View Single Post

Nothing Beats Lens And, Or Sensor Surface Area
Old 12-09-2018   #19
Registered User
willie_901's Avatar
willie_901 is offline
Join Date: Dec 2005
Posts: 5,373
Nothing Beats Lens And, Or Sensor Surface Area

Originally Posted by Bill Pierce View Post
...In your experience, how much does sensor size effect image quality? ...
It is best to think about the lens and sensor together. The surface area for both affect the maximum possible signal level. In comparisons where lens differences (aperture and T factors) are irrelevant, then sensor area alone affects the maximum attainable image quality.

Originally Posted by Bill Pierce View Post
What are the advantages of large sensors, and what are the advantages of small sensors?
As sensor area increases, the maximum possible signal level increases.

In terms of technical image quality (where composition, creativity, use of light and many other human factors are ignored) sensor surface area determines the maximum signal-to-noise ratio for the data. The signal contains the information we need (the spatial illuminance for the scene). The noise determines the uncertainty (error levels) for that information.
  • More surface area means more signal
  • More signal means lower relative levels of photon (shot) noise
  • For contemporary platforms, our camera's electronic noise levels are similar. Increasing sensor area does not necessarily increase electronic noise levels

Whenever sensor underexposure is unavoidable, SNR makes a difference. The most obvious impact is on shadow region detail. Of course, at low ambient light levels, sensor underexposure reaches levels where even highlight regions are have low SNRs. With newer cameras most of this noise is photon noise.

In bright scenes SNR is also important. Analog dynamic range is directly determined by SNR. SNR advantage that may be obvious in low light is also relevant for the brightest scenes.

Small sensors mean smaller and lighter cameras. In some circumstances this is a significant advantage. The extreme examples are phone cameras.

But smaller sensors do not necessarily mean smaller and lighter lenses. Increasing sensor area and increasing aperture area both increase signal levels.

So, signal levels can be identical for a large sensor with a small maximum apertures lens and a small sensor and a large maximum aperture lens.The size and weight advantage of smaller sensor cameras can be offset by the size and weight of larger lenses.

But maximum possible lens apertures have practical limitations. When DOF is not considered, large sensors will have superior SNRs.

Right now another advantage of smaller sensors is cost. A mirrorless medium format camera (44 x 33 mm sensor) is more expensive than many 24 X 36 mm cameras. The Phase 1 IQ3 has 54 x 40 mm sensor. This is the highest SNR and dynamic range still camera one can buy. It also costs $50K.

Originally Posted by Bill Pierce View Post
How important overall is sensor size in your photography?
It is very important.

I own APS-C cameras because the compromise between maximum SNR and convenience (size, weight, cost and OVF with RF style composition) is perfect for my interests.

If I was interested in landscape photography I would own a mirrorless medium format camera.

If I was interested in action photography I would own a camera with a 24 X 36 mm sensor.

In 2008 I started to use APS-C DSLRs. When I switched from a D200 to a D300 I noticed a significant difference in technical IQ due to SNR differences. Then I switched to a D700. Again the improvement in technical IQ was obvious.

I noticed similar technical IQ differences between the X100, XT-1/X-100T and the X-Pro 2. Here the difference SNR is due to improvements in sensor assembly electronics.

When I first started using the XT-1 I discovered it slightly out-performed the D700 at base ISO. At the time analog dynamic range at base ISO and shadow region IQ was very important. The D700's surface area advantage was offset by the improvements in sensor noise levels.
Basically, I mean, ah—well, let’s say that for me anyway when a photograph is interesting, it’s interesting because of the kind of photographic problem it states—which has to do with the . . . contest between content and form.
Garry Winogrand
  Reply With Quote