BS ISO 19093:2018 pdf free
BS ISO 19093:2018 pdf free.Photography – Digital cameras – Measuring low-light performance
A way to hide the noise in the image or to prevent blur due to camera shake is to intentionally underexpose the images. That way, the bright parts will be visible but high noise levels can be hidden in the dark areas. Underexposed images look too dark and are therefore undesirable.
The 18 % grey background of the test chart is typically reproduced as an L*50 value in the output image colour encoding. In case of an 8-bit sRGB image (see IEC 61966-2-1) the aim value is a digital output value of 118.
Due to potential shading of the optical system, the grey value shall be measured at the 18 % grey area that is closest to the centre of the image.
If the raw signal is amplified without any noise reduction, then the result will be a noisy image. Below a certain light level, the noise will obscure the signal in a way that renders the image unacceptable.
For the evaluation of noise, the visual noise shall be measured according to ISO 15739 as a mean value over the grey scale in the test chart excluding patches with L* values above 80 and below 20 (densities above 1,5 and below 0,25). The applied contrast sensitivity function shall be the one for viewing the images with pixel density 150 um per pixel from a 500 mm distance. At this resolution and distance,the pixels subtend about 1′ (1/60°), which is approximately the limit of the human visual system. (This viewing condition is also equivalent to a 100 % viewing of images on a 4k 76 cm (30″ class) display from a 500 mm distance). The actual image size in this case depends on the pixel count of the image. Should a manufacturer decide to down sample an image (e.g. in low-light mode of a camera) to average noise over multiple pixels, this would result in a lower pixel count. Therefore it is necessary to report the pixel count together with the result.BS ISO 19093 pdf download.
