# Why was the Vera C. Rubin Observatory LSST Camera tested with a broccoli and a pinhole?

The SLAC National Accelerator Laboratory video Vera C. Rubin Observatory LSST Camera (caution! annoyingly loud music, lower volume before playing) shows a Romanesco broccoli (which I'd assumed was a 3D printed mathematical surface of that name until I checked and found out that this is the name of an actual vegetable i.e. Brassica oleracea) as the object of a pinhole camera projecting its image on to the astoundingly large CCD array.

There are likely several advantages to using the pinhole rather than a well-corrected projection lens, is it possible to say what they are?

Is it also possible to elaborate on the choice of object for the camera?

Caution! annoyingly loud music, lower volume before playing:

• Broccoli is fractal. No, really, it's a decent physical object with repetitive, high-detail structures. Lots of stuff to try to resolve in the image. Sep 9 '20 at 15:38

From the first article you linked:

Taking the first 3,200-megapixel images of a variety of objects, including a head of Romanesco – a type of broccoli – that was chosen for its very detailed surface structure, was one of these tests. To do so without a fully assembled camera, the SLAC team used a 150-micron pinhole to project images onto the focal plane.

Which implies:

• the advantage of using a pinhole is that it allows the sensors to be tested without requiring the full camera to be assembled yet.
• the Romanesco was chosen to test the sensors' ability to image an object with detailed structure.
• I think that the circa 200 wavelength wide pinhole also offers the advantages over a projection lens of being very easy to model unambiguously with basic optics (MTF, scattered light, etc) not require any focus or introduce any errors if focus isn't correct, and automatically produces a fairly low light flux which might be suitable for the array, but those aren't mentioned in the linked articles.
– uhoh
Sep 9 '20 at 7:11
• @uhoh - you can speculate all you want but I'm going with the reason the people involved actually stated for why they used a pinhole.
– user24157
Sep 9 '20 at 7:13
• Ya I know. There may be additional sources about this test beyond those I'd included in the original post, and maybe this is a technique tried before on other cameras, or maybe it's totally unique. I suppose the "broccoli" will turn out to be unique for sure. Let's see what else turns up
– uhoh
Sep 9 '20 at 9:38

Supplemental speculation:

In order to produce an image on the sensor from a finite distance, one would have to design and build a special and expensive lens with a conjugate focus close enough to be practical, and a well characterized and flat field (or a curved object plane).

Then one would presumably discard this lens after this simple test.

Instead, the ~200$$\lambda$$ pinhole offers the following advantages over developing and adding some kind of projection lens

• being very easy to model unambiguously with basic optics (MTF, scattered light, etc)
• not requiring any focus or introduce any errors if focus isn't correct
• automatically produces a fairly low light flux which might be suitable for the array

While none of these are mentioned in the linked articles in the other answer, they are sort-of self-evident from basic optics.

Also somewhat self-evident is that the internet can not get enough fractal vegetables (or fruit, see Vi Hart's video below), and that the simple yet beautiful pinhole camera + fractal vegetable demonstration is cool and has some public outreach value.

Click images for full size. The first three are screenshots from the video and the last is from Wikimedia's Romanesco broccoli (Brassica oleracea).