Realism is vital when visualizing architecture. We need to be able to effectively communicate thoughts, ideas, and visions in a way that's both understandable and relatable. But with an industry as vast and populated as arch-viz, how do we define realism?
Architectural visualization has come a long way over the past few years. Through the many changes, however, the goal of photo-realism has remained. But if that is the only goal we've set, what come next?
In my early years as a 3D artist, I remember getting lost in platforms such as Art Station. I would stare in wonder as I came across image after image that was seemingly indistinguishable from photography. I would marvel at the fact that there were people out there who were so skilled and had such a unique understanding of the world around them that they were able to translate that into their imagery. It began an obsession with arch-viz that drove me to practice as much as I possibly could. Eventually, I finally produced an image that I was proud of. The more I practiced the more I improved until eventually I finally produced an image that I was proud of. I’d started to crack the secret mystery that was photo-realism.
Or at least I thought.
As the years went on, my skills continued to improve until finally, something happened. I began to find the images I once admired... boring. Don't get me wrong, the technical ability of the artists still blew my mind, but with each passing year, that level of accuracy that once seemed like magic was now becoming a standard. As software advanced, the ability to achieve highly convincing results was becoming more accessible and yet my taste for this type of realism started to fade.
I was underwhelmed, frustrated, and confused. After I spent so much time and work trying to attain that goal of photo-realism, it felt like the moment I began to arrive there I became dissatisfied. It wasn't as if I disliked the images I was creating, but I felt that there had to be more to this idea of perfection than just creating highly accurate pictures. I decided to re-evaluate what photo-realism meant both to me as an artist, as well as to how it applied to arch-viz in general.
Oxford defines photorealism as:
A style of art and sculpture characterized by the highly detailed depiction of ordinary life with the impersonality of a photograph.
(in computer graphics) the rendering of images in an extremely realistic way.
Highly detailed depiction of ordinary life with the impersonality of a photograph.
This was my turning point. At best, photo-realism established a basis by which we could familiarly illustrate an idea. At worst, it meant the inhibition of creativity. It meant that as an idea, photo-realism would contradict what I was trying to achieve as an artist.
I still wasn't convinced. Photo-realism is a goal that is maintained by artists all over the world, many of whom are far more talented than I was. Surely I was missing something.
I decided to take a step back and look at photo-realism from a different perspective. I began to investigate other mediums such as film, concept design, and photography, and I realized something important. Photo-realism as we see it in 3D isn't simply a set of rules or themes that are followed to achieve 'realism' in our work. In fact, the more I learned, the more I started to entertain the idea that photographic realism may not even be attainable as a CG artist.
I know, blasphemy in the eyes of any technologist in the CG industry, but bear with me.
A common theory articulated by film and photography theorist such as Andre Bazin, indicate that a photograph is a mechanical instrument. It's a prosthetic of sorts that allows the viewer to witness an actual moment in time, with an actual subject on the other end of a lens. This idea of realism in photography can never translate to the works of a CG artist as the images we create have no physical referent to draw on. They're a projection of an idea that doesn't exist in reality and so what is perceived as realism becomes somewhat subjective.
I was intrigued, but It was all becoming a little too philosophical for my liking.
Photo-Realism in Context
With that in mind, I wanted to define what is it that we interpret as photo-realism?
Is it the technical elements that we see in photography? Aspects such as lighting, shadow, reflection, and color? These are all things that the real world (and by association, our perception of photo-realism) are limited by. Lighting and shadow need a source and are limited by what's attainable in a physical space. Colour is finite and influenced by factors that we have limited control over. In modern photography, we remove these constraints through digital editing, artificial lighting, and compositing. At the moment this occurs however, the photos are no longer 'realistic'. They're a manipulation of realism to create something aesthetically pleasing.
I came to the opinion that realism at its core is reliant on the understanding and believably of the viewer. Let's take a look at modern-day cinema. In Avengers: End Game, it's clear that much of the film is CGI. We know this before we watch it, yet when we see Ironman fly across a city, or Thor become engulfed in lightning, we voluntarily allow ourselves to believe its validity. At that moment we aren't concerned about the accuracy of the smoke particles, whether the colors are realistic, or whether the light sources are proportionate to what would occur in the real world. At that moment what we see is real, and powerful, because our acceptance of its impossibility is what allows us to manifest that response.
Now, is Architectural Visualization the same as Blockbuster cinema? not quite. However, I believe the same principals can apply to our industry. When we create our imagery we are projecting an idea that doesn't exist and as such the theoretical idea of photo-realism also can't exist. But subjective realism is what we do. In translating a raw concept into a medium that's tangible and relatable to our audience, we have created realism.
From a technical perspective, our tools and software bring us as close as possible by allowing us to emulate the parameters of what we consider to be photo-realistic. Each software release brings us closer to technical accuracy, but I don't believe those aspects are what makes our work realistic, or even appealing for that matter. No matter how beautiful our textures are, or how accurate our lighting is, the images we create will never be ‘real’. So why limit our creativity by trying to make them that way?
We're lucky enough to be part of an era whereby the tools we are using are achieving the technical aspects of arch-viz more or less on their own. 'Accuracy' is no longer the metric that defines the images we create. Soon, anyone will be able to adjust a few settings and get something that resembles the idea of photo-realism. But where do we go from there?
Photographs are about showing what is. Visualization is about showing what could be, and I think sometimes the strive for photo-realism causes us to forget that.