![]() ![]() ![]() The computation of luma is indeed closely related to the human visual system sensitivity as you can see from the weighting factors used: Similarly the U and V components are non-linear as they are formed from already non linear R'G'B' components. In the Y'UV colourspace, (notice that I put a prime on Y), Y' doesn't represent relative Luminance but luma, which is a non linear component: it is already encoded through mean of an OETF (i.e. For example, HDTV broadcasting is usually using BT.709 (gamma ☐.5) as the encoding function for the source data (like filming with an HDTV camera) and is decoded by your TV using BT.1886 (gamma 2.4) which yields an end-to-end system gamma of ☑.2, suitable for dim viewing conditions. Gamma correction purpose is not to bring relative Luminance down but to encode values in a perceptual uniform way which will optimise code allocation and thus reduce quantization artefacts compared to a linear encoding of those same values.Īs a matter of fact most opto-electronic transfer function (OETF / OECF) and electro-optical transfer function (EOTF / EOCF) such as sRGB, BT.709, BT.1886 which are power function based (gamma) don't affect the black and white levels.Ī combination of different exponents at the display chain / system input and output will result in an end-to-end system gamma that might increase or decrease visual contrast depending the suited result (this is called Picture Rendering). ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |