How Exhibition Designers Mock Up a Gallery Before a Single Crate Arrives
A gallery hang, National Museum in Kielce. Photo: Krzysztof Poplawski, CC BY 4.0, Wikimedia Commons.
One room of 180 square metres, 24 loaned works, six weeks to the opening and two lenders who would not ship until they had seen a hang plan. The designer on this show built the whole gallery in SketchUp Pro during the first week, tried three layouts inside it, and handed the installers a plan that moved one plinth on the day and rehung nothing, which for a mixed show of paintings, sculpture and vitrines is about as good as installation week gets. Fourteen hours went into the model. How those hours were spent is worth setting out, since the method is the same for a municipal gallery and a commercial one and the only thing that changes is the numbers.
The survey comes first and takes the longest
Two hours with a laser measure and a notebook, most of it spent on things other than the size of the room. The room happens to be 18 by 10 metres under a 4.2 metre soffit, but a designer who stops there has measured the easy part and skipped the part that decides the hang, which is everything that cannot be moved. On this room that meant two double-leaf doors at 1.8 metres, a pair of 450 mm columns standing at 6 and 12 metres along the length, eight floor sockets, four lighting track runs at 3.8 metres, a hanging rail at 3.4 metres and a fire exit route along the north wall that has to stay 1.2 metres clear for the whole run of the show. The fire route went into the notebook as a fixed line rather than a preference, because the fire officer walks it on the morning of the opening and has no interest in the sight lines.
An hour on the shell, and why it gets locked
There is nothing clever about the shell and it should not take more than an hour. The floor slab was drawn with the Rectangle tool, the walls pulled up to the soffit with Push/Pull, the two door openings cut and the two columns placed where the survey had put them rather than where the venue's old plan said they were, which differed by 300 mm on the second column and would have mattered later. Guide lines went in with the Tape Measure at every socket and under every track run, so that the lighting and the power were visible on the plan from the start instead of being remembered at the end. The shell then went on its own tag and was locked, for the simple reason that a hang involves dragging a great many boxes around and sooner or later one of them nudges a wall. A plan scene and one scene per wall were saved before anything was placed, which meant that after every change the same views could be exported again and laid side by side.
Every work becomes a box at its loan-agreement size
Nobody models a painting, and nobody should. Each work became a box drawn at the height, width and depth in its loan agreement, with the work number in large type on the front face so it could be read across the room in any scene. The works on paper caused the only argument, since the loan agreement listed the sheet size and the framer's glazed dimension turned out to be 90 mm bigger in each direction, which would have thrown every gap on the short wall if anyone had trusted the paperwork.
Group
Count
Box size
Frame depth
Paintings, large
6
1,800 by 1,400 mm
60 mm
Paintings, medium
8
1,000 by 800 mm
50 mm
Works on paper, framed
4
600 by 800 mm, glazed
40 mm
Sculpture, floor
2
1,600 mm tall
None
Small objects for plinths
4
Under 300 mm
None
Frame depth gets less attention than it deserves, because a 60 mm frame on a large canvas throws a noticeably heavier shadow line under track lighting than a slim one and pushes the label further out than the curator wanted. The weights went onto each box's tag as well, since weight is what decides the plinth, and the 42 kg stone fragment was the sole reason a low 1,200 mm square plinth existed in this show at all.
Plinths, vitrines and benches
The plinths were boxes drawn in place, four standard ones at 500 mm square and a metre tall and two low ones at 1,200 mm square and 300 mm, painted MDF in the real room and a flat grey in the model so they never competed with the works for attention in a scene. The three table vitrines and the two benches were a different matter, since a vitrine drawn as a solid grey box hides the very sight line through the glass that bench 2 exists to offer. Those came from a free library of SketchUp furniture models, imported as .skp files and rescaled to the joiner's drawings with the Tape Measure, which takes about a minute per piece once the habit is there: two points across the width, the real width typed in, a yes to SketchUp's question about resizing the whole model, and a move onto the floor guide. Six standing visitor figures at 1,700 mm went in last on their own tag, partly for scale in every exported view and partly for the curator, who like most curators does not read plans and does read a picture with people in it.
Hanging heights and the camera test
The centre line went at 1,500 mm, which is where most galleries put it when the audience includes children and wheelchair users, and the big paintings dropped a little below that so their tops cleared the 3.4 metre rail with room for the hanging wires. The trouble came from an unglamorous number, the 600 mm minimum gap between works. On paper it looks generous. On the long wall it turned out to be the thing that decided how many large paintings would fit, and the answer was four. The curator had been counting on five since the loan list was agreed, and an hour of nudging boxes about in the model never found the fifth.
Hanging height
Value
Centre line
1,500 mm
Large paintings, bottom edge
800 mm
Medium works and paper, bottom edge
1,100 mm
Labels
1,400 mm, 100 mm to the right of the work
Gap between works
600 mm minimum
Sight lines got checked the only way that has ever worked, which is to put the camera where a visitor will actually be standing or sitting and look. From the first door at a standing eye height, the far wall read well and sculpture 2 vanished entirely behind the second column, which nobody had predicted from the plan. From the first bench, seated, the run of paintings 7 to 10 read well. From the second bench the vitrine glass threw back a reflection from the track above it, so the bench moved a metre further back and two fixtures moved with it. Three layouts in four hours, which is cheap for what they turned up. Layout one was the curator's chronological run, left to right, and it put the heaviest paintings on the short walls, where they looked thin from the door and crowded from the benches. Layout two moved the big paintings to the far wall and brought the objects into the middle of the room, which fixed the entrance view and promptly hid sculpture 2 behind a column. Layout three kept two and dragged sculpture 2 forward, 12.5 metres from the first door down to 9, and that is the version the installers got.
Circulation, measured rather than eyeballed
Widths were measured in the plan scene with the Tape Measure against the minimums the venue supplied, and one of them failed. The distance from the benches to the wall works came to 2,400 mm against a 2,000 mm minimum, and the fire route along the north wall came to exactly 1,200 mm, which is the number the fire officer checks and the reason nothing else was allowed within a hand's width of it. The gap between plinths 1 and 2, the narrowest in the room, was 1,500 mm against 1,200. The west side of sculpture 1 was the failure, at 1,100 mm in the first layout, and the plinth beside it moved 200 mm east to fix it. A 1,500 mm wheelchair turning circle was drawn at the vitrines on its own tag and left visible on the installers' plan, so that nobody could park a trolley in it on the day and claim it was not marked.
Lighting in the model
Shadows were switched on with the date set to opening night at 7 p.m., which showed where the window light would still be crossing the south wall at that hour in that month, something no amount of experience predicts reliably. Fixture positions were then marked on the track runs as small cylinders at 3.8 metres, aimed at 30 degrees from vertical at the wall works, the usual compromise between glare off the glazing and shadow from the frames. The count came to 36 fixtures with 12 of them for the plinths and cases, and the lighting plan drawn from those marks let the electrician focus the whole room in one evening, which the venue said was the shortest focus it had had.
What left the model
Five documents went out, and the installers used one of them far more than the rest. The plan at 1:50 had every plinth, bench and vitrine placed and every work numbered, the four wall elevations at 1:20 carried bottom-edge heights and label positions, and the lighting plan carried the fixture marks. The schedule, a single sheet listing work number, size, wall and bottom-edge height, was the document in the installers' hands all week while the plan stayed pinned to the wall. The scenes went out as images too, so that the curator could sign off the entrance view without opening SketchUp, which she had no intention of doing.
Installation week
One plinth moved on the floor, no work was rehung and two fixtures were added, which on a show of this size is a quiet week. The crates were opened in the order the plan had set, sculptures first so the floor pieces were in before the walls got busy, vitrine material second so the cases could be closed early and stop being a hazard, large paintings third for the far wall, and the medium works and paper last. The show opened on time. The two lenders who had wanted a hang plan before they would ship got one in week one with their own works numbered on it, which is why their crates arrived in week four rather than week five, and why the designer will build the model in week one on the next show as well.
The habit worth taking from all of this is not the software but the sequence: survey everything that cannot move before drawing anything, draw every work at its real size with its weight on it, and stand the camera where the visitors will stand before anyone books a van.