White Balance

From RawPedia
Jump to navigation Jump to search
White Balance


Digital images generally consist of a mixture of the three primary colors: red, green and blue. For various reasons which you can read about in-depth elsewhere, the red, green and blue values which serve as the starting point in any raw photo development program need to be corrected in various ways before they resemble the photographed scene. One of these corrections is performed by adjusting the white balance - ensuring that neutral-colored (white) objects in the photographed scene still appear neutral on the photograph. Adjusting the white balance affects all colors, though it is easiest to discern whether the white balance is correct if an object you know to be of a neutral (white, gray) color looks non-neutral.

White balancing works by multiplying each of the primary colors by a different amount, until a satisfactory result is reached. In order to make this operation more human-friendly, instead of operating on the three multipliers directly, the user is presented with an abstraction in the form of a temperature slider which adjusts colors along a blue-yellow axis, and a tint slider which adjusts them along the magenta-green axis.

A neutral color is one whose red, green and blue values are equal. For example, R=G=B=65% and R=G=B=90% are both neutral, the former being darker than the latter. You can tell whether the white balance of a spot which should be neutral is correct by checking whether that spot's RGB values match, or whether the a* and b* values in the L*a*b* color space match, or whether the RGB indicator bars under the main histogram are directly over each other. You can do this even if you have a very miscalibrated monitor. Your perception of color changes depending on the color of the surroundings and of the illumination in your room, so don't always trust your eyes - verify using the method described above.

Having an incorrect white balance results in the image having a color tint, typically warmer (orange) or colder (blue). Some people use this for creative effect, however there are various tools and operations which rely on the assumption that the white balance of the image is correct (for example highlight recovery in the Exposure tool, skin targeting in the Contrast by Detail Levels tool, sky targeting in the Wavelets tool, the CIECAM02 tool), so you should not misuse the white balance tool to create a color cast for artistic effect but rather use it to ensure that neutral areas remain neutral, and then use Color Toning or any of the other tools to render a creative color tint.

The white balance tool can be turned on/off. When off, the multipliers are set to R=1 G=1 B=1 when working with raw files. This can be useful for diagnostic purposes or when working with UniWB images.

Interface Description


  • Wb-camera.png Camera
    Takes the white balance used by the camera. If you shoot only in raw (so no raw+JPG), put the white balance settings of your camera on auto. This should generally give good results.
  • Wb-auto.png Automatic
    • RGB grey
      Automatically corrects the white balance, by assuming that the average color of the scene is neutral gray. Works well for a wide range of scenes, and can be a good starting point for manual adjustments.
    • Temperature correlation
      Provides a generally better color balance than auto “RGB grey". The algorithm is based on the best correlation (Student's test) between the colors of the image and an array of 200 spectral reference colors.
      • This algorithm may give erroneous results:
        • If the illuminant does not have a CRI (Color Rendering Index) close to 100, e.g. "Underwater", "Fluorescent", "Led" lighting conditions may give bad results.
        • Some DNG-type files obtained after conversion with a DNG or other converter.
        • If the shooting conditions are extreme (very low luminance values, etc.).
      • The GUI displays the correlation value:
        • A value of 1000 means either that calculation is not performed again and that the previous results are used, or that the algorithm has failed to compute a result in which case T=5002 is displayed.
        • Values less than 0.01 are good.
      • You can use "Awb temperature bias" to adjust the results. Each movement of this command brings a new calculation of temperature, tint and correlation.
      • A description of the Itcwb algorithim can be found here algorithm
  • Wb-custom.png Custom
Set your own color temperature and green tint by moving the two sliders and/or using the Spot WB tool.
  • Light source presets
    • Wb-sun.png Daylight (Sunny)
    • Wb-cloudy.png Cloudy
    • Wb-shade.png Shade
    • Wb-water.png Underwater
    • Wb-tungsten.png Tungsten
    • Wb-fluorescent.png Fluorescent
    • Wb-lamp.png Lamp
    • Wb-led.png LED
    • Wb-flash.png Flash


This tank wagon was the whitest object in the scene. We can tell that the white balance is wrong because the RGB levels are not equal, and the RGB indicator bars directly under the histogram are spread apart.
Picking the white balance off the side of this tanker, which we know to have been the whitest object in the scene, adjusts the colors across the whole image such that the RGB levels in that spot are now equal.

When you click on the Pick button Color-picker.png (shortcut: w), the cursor changes into a pipette when it's over the preview. Click on a neutral area to set the correct white balance for the whole image based on the clicked area.

Pick a spot which should have a neutral tone - gray or white. This spot should not be clipped in any of the three channels, as clipping means that information from the clipped channel is missing. As far as white balancing is concerned, "white" does not mean R=100% G=100% B=100% as that would be clipped, but instead means a shade of gray - even a very light one, but still one without any clipping. The picked spot should also not be black, as black means that insufficient data was captured for that area, and so a correct white balance calculation cannot be performed.

You can use the picker multiple times on different places in the photo until you find an ideal spot. Use the Size drop-down box to change the size of the pipette.

This tool can be used as well inside a detail window. Right-click to cancel the tool and to get the regular cursor back.

Temperature and Tint

The temperature slider adjusts colors along the blue-yellow axis. Moving it to the left makes the image cooler (bluish); moving it to the right makes it warmer (yellowish).

The tint slider adjusts colors along the magenta-green axis. Moving it to the left makes the image more magenta; moving it to the right - more green.

Blue/Red Equalizer

The red/blue equalizer allows to deviate from the normal behavior of "white balance", via increase or decrease of the ratio between red and blue. This can be useful when shooting conditions are far from the standard illuminant, e.g. underwater, or are far from conditions where calibrations were performed, for which the color matrices in the input profile are unsuitable.

AWB Temperature Bias

The auto white balance temperature bias slider allows you to specify how much the automatically-calculated temperature should deviate. Use this if you would like the automatically-calculated white balance to be cooler or warmer.

White Balance Connection to Exposure

The white balance is described in temperature and tint, but when working with raw images it will be translated into weights of the red, green and blue channels. The weights will be adjusted so that the channel with the smallest weight reaches clipping in the working space (usually ProPhoto RGB) when the raw channel is clipped. In other words, with exposure set to 0.0 and no highlight recovery enabled the full visible range is fully defined by the raw backing. As white-balancing changes the weights you may see a slight exposure change if you make drastic changes to white balance.

The Temperature Correlation Algorithm

This section is a technical description of the temperature correlation algorithm and its implementation. It is not necessary to know this if you just want to use the "Automatic > Temperature Correlation" white balance method, but will be of interest to those studying the matter.

This algorithm is referred to in abbreviation as "ITCWB".

Unlike the majority of white balance algorithms based on gray tones, this one is based on color. Put simply, the algorithm compares a large number of sample colors in the image with a set of reference colors and their associated spectral data.


This algorithm was developed by Jacques Desmis. It was based off an unpublished research summary, which divides the process up into 3 phases:

  1. xyY comparison
  2. Spectral data analysis
  3. Color histogram analysis

These phases form the basis of the algorithm described below, which was developed from scratch and is not based on any existing algorithms or code.


The performance of the algorithm depends on:

  • The choice of colors in the image obtained by sampling and selecting the dominant colors (skin, sky, plants etc.).
  • The determination of certain parameters, which will be used as the basis for the calculations i.e. camera white-balance temperature, which acts on the red and blue components and tint, which acts on the magenta and green components, etc.
  • The choice of the RGB channel multipliers and their calculation based on the temperature of the illuminant.
  • The calculation of the XY values of the reference colors (spectral data), using an "exact" formula and samples of spectral data at 5nm. Matrix [Color seen] = Matrix [illuminant] * Matrix [color] / Matrix [Observer 2°].
  • Multiple iteration of the calculations taking into account, in equal proportions, the balance between green-magenta and red-blue.
  • Rigorous calculations if the illuminant has a CRI (Color Rendering Index) close to 100 i.e. illuminant close to Daylight in the limit 4000K - 15000K or Blackbody from 2000K to 4000K.
  • Statistical correlation using a Student's test.

Reference Spectral Colors

The origin and nature of the 429 reference spectral colors:

  • Data found on the web for flowers, foliage.
  • A ColorChecker24 or other color patches.
  • The 468 calibration chart that I developed for calibration a few years ago.
  • The Colorlab utility (Logo Gmbh).
  • These colors are distributed almost equally over the entire color palette (Red, Orange, Yellow, Green, Cyan, Blue, Magenta…).
  • These colors are also sorted into neutral or close to gray, slightly saturated, pastel and saturated.
  • The luminance has little significance because the comparison is made on the chroma component.

General Principles

  • Using the RGB values just after demosaicing, 3 tables are generated (Red, Green, & Blue) for 1 pixel out of every 3 in the image (horizontally and vertically). It is possible to change this value if necessary for more precision. The values are then adjusted so that they are in the range 0 to 65535.
  • Then we switch to a procedure called "autowb", which is common to both automatic white balance algorithms. It calculates the RGB channel multipliers, and passes on the values to either to "Itcwb" or "rgbgray".
  • The parameters that "wbauto" passes on to "Itcwb" include the important reference temperature (the value present in the Exif camera data) and the tint (also present in the Exif data), whose values are limited to the range between 0.77 and 1.30. There is no Daylight or Blackbody illuminant beyond these arbitrary limits and any calculations would therefore be fanciful or false.

Simplified Temperature Correlation Algorithm

  1. Phase one
    1. Calculate the RGB multipliers for each temperature between 2000K and 15000K (blackbody and daylight illuminant's) and for the tint.
    2. Calculate the XY values from the 429 spectral-data values for each temperature.
    3. Select a temperature data range relative to the reference.
    4. Calculate the xy values in the form of a histogram and select from among the 236 possible values, the most commonly used colors (skin, sky, etc.) for each temperature.
    5. Sort the data in ascending numerical order.
    6. For the most frequently occurring data values, calculate the chromatic values of the image.
    7. Use the deltaE chroma values to select the reference colors from the 429 available possibilities.
    8. Calculate the reference RGB values as a function of the reference temperature.
  2. Phase two
    1. Calculate the XY values for each selected reference color as a function of temperature and tint.
    2. Calculate the RGB values of the image from the XY values using the RGB multipliers.
    3. First calculation of the Student correlation.
    4. For each tint and temperature range, calculate the channel multipliers and the XY values from the corresponding spectral data.
    5. Calculate the correlation coefficient as a function of the color green.
    6. Sort these values.
    7. Optimize the values to determine the correct temperature and tint values.
    8. Send these parameters to "wbauto”.
    9. Display the results and update Improccoordinator.cc.

Latest Improvements 05 2023

  • By default, because it is necessary to have a starting reference for the algorithm, the parameters chosen are those of "Camera" (temperature and tint). It turns out that in a few cases, these values are obviously wrong. In case "camera" tint is higher than 1.5 or "Camera" temperature is lower than 3300K or "Camera" temperature is higher than 7700K, the new starting reference is the one calculated with "Automatic RGB grey"(or a mix "Camera" and "Automatic RGB grey").
  • When "Camera" temperature is lower than 4000K or "Camera" temperature is higher than 6000K, a 2 pass process of the algorithm is set up, looking if another value far enough away gives better results. When "Camera" temperature is higher than 4000K or "Camera" temperature is lower than 6000K, another 2-pass process is implemented with smaller deviations from "Camera".
  • If you activate "Low sampling & No use Camera settings", upstream of the algorithm and in the algorithm, the system will use a mix of the "Camera" and "Auto WB grey" settings
  • the calculation of green (hue) has been revised;
  • an attempt to optimize the patch is made from the chromatic analysis of the image (deviation from the white point). The lower the value, the more credible the patch is in theory.
  • before developing the histogram of the image data, a slight denoising (median 3x3) is applied.
  • a gamut control is also applied when calculating the histogram data, eliminating outliers.

Data displayed in the GUI - limitations of interpretation

  • Multipliers r, g, b. These are given for information only and cannot be modified;
  • Correlation factor: gives a probability of correlation between the image data and the spectral data. This coefficient is indicative, because it assumes that the patch is "good", but this patch is relatively indeterminate: what data to take? On what amplitude? Certainly it will translate, for a choice of temperature / start-up shade, the best compromise, but is the choice of start-up optimum?
  • Passes ( 1 or 2) and Alt_temp: displays the number of passes of the algorithm carried out as well as the possible alternative temperature. In the "2 passes" case, the "Remove 2 passes algorithm" check box is active, otherwise it is inactive.
  • Read colors: number of colors read in the image (depends on "Sampling"), the maximum is 237 which covers the entire CIExy diagram.
  • Chroma patch: indicator that translates the attempt to optimize the patch. A weighting according to an exponential law, affects the values to try to take into account "at best" the flat areas (sky, skin, etc.) and the more isolated data.
  • Size: number of colors chosen at once in the image (from the highest in number of corresponding pixels, to the lowest).
  • patch ΔE: shows the deltaE of the patch between the image data and the spectral data.
  • datas x9: displays the number of records found for each color. Maximum and minimum. The absolute minimum is (arbitrarily) set at 400. To have a real evaluation, multiply these values by 9, because only 1 pixel out of 3 (horizontal and vertical) is analyzed to minimize the processing time.
Interpretation limits

The Itcwb algorithm is complex, but devoid of intelligence. There is no interpretation of the image: is it a portrait? a landscape ? what is the nature and distribution of the data? So it is necessary to be very careful on the "steering" of the system by these indicators. Certainly they seem robust in a majority of cases but there are many exceptions:

  • just because "chroma patch" is lower doesn't mean the patch is necessarily better. This indicator shows for a "temperature / tint" couple a possible optimization, but is it the best? It will allow you to choose the size of the patch.
  • it is not because "deltaE" is the lowest, that this choice is the most optimal. Admittedly, it translates for a given temperature/tint amplitude around a reference value, the optimization of the deviations, but that does not demonstrate that it is the best choice.
  • it is not because "Correlation factor" is the lowest that the result is optimal. This correlation translates the optimization of a "hue" and a temperature around a reference value, but this does not demonstrate that it is the best choice.

The system first determines (before approaching the algorithm itself) which base references to use (Camera or auto WB grey). Then he determines if he will use 1 or 2 passes. In the first step, for each pass (or only one) the optimization of the patch is done with "patch chroma". Then the algorithm tests a range of temperature and tint and calculates deltaE of the patch and correlation. A compromise is found by taking into account the minimum of the multiplication of the detaE and the correlation. On the other hand, when we are far from D65 (value taken into account by Adobe) to determine the color matrix, the probability that the data read and interpreted are incorrect increases and the results displayed are probably flawed. There is currently no possibility in Rawtherapee to read 2 color matrices. Even if this possibility existed, the work of calculating and entering the stdA color matrices would be considerable.

There will be images that will (obviously) not be processed well. For example, images with one or two strong colors (red, yellow, purple..) and virtually no white (gray) will be poorly optimized. For example the flowers on a foliage background. Similarly illuminants far from Daylight (LED) or illuminant mixtures will lead to incorrect results.

Using AWB Temperature bias

AWB temperature bias is a simple way to get a quick result.

Indeed, acting on this slider completely reexecutes the algorithm by shifting the initialization temperature. All parameters are re-calculated temperature, tint, correlation.

  • it can be to readjust the colorimetry - a bit like CIECAM chromatic adaptation, but it's not exactly the same thing.
  • or to visually obtain an image more in line with your expectations, without color shifts.
  • it has a role similar for "temperature" to "Green refinement".

User-Modifiable Settings

The development branch whitebalanceopt allows one to modify the parameters used by this algorithm.

By default the settings should be suitable in most cases. However, it is possible to make custom modifications to the operation of the algorithm.

In the Color / White Balance tab you can make a series of settings appear that allow you to adapt the algorithm. Eventually (I hope) to be able to remove the majority of these settings (perhaps all). The purpose of this provision of these settings (optional) is essentially the development of the algorithm.

To make these additional settings appear, go to Preferences / Color Management / White balance - Automatic temperature correlation, and check the corresponding box. If "Temperature correlation" is not selected in "White Balance", the choices appear in gray.

Relevant Parameters

I think I can give a relevant opinion, being the designer of the algorithm. Nevertheless, I leave the 'door open' to other hypotheses because the colorimetry in general and the adaptation to Rawtherapee can reserve hypotheses.

Parameters having an influence (a priori):

  • The 3x3 color matrix
    • The 3x3 matrix ensures the conversion of raw Raw data, into useful data. In Rawtherapee, these indispensable matrices which are used in several algorithms of the Raw part have their origin in Adobe, either they come from Dcraw, or they come from Adobe DNG converter (Tag matrix). How were these matrices obtained by Adobe ? Internal research, link with the manufacturer,... ? One can think that these matrices are built in a common process, and thus that all things being equal, the same scene, under the same illuminant, taken under the same conditions (speed, diaphragm, lens,...), with a common Raw process, must bring usable color data (and not luminance or Dynamic Range) Raw approximately identical (with the possible reserves due to the parameters developed below). That it is a matrix of Bayer, or other, a case of mark X or Y. Obviously there are differences, in particular for the non-Bayer matrices which have a specific demosaicing algorithm, as well as for the optics... but the differences must be minimal.
    • This matrix is for illuminant D65. In the exifs there is a second matrix for the illuminant stdA (tungsten). A test carried out on a raw at 2600K, leads me to think that this choice would be (a little) better. However Rawtherapee only knows how to manage a color matrix. On the other hand how would one make the choice? And how to update the hundreds of cameras?
    • This matrix is "the factor" that identifies make, model. It allows (unless this matrix is poorly calculated, a rare case…) to say that the brand and the model are taken into account and that they have no influence on the algorithm. This does not mean that the brand, the model are not important, obviously they are in the gamut sense. But it is taken into account by the algorithm if the spectral data are sufficient.
  • The number of spectral colors: it seems obvious that the more the perimeter of the image data is increased (reading the xyY data in the CIExy diagram), the newer the devices will be with a very wide gamut, the more spectral data will be needed matches – otherwise the algorithm malfunctions. The number of spectral color data has increased from 201 until the beginning of 2023, to 429 currently. Maybe more data is needed?
  • The demosaic algorithm also has a influence, especially the algorithms designed to handle noisy images (LMSSE, IGV), and of course the demosaic algorithm for non-Bayer, for example Xtrans-demosaic (Frank Markesteijn's algorithm, and Ingo Weirich). Each demosaicing algorithm has its particularities (this is not the place to discuss it here), but brings its own colorimetry. Between several methods (Amaze, DCB, VNG4,...) there are differences which are quantified in delatE, and induce different colors. This will affect Itcwb. Similarly, Raw processing performed before Itcwb also has a certain impact, for example Capture Sharpening and chromatic aberration correction. But this is not a malfunction, on the contrary it shows that the algorithm works and takes into account the color changes induced by these processes.
  • Each sensor - associated with a housing - has :
    • a specific DR (Dynamic range) - often around 12Ev for older cameras, close to 15 or 16Ev for recent cameras. This DR must have a low impact on the algorithm, the luminance component is almost ignored (otherwise to determine the gamut).
    • White-levels and Black-levels : if these 2 components have a strong importance for the whole of the Raw processing, except if they are very badly adjusted, they should not have an influence on the algorithm.
    • the gamut of the sensor : I do not know of official documents showing the limits of transcription of the colors. It is reasonable to think that before conversion matrix, these limits are largely sufficient and beyond Prophoto, and thus have no or little influence on the algorithm (not on the result of course).
  • The nature and intensity of colors: the distribution in the xyY diagram. Are we with an image where the colors are close to the white point (pastel colors or neutral), or images with colors at the limits of those of human perception?
  • The distribution of these colors in relation to each of the primary Red / Green / Blue.
  • As the algorithm divides the 'xy' space into 236 areas, covering the entire visible space, the algorithm must take into account this distribution: for example a majority of tones very close to the white point (neutral) or a dominant important (sky, or skin).
  • And of course the illuminants. An image with parts in the sun and shade is actually with 2 illuminants Daylights (near 5000K in the sun and 7000K in the shade). This is almost insoluble with a single setting of the white balance (temperature, hue). Of course it is even more complex if Fluorescent or LED illuminants are present. But these remarks are not specific to Itcwb, but to all white balance algorithms. Local adjustments (Warm/cool) allows you to correct double illuminants (sun, shadow...) quite well.
  • Still on the subject of illuminants, they have a theoretical definition that links spectral data to a temperature. These formulas by principle can not be perfect and respond to all environments: latitude, altitude, time, meteorological conditions (fog, gradients, ...) that must have an impact on the shade (green) that becomes different from 1.
  • Observer 2° (1931) or Observer 10° (1964): the second provides a better perception of human vision.
  • As a reminder, a perceived color with its XYZ data is the combination of 3 matrices:
    • spectral data of the illuminant (function of temperature and nature of the illuminant).
    • spectral data of the base color (measured with the spectro).
    • spectral data of the observer (2° or 10°).

By principle this algorithm is not designed to correct the malfunctions of the processing (which is always very complex). Of course it can (possibly) correct a problem, but this is not its purpose.


  • CIExy diagram and gamut: You can see on the 2 diagrams below, that it is not because a color is in the CIExy diagram that it is in the gamut.
    • Example for a luminance of 10 and a luminance of 50 [0..100].
Gamut comparison for L=10 : yellow=ACESp0 gray=Prophoto green=sRGB
Gamut comparison for L=50 : yellow=ACESp0 gray=Picture green=sRGB
Pointer's gamut and Rec2020
Gamut comparison for L=17 : pink=JDCmax gray=Prophoto yellow=Beta RGB green=sRGB

Impact settings: After several weeks of user testing and optimization of the algorithm, 3 parameters remain directly accessible to the user:

  • Green refinement : Allows you to change the "tint" (green) which will serve as a reference when starting the algorithm. It has substantially the same role for greens as "AWB temperature bias" for temperature. The whole algorithm is recalculated. Depending on the case, in order to guide the algorithm, the action on "Remove 2 pass algorithm" may be necessary.
  • Remove 2 passes algorithm: this checkbox, manages a complex background process, which leads to that in cases where there is only one pass chosen by the algorithm, the action on this box to check has no effect. In the case where 2 passes are detected, this checkbox allows switching from one setting to another. Be careful, when one of the 2 settings comes out towards temperatures close to stdA (Tunsgtene 2855K), the indicators (correlation deltaE..) must be taken with care, the results are probably marred by errors linked to the color matrix.
  • Choice of image data sampling: 2 choices are offered:
    • Low sampling – limits data to sRGB values. In some conditions (green camera > 0.8), forces the algorithm in some cases not to use camera settings.
    • Medium sampling (default)- near Pointers's gamut. Beta RGB primaries are used, thus obtaining a color range close to human vision in reflected color (Subtractive color mixing). I could also have chosen Rec2020 which is a bit "bigger" than "Beta RGB".
    • Camera XYZ matrix - uses the matrix directly derived from Color matrix (Adobe)
    • Close to full CIE diagram – the limit is that of JDCmax, i.e. close to entire CIE diagram.
    • These 3 samplings have no relation to the 'working profile' and have no effect on the rest of the processing, except on the balance of the white balance.
    • I recommend, insofar as there is enough spectral data to choose this last choice 'Close to full CIE diagram', even if it generates imaginary colors (most in greens), because it is much in the continuity of the process before the application of the "working profile". The exceptions can be of 2 orders: a) an image contains data which is not referenced in spectral values (must be very rare) and in this case the deltaE and correlations will be biased; b) you voluntarily wish not to take into account parts of the image with high gamut, which could disturb the result or no use of camera settings.
    • Note that when there is no "Input profile" (or when the user chooses "Camera standard") the reference data used for sampling are processed in such a way as to be similar to those used later.

Settings accessible from pp3:

  • Itcwb_findgreen - Find green student: number of iterations to find the best compromise between the correlation (student) and the value of green which for Daylight / Blackbody illuminants is close to 1. Default: 3. Range of settings taken into account [2..5]. It seems that the value 3 is a good compromise, which would make it possible to remove this setting.
  • No purple color used: by default this setting is not taken into account. If the image needs highlight reconstruction it may need to be enabled.
  • itcwb_minsize: set by default to 20. Sets the minimum value for the patch size.
  • Itcwb_rgreen - Geen range: sets the amplitude of examination of the value of green in the iterations, from a low amplitude of 0.82 to 1.25 up to a maximum of amplitude 0.4 to 4. Default: 1. Range of settings taken into account [0..3]
  • Itcwb_delta - Delta temperature in green loop: Fixed for each "green" iteration tried, the temperature difference to be taken into account. Default: 4. Range of settings taken into account [1..8]

Settings accessible from the 'options' file

  • Itcwb_deltaspec: if Verbose is "true", displays the results (in the console) of the image and spectral data whose deviation in deltaE is greater than this value. Default 0.075.
  • Itcwb_maxsize: sets the maximum patch size. By default the value is 70.

Chromatic Adaptation

The results of the Itcwb algorithm, reflect the relevance of calculations on objective mathematical foundations. but this result - as indeed all settings of the white balance - do not take into account in full human perception: surround, simultaneous contrast, ... and especially the adaptation of our eye / brain to temperature differences from D50 (which is the reference in colorimetry). To overcome this gap, it is possible to set up a chromatic adaptation "integrated" with the white balance (this is what I had proposed in 2018...) or let the user set up this adaptation with the module "Color Appearance & Lighting".

To achieve this and limit the role of Ciecam, put Ciecam in "Automatic symmetric" mode, the system will apply 2 chromatic adaptations, the first from "Scene conditions" to the reference illuminant (usually D50, but you can change it), the second from the reference illuminant to "Viewing conditions". By default the 2 adaptation percentages are set to 90%, you can increase or decrease these values. You can also modify the temperature in "Viewing conditions" to obtain a warmer or colder rendering. You can, if you wish, change the Ciecam settings like "Absolute luminance", "Surround", etc. See the tutorial on Ciecam.


I have (arbitrarily) chosen these 6 examples, to show what Itcwb can (and cannot) do, associated or not with Color Appearance & Lighting.

  • Salt mountain in Turkey (_ASC4145.NEF CC BY-SA 4.0 Jacques Desmis)
    This image that seems harmless is complex in terms of photography, for several reasons:
    • the white of the salt mountain is difficult to process, and is affected by a complex structure that makes it difficult to process in general.
    • The point that interests us here is the white balance and color distribution. The majority of the image sky, trees, mountain is in sRGB, but the flowers at the bottom of the image (red, orange, yellow) are well beyond: how to treat, incidences of settings
  • Lunching Room (LunchingRoom.CR2 CC BY-SA 4.0 Rawtherapee)
    This image shows that the algorithm can handle complex situations.
    • By default with White Balance set to "Camera" the image is green.
    • Try successively, White Balance auto : 'Rgb grey' and "Temperature correlation".
  • London Bridge (london_bridge_moving_1.pef CC BY-SA 4.0 Maciej Dworak)
    This image shows both the need for chromatic adaptation and the relevance of the Itcwb algorithm.
    • The default setting 'Camera' gives a green/yellowish cast.
    • Itcwb allows to find a good mathematical compromise, but the temperature is high, giving the image a warm coloring, which can be seen on the faces, the deck stays.
    • Try Color Appearance in "Automatic symmetric" mode
  • Calibration test pattern (DSCF5334.RAF CC BY-SA 4.0 RawTherapee)
    • I chose this test pattern for two reasons: the first is that it is a RAF file, so it is not a Bayer file; the second is that it contains almost pure whites and blacks (the grays are dominant in the image, which could interfere with the algorithm).
    • Nevertheless it does not seem to examine the results (neutral) that the camera has been calibrated (or I do not have the profile), the information is therefore orders of magnitude, but quite close to reality (whites, blacks, ColorChecker)
    • Itcwb doesn't know it's a calibration test pattern, try because of the high gamut 'Force use of the entire CIE diagram' on or off, 'Sort in chroma order' off or on and 'no purple used' off - the results are very close to those with Camera (which are better?), but there is no drift.
  • Caribbean Backlight (DSC02973.ARW CC BY-SA 4.0 Jacques Desmis)
    • I chose this image, taken with my old Sony during a trip to the Caribbean, where I had chosen the automatic white balance.
    • Try with 'Camera' then 'Itcwb'.
  • Using Inpaint-opposed (Nikon - D800 - 14bit uncompressed (3_2).NEF CC0 1.0 Pascal Obry)
    • I chose this image to show the impact of the white balance on "Inpaint opposed" (highlight reconstruction), especially the green (tint) influence
    • Try "Inpaint opposed" with White Balance set to Camera, notice the large artifacts in the sky
    • Choose "Itcwb" and try various settings.

5.9 Compatibility

The Temperature Correlation Algorithm (Itcwb) versions after 5.9 have limited compatibility with version 5.9:

  • Only the "Observer 10° instead of Observer 2°" setting is available.

When the system encounters a pp3 file of version 5.9 using Itcwb, then the "Temperature correlation" method is displayed. The only relevant information is:

  • white balance multipliers;
  • temperature and tint;
  • the "Observer 10° instead of Observer 2°" check box.

You can switch to the current Temperature Correlation Algorithm (Itcwb) method by changing the "White balance" method, for example by going through "Custom", then "Temperature correlation"