The results of the experiment show that increasing a camera’s gain will increase the noise in the image. Depending on the application, this may be an acceptable trade-off based on the requirements of the camera’s aperture and shutter speed. The only way to further reduce the noise is to actually select a camera with less noise.

As expected, the Lt425C produce a noisier image in a shorter period of time since it is more sensitive than the Lt965RC and has a higher read noise level.

When configuring a camera, there are a number of ways to make the image brighter: increase the exposure time, open the lens’ iris, or increase the camera’s analog gain. Each application usually has limits for each parameter. For instance, increasing the exposure time can make moving objects too blurry to image. Opening the lens’ aperture will reduce the camera’s depth of field (some lenses also do not have the ability to modify their aperture). And, increasing the analog gain will increase the noise in the image.

This technique, now known as the Foucault knife-test, is incredibly sensitive: bulges or hollows in a mirror surface with a relief as little as one millionth of an inch are easily detectable. Armed with his knife-edge, Foucault was able to produce mirrors with an accuracy of figure never before achieved. His technique is still much used today.

Like John Hadley some 200 years earlier, Foucault placed a pinhole source at the mirror's center of curvature and arranged the image to be formed alongside the source. However, unlike Hadley, Foucault examined the rays converging to a focus by placing his eye behind a knife-edge, which he then slowly introduced into the image. If the surface of the mirror darkened uniformly, he knew the mirror was spherical; if it didn't, he was able to deduce where and by how much the mirror surface deviated from sphericity.

JavaScript seems to be disabled in your browser. You must have JavaScript enabled in your browser to utilize the functionality of this website.

If a situation arises where the exposure time cannot surpass a certain duration, the camera’s analog gain would need to be increased to create two images of matching responsivity. To illustrate this, the exposure time of the Lt965RC was set to 23.5ms and its gain increased to 9.2 to maintain the same scene brightness level as the first image captured by the Lt425C. The results of these changes can be seen below in the cropped image of the same area seen in the previous images (digital gain of 4 added, once again).

Below are two cropped images of the resolution target taken with each camera. The target’s white background is highlighted by the red rectangle in the experiment setup image above. They were both captured using an analog gain of 3.2 and an aperture of f/16. The exposure time used for the Lt425C was 23.5 ms and 91 ms for the Lt965RC. A digital gain of 4 was added to both images to help the human eye see the visual data.

Comparing all three images, it can be seen that when gain is added to an image the noise is also amplified. This is seen numerically by the increase in the range between the minimum and maximum pixel values as well as in the variance of the image.

In the following experiment we set each identical lens’ aperture to f/16 to create a dark scene for the cameras. We then imaged a resolution target with two different cameras (Lumenera Lt425C and Lumenera Lt965RC) to demonstrate the impact of a camera’s sensitivity on exposure time and gain required to achieve similar results. Because we worked with a darkened scene, we will also be able to observe the impact that gain has on noise.

The Foucault knife-edge test is a test of optical surfaces, developed by Léon Foucault, that uses reflection and geometrical optical principles to amplify shadows of defects on a telescope mirror so that they are made easily visible.

The Lumenera Lt425C camera has larger pixels and is therefore more sensitive than the Lumenera Lt965RC camera. This means that it requires less light than the Lt965RC to produce an image of similar brightness with the same camera settings.