spacesRGB (version 1.7-0)

Linear RGB, XYZ, and Lab Calculation: Convert Signal RGB coordinates to Linear RGB, XYZ, or Lab

Description

Convert signal RGB coordinates to linear RGB, XYZ, or Lab

Usage

LinearRGBfromSignalRGB( RGB, space='sRGB', which='scene', TF=NULL, maxSignal=1 ) 
XYZfromRGB( RGB, space='sRGB', which='scene', TF=NULL, maxSignal=1 ) 
LabfromRGB( RGB, space='sRGB', which='scene', TF=NULL, maxSignal=1 )

Value

XYZfromRGB() returns a data.frame with N rows and these columns:

OutOfGamut

a logical vector. TRUE means the input signal RGB is outside the cube [0,maxSignal]\(^3\). If TF is not identity.TF, then the input signal RGBs are clamped to this cube before further calculations.

XYZ

the calculated XYZ vectors, using the appropriate 3x3 matrix for the RGB space. These are for viewing under the white point (either 'scene' or 'display') of the given RGB space.

LabfromRGB() returns a data.frame with N rows and these columns:

OutOfGamut

a logical vector. TRUE means the input signal RGB is outside the cube [0,maxSignal]\(^3\). If TF is not identity.TF, then the input signal RGBs are clamped to this cube before further calculations.

Lab

the calculated Lab vectors. First XYZ is calculated using XYZfromRGB(). Then XYZ → Lab is calculated using spacesXYZ::LabfromXYZ(), with the white point (either 'scene' or 'display') of the given RGB space. If R=G=B exactly, then a=b=0 exactly.

LinearRGBfromSignalRGB() returns a data.frame with N rows and these columns:

OutOfGamut

a logical vector. TRUE means the input signal RGB is outside the cube [0,maxSignal]\(^3\). If TF is not identity.TF, then the signal RGBs are clamped to this cube before linearizing.

RGB

the calculated linear RGB vectors, either scene linear or display linear, as specified by which.

In case of error, all functions return NULL.

Arguments

RGB

a numeric Nx3 matrix with non-linear signal RGB coordinates in the rows, or a vector that can be converted to such a matrix, by row. These should be in the appropriate cube [0,maxSignal]\(^3\); but see OutOfGamut in Value below.

space

the name of an installed RGB space. The name matching is partial and case-insensitive.

which

either 'scene' or 'display'. For LinearRGBfromSignalRGB(), which describes the output RGB. For XYZfromRGB(), which describes the output XYZ. For LabfromRGB(), which describes the output Lab. Usually the OOTF for the space is the identity and so 'scene' and 'display' are the same.

TF

if not NULL, TF is a TransferFunction that overrides the appropriate transfer function of space.
TF can also be a positive number. If TF=1, then TF is set to identity.TF, so the returned RGB values are not clamped (see Value). If TF!=1 it is used to create either power.EOTF() or power.OETF() as approriate. If TF is not NULL in LinearRGBfromSignalRGB(), then space is ignored.

maxSignal

maximum value of the input signal RGB. It is just a simple scaling factor; other common values are 100, 255, 1023, 4095, and 65535. Even when maxSignal=1, the RGBs are still taken to be non-linear signal values.

Details

In XYZfromRGB(), the conversion is done in 2 steps:

• signal RGB → linear RGB using LinearRGBfromSignalRGB() and all other the given arguments

• linear RGB → XYZ using the 3x3 matrix for the given space and which

References

Wikipedia. RGB color space. https://en.wikipedia.org/wiki/RGB_color_space

See Also

RGBfromXYZ(), SignalRGBfromLinearRGB(), installRGB(), spacesXYZ::LabfromXYZ()

Examples

XYZfromRGB( c(128,200,255, 0,0,0,  255,255,255), max=255 )$XYZ * 100
##              X         Y        Z
##  [1,] 47.60334  53.11601 102.3549
##  [2,]  0.00000   0.00000   0.0000
##  [3,] 95.04559 100.00000 108.9058


XYZfromRGB( c(128,200,255, 0,0,0,  255,255,255), space='Adobe', max=255 )$XYZ   * 100
##              X         Y        Z
##  [1,] 42.36398  50.82876 103.8704
##  [2,]  0.00000   0.00000   0.0000
##  [3,] 95.04559 100.00000 108.9058