We have produced, in the eight-year history of the studio, exactly four photoreal renderings. Three of them were for clients who had asked for them. One was a mistake we made on our own. I want to write briefly about why this number is so low, why it is unlikely to grow much in the next eight years, and what we use instead.
The case against the rendering, properly stated, is not a case against the technique. The case is against a habit: the habit of designing toward the picture rather than designing toward the building. Once a rendering exists in a project — once it has been printed, emailed to a client's spouse, sent to a contractor, hung on the wall of the office — it begins to behave like a contract. The room must look like the picture. The picture begins to make the decisions.
§ I.A drawing proposes; a picture promises.
When I was a student at RISD, my first instructor used a phrase I have never forgotten. A drawing proposes. A picture promises. A pencil section, a half-finished plan with a north arrow drawn freehand, a 1:20 elevation with the joinery scribbled in but the glazing left blank — these are drawings. They invite you to argue with them. They are explicit about what is not yet decided. The empty parts of a drawing are doing real work; they are saying this is open.
A photoreal rendering does not, by its nature, allow that. It cannot say this is open; it has to say this is what it will look like at four o'clock on a Thursday in October. It has to commit to a sky, a tree species, a piece of furniture in the corner, the colour temperature of the light. None of these commitments are real — the sky is from a stock library, the tree is a polygon mesh from somebody's plug-in folder, the furniture is from a Vitra catalogue scan — but the client cannot tell which commitments are load-bearing and which are decorative. They have to assume all of them are load-bearing. They have to.
The empty parts of a drawing are doing real work. They say: this is still open. Renderings cannot say that.
§ II.The four renderings, briefly.
The four renderings the studio has produced were:
One, in 2018, for a developer client in DUMBO who required them in the proposal package. The image showed a ground-floor lobby in a building we did not, in the end, get to design. We do not regret making it; we never built the room.
Two, in 2020, for the family who eventually commissioned the Carroll Gardens townhouse. The renderings showed the rear extension at the schematic stage. We regret these. The clients liked them so much that, when we proposed a change to the glazing pattern at design development — a change we believed in, a change that made the room demonstrably better — they spent four meetings trying to talk us back to the picture. The picture was made in three days. The room is going to last a hundred years. We had given the wrong artefact authority.
Three and four, in 2022, for the Pavilion at Pier 6 — both produced for the Brooklyn Bridge Park Corporation as part of a public design review. These we regret a little less, because the rendering convention is genuinely useful in public-process settings: it is the only language a community board can read at speed. We still believe that, in those meetings, we lost some nuance about the materiality of the cross-laminated spruce. The renderings made the timber look slicker than it actually is. The actual posts have grain in them that the renderer never managed.
§ III.What we use instead.
I do not want this essay to read like a complaint. The genuinely useful question is: what do we present to a client, in lieu of a rendering, that does the work of communication without doing the harm of premature commitment?
The answer is: a sequence. A rendering tries to do, in one image, the work that we believe should be done in five or six different artefacts, each of which is honest about what it is and is not deciding. The sequence we use, in order, is roughly this.
First, the site model: chipboard, 1:200, unstained, with the proposed massing in basswood. The model deliberately abstracts away every interior question — there are no windows, no doors, no roofing. It commits only to footprint, height, and the relationship to the neighbours. Every other question is held open.
Second, the 1:50 plan and section, drawn in pencil over half a dozen sheets of tracing paper. The plan commits to circulation, structure, and the basic geometry of the rooms. It does not commit to finishes; it does not commit to the joinery; it usually does not commit to whether a wall is plaster or board.
Third, the material board: a long oak tray, two feet by four, with samples laid out in approximate spatial position. Stone for the hearth, plaster for the walls, the actual species of timber we are proposing for the floors, the actual glass for the windows in the actual frame. The client picks the samples up. They handle them. They are allowed to disagree with them. The board is an argument, not a picture.
Fourth, when the project warrants it, a perspective sketch — drawn freehand, in pencil and watercolour wash, usually by Anders or by the project architect, sometimes by both in conversation. These are sketches, explicitly. They have a line weight. They have erased corrections. They look like what they are: a proposition, drawn quickly, by a human, on a piece of paper that may get coffee on it before the meeting is over.
Fifth, and only at the very end, a set of 1:5 detail drawings. The detail drawing is the artefact in our process that comes closest to the rigour of a rendering, because by the time you are drawing a 1:5 of a window jamb, you have committed to almost everything. The window is going to look a particular way. The plaster is going to die into the frame in a particular way. The drawing is then doing the same job a rendering would do — except that the drawing knows it is a construction document, not a picture. It is honest about its purpose.
§ IV.The longer view.
I am aware that the rest of the profession has, over the last fifteen years, decided this argument. Renderings have won. They have won the planning meeting, the developer pitch, the Instagram feed, the architecture school crit. There are practices I admire enormously who produce renderings of extraordinary quality — renderings that are, in their way, real architectural thinking, made in pixels. I do not want this essay to be read as a complaint about them.
I just want to be honest about why the studio does not work that way. We are slow. We are physical. We design at a long table, with samples on it, and the rooms we end up with tend to feel — we hope — like rooms that have been argued about, not rooms that have been signed off in a screenshot. The renderings will, of course, exist after we are done; the photographer will arrive, the magazines will publish them, and someone will put a polygon version in their portfolio. But by then the room will be the thing. The picture will follow.