buttery smooth without pre-rendering.
AMD: What has the experience of running 3ds Max on Parallels been like?
JL: I was nervous about it at first, but I was keen to keep my photography workflows matching my CG ones. It does all the things I need. I'm using Parallels Pro [Parallels Desktop for Mac Pro Edition], which allows more cores and RAM to be dedicated to Windows, which is a massive benefit. It's as if I'm virtualising a 20-core, 50GB RAM Windows machine, but in my iMac Pro.
And I get all of the other benefits of working on a Mac with a high-end GPU. I run it so the Windows side isn't connected to the internet, to reduce network traffic and to avoid the need to run antivirus software and a firewall.
AMD: What hardware do you use for rendering?
JL: I have two Windows towers running AMD processors, one of which I used as a workstation before I got the iMac Pro. The CPUs are quite old, but I currently have an Amazon basket full of new AMD Ryzen™ Threadripper™ chips that I'm pricing up.
AMD: What is it about the Ryzen Threadripper chips that attracts you to them?
JL: The sheer power for the cost. It's a fraction of what you'd pay for the equivalent Intel® chips. I’ve been looking at the 16-core Ryzen™ Threadripper™ 2950X for my rendering node, and the bang-for-the-buck on those processors is crazy.
One thing is clear when exploring James Lawley's work: an eye for detail and an understanding of light will always create a powerful image, no matter in which medium it is created. In many of James's images, it is difficult to distinguish between what is real and what is CG. This is the type of work that AMD applauds. Through its support for non-proprietary technologies like OpenCL, AMD is helping to break down barriers between disciplines. Technology should never be a barrier to creativity.
James Lawley is a valued customer of AMD. The views of James Lawley are his own opinions and may not represent AMD’s positions, strategies or opinions. Any benchmarks or performance statements from James Lawley have not been verified by AMD. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied. GD-5
©2018 Advanced Micro Devices, Inc. All rights reserved. AMD, the AMD Arrow logo, Ryzen, Threadripper, Radeon and combinations thereof are trademarks of Advanced Micro Devices, Inc. Other product names used in this publication are for identification purposes only and may be trademarks of their respective companies. GD-28