Panoramic Motion Device

Figure 1 – Liedke's Box and Map of Site
Note the location of the music seller's stall marked with an X on the period Map.

1. The Problem

Here I describe a new approach to the problems that the Fabritius Pantheon helped me to think through. The question was, how did the missing viewing box for Fabritius's View in Delft work? What was the curvature of the image and where was the point of sight located? To better follow, you should read about the Pantheon first.

 

Walter Liedtke[1] thought that the missing box was like another Dutch "perspective box." In these boxes a distorted image, or anamorphosis, is projected on the inside surface and only when an observer looks from a specific point of sight, defined by a peephole, does the image coalesce as a normal perspective view. In theory, you can project an anamorphosis on any surface, and with any point of sight. So the point of sight wouldn't have to be at the centre of Fabritius' curved image. In Figure 1 (#5.) you can see that Liedke's point of sight, at the apex of the triangle, is well outside the centre of his proposed curve.

 

I believe that Fabritius made the image with a rotating camera obscura so that the point of sight in making the image was in the centre of rotation. The point of sight when exhibiting the image would likewise have to be in the centre of rotation. All the evidence pointed in this direction (see my discussion) except for one thing. With Liedke's curve people reported an immersive wrap-around experience as though they were sitting in the instrument seller's stall with Fabritius.

 

Figure 2 – Panoramic Motion Device

Struggling to make sense of this anomaly, I returned to Liedke's paper. Looking at his period map of the site (Figure 1) I realized the error I had made. Leidke’s box has an angle-of-view of 60°, but his estimate of the angle-of-view of the painting is 90°. Note the lines that he has drawn outward from the "x" on his map. I had set my panorama up for 60° of arc; if I remade it instead for 90° of arc, the viewer would gain more of that wrap-around experience that had been lacking. There was no way to fix the old device, so I had to make a new one.

 

Figure 3 – Pivoting Sight

2. Design of the New Device

As I mapped out the geometry of the new device I realized that it could never work with a peephole. So instead I designed a viewing mechanism that allowed the eye to swivel and the head to rotate. The pivot is located at the centre of rotation of the right eye (see Figure 2). In the first device the box serves to protect the imagined space of the painting from intrusive evidence of the real space of the world. Here I let the hood of the viewing device perform that masking function and removed the walls of the box for better illumination of the image.

 

Just for fun, instead of a knurled thumbscrew to secure the viewing-piece on its vertical post, I decided to try a decorative brass wedge (Figure 3). My aim is to bring attention to the way that viewing is mechanically mediated.

3. The Results

The scene now does seem to surround you as you scan it. A still image of the lute shows it to be far less distorted than in Liedtke's version. But the strongest evidence that the painting is best viewed this way as a panorama, is the way it rectifies the shadows. Shadows in Fabritius' painting – of the lute against the wall and of a large tree against building facades – are cast in almost opposite directions. That is strange given that the sun's rays fall parallel on the earth. To test how shadows look in a panorama, I made a 90° iPhone panorama in my orchard (Figure 4).

Figure 4 – Shadows in a Panorama
Note how the shadows of the tree trunks fall in opposite directions.

 

Look at the shadows cast by the trunks of the two foreground trees. You can see how they fall in opposite directions. When I took the photo, I rotated my head and body slowly from the left a full quarter turn to the right. My perceptual system "knows" at some pre-conscious level that this action will make the scene appear to rotate, and lines, like the parallel shadow lines, to change their orientation. I perceive the space correctly only when my body moves. Only the pivoting head-piece will recapitulate that motion for Fabritius' panorama and make it look right.

4. Assessment

The evidence points to Fabritius having used a rotating camera obscura to make this painting. However, If he had tried to exhibit it in a correspondingly curved viewing box he would have faced difficulties, mostly because of the small scale of his painting. His was only 15 cm high. My reproduction of it is 21 cm high and that gave me some difficulties. The viewing device has to be too close to the image plate, so the materiality and flaws of the canvas become evident. It would be worse with Fabritius' original. Also, the viewer's chin barely fits over the pivot of the device and would not fit at all if scaled down. Fabritius might have used a different design – something closer to a pinula. These were common as sights for surveying and navigating devices and could be made to swivel. It also might be that Fabritius tried and failed to come up with a suitable exhibition device. One thing that I have learned from experimental media archaeology is that it is often difficult to get these things right the first time.

 

Contact: rbantjes@stfx.ca