This summary should be read in conjunction with the full explanatory article that you can see  here. Note that the analysis on this page does not include a discussion of the various complex issues that can arise in practice when estimating pixel density and the the pixel pitch or area of individual pixels. It is recommended that you study a detailed technical article if you would like to become familiar with these issues. For example, you may find this  DPR forum discussion  about pixel density and pixel size to be helpful. Therefore, the calculations set out below are presented for the purpose of calculating only a very approximate measurement of pixel density, pixel pitch, and the area of one pixel, which can be used for comparing the approximate mathematical relationships between the pixel density and pixel size of different cameras.

This summary provides an example of how to apply the template that is published  here. In this theoretical template, the reconciliations between the percentages shown for pixel density and pixel size, work out exactly, only because the number of megapixels on the sensor is exactly the same as the image width in pixels, multiplied by the image height in pixels.  In addition, the image width divided by the image height, gives the same answer as the sensor width divided by the sensor height. In the theoretical template, the approximate area calculation for the size of one pixel is exactly equal to the pixel pitch squared. In addition, the approximate area calculation for the pixel density is exactly equal to the linear pixel density squared.

However, in the practical example that follows, the arithmetical reconciliations demonstrated in the template do not work out exactly because of roundings in the specifications used, and also because of the way the effective number of pixels of the cameras is calculated (that is, the image width multiplied by the image height, does not exactly equal the effective number of pixels published for the cameras). For example, the specifications for the Sony A900 state that it has 24.6 million effective pixels, and that the image size is 6048 pixels x 4032 pixels. But, when you multiply 6048 pixels x 4032 pixels, you obtain 24.386 million pixels, not 24.6 million pixels.

Note: The information below is not designed to provide information about the quality of images or the quality of the cameras, because these are separate issues.

Note: If the (full frame) Sony A900 / A850 had the same pixel density as the (APS-C) Sony NEX-3 / NEX-5, it would have approximately 33 megapixels, and image dimensions of approximately 7045 pixels x 4697 pixels.

Pixel pitch  (in microns)

Refer to the reservations  here  about calculating the "true" width and area of an individual pixel.

Pixel pitch in microns  = width of sensor in millimetres  divided  by  image width in pixels  multiplied by 1000

NEX-3 = 5.096    (23.4 / 4592  x 1000)

A900 =   5.936    (35.9 / 6048 x 1000)

Relationship: A900 is approximately 16.5% greater than NEX-3

Crop an image from A900 to the same   field of view  as an image from NEX-3, and compare the changed field of view of A900 with that of NEX-3: Assume that a 70mm lens is on both cameras and that the field of view of an uncropped A900 image is 70mm

Field of view of NEX-3 = focal length of lens  x  crop factor of NEX-3 = approx. 107mm  (70mm x 35.9mm / 23.4mm)

Changed field of view of an A900 image when it is cropped to the same field of view as a NEX-3 image

= uncropped image width of A900  /  cropped image width of A900  x  focal length of lens  =  approx. 107mm  (6048 / 3942  x  70mm)

Relationship: The fields of view of NEX-3 and A900 are the same, that is, approx. 107mm.

Note: The image width of an A900 image, when it is cropped to the same field of view as a NEX-3 image, is approx. 3942 pixels (6048 x 23.4 / 35.9). Click  here  to go to an article titled "Advantages and disadvantages of cropping images to gain extra reach". This article gives further details in support of the formulas used above.