CCD Advantage

[Sonnar Brian]

Why would anyone design a detector array without two-axis symmetry? Aside from the D1x, which used odd binning.

Time to drag out the DCS200 which does not have a Mosaic filter.

 

[antiquark]

Thanks, Antiquark.

Here are Antiquark's COF images, as surface plots with intensity coded as height. 1.4 on the left, 1.8 on the right. What an interesting result: the COF's are asymmetric, and (apparently) more so at wider aperture!

At f/1.4 we see very obvious horizontal bands of light falloff at the top and bottom, and a trace of this behavior is still visible at f/1.8. This result is consistent with a photosite geometry in which light rays can be detected at shallower angles when coming from the left or right sides of the lens's exit pupil than when they are coming from the top or bottom of the exit pupil.

This behavior could be due to the design of the photosites, the microlenses, or both.

I noticed that too, but wasn't sure if it was due to my setup. The light source was a flashlight with aluminum foil on the front punctured with a 2mm hole. I thought maybe the beam was directional, so I re-tested and added two layers of wax paper behind that to diffuse things.

Still the bokeh showed the darkening at the top and bottom. However, when I tried my D90 (much better sensor) the effect seems to have vanished. So yes, something's not optimal with the D40 sensor design.

An odd effect of that "anisotropic" bokeh is that the DOF would (theoretically) be different along the X vs the Y axis. However, it might not be measurable in a practical sense.

 

[semilog]

@ Brian,

It's probably not the array that lacks symmetry (although many arrays with rectangular pixel pitch have been made, predominantly for video applications) — it's the photosites themselves.

Even if the pixel *array* is a square grid, the photosites may have anisotropic light sensitivity due to the geometry of the wiring at each pixel. Asymmetric layouts may make possible lower shading factors (and thus higher overall quantum efficiency) at the cost of some angle-of-incidence anisotropy.

 

[semilog]

I noticed that too, but wasn't sure if it was due to my setup.

Easy to check: rotate the camera 90 degrees to the right or left.

 

[antiquark]

Easy to check: rotate the camera 90 degrees to the right or left.

Good idea. I tried it, and the dark edges stay aligned with the top and bottom of the CCD, so it's definitely the sensor.

 

[semilog]

especially in APS-C sensors. I think the pixel alignment within the sites is different in an FX sensor.. just guessing though!
p.

I suspect it's not so much APS-C vs. FF, but (1) sensor generation, and even more so (2) pixel size, with smaller pixels having more shading issues. This stands to reason: think of the acceptance angle of a cardboard tube like the one at the center of a toilet paper roll, versus the acceptance angle of a coffee cup. Same well height, different cross-sectional area.

Look at the charts in the Luminous Landscape article. The Canon 7d, with its teeny tiny pixels, is recording enormous losses (a full stop at f/1.2) for such a recent camera.

 

[antiquark]

Here's a better pic of the bokeh blob. I decreased the focus distance to the minimum, 0.45m, used a tripod, lower ISO to decrease noise.

You know, if you look closely, it looks like the left and right edges ALSO have some darkening. So it's like there's a rectangular mask that's slightly smaller than the circle of confusion.

Now I'm REALLY interested to see what an f1.2 lens would look like on an older sensor like a D40. If you expanded the disc below, the top and bottom edges might get quite dark with at f/1.2... in effect, confirming the article at luminous landscape!