COMPOSITION
DESIGN
COLOR
-
Image rendering bit depth
Read more: Image rendering bit depthThe terms 8-bit, 16-bit, 16-bit float, and 32-bit refer to different data formats used to store and represent image information, as bits per pixel.
https://en.wikipedia.org/wiki/Color_depth
In color technology, color depth also known as bit depth, is either the number of bits used to indicate the color of a single pixel, OR the number of bits used for each color component of a single pixel.
When referring to a pixel, the concept can be defined as bits per pixel (bpp).
When referring to a color component, the concept can be defined as bits per component, bits per channel, bits per color (all three abbreviated bpc), and also bits per pixel component, bits per color channel or bits per sample (bps). Modern standards tend to use bits per component, but historical lower-depth systems used bits per pixel more often.
Color depth is only one aspect of color representation, expressing the precision with which the amount of each primary can be expressed; the other aspect is how broad a range of colors can be expressed (the gamut). The definition of both color precision and gamut is accomplished with a color encoding specification which assigns a digital code value to a location in a color space.
Here’s a simple explanation of each.
8-bit images (i.e. 24 bits per pixel for a color image) are considered Low Dynamic Range.
They can store around 5 stops of light and each pixel carry a value from 0 (black) to 255 (white).
As a comparison, DSLR cameras can capture ~12-15 stops of light and they use RAW files to store the information.16-bit: This format is commonly referred to as “half-precision.” It uses 16 bits of data to represent color values for each pixel. With 16 bits, you can have 65,536 discrete levels of color, allowing for relatively high precision and smooth gradients. However, it has a limited dynamic range, meaning it cannot accurately represent extremely bright or dark values. It is commonly used for regular images and textures.
16-bit float: This format is an extension of the 16-bit format but uses floating-point numbers instead of fixed integers. Floating-point numbers allow for more precise calculations and a larger dynamic range. In this case, the 16 bits are used to store both the color value and the exponent, which controls the range of values that can be represented. The 16-bit float format provides better accuracy and a wider dynamic range than regular 16-bit, making it useful for high-dynamic-range imaging (HDRI) and computations that require more precision.
32-bit: (i.e. 96 bits per pixel for a color image) are considered High Dynamic Range. This format, also known as “full-precision” or “float,” uses 32 bits to represent color values and offers the highest precision and dynamic range among the three options. With 32 bits, you have a significantly larger number of discrete levels, allowing for extremely accurate color representation, smooth gradients, and a wide range of brightness values. It is commonly used for professional rendering, visual effects, and scientific applications where maximum precision is required.
Bits and HDR coverage
High Dynamic Range (HDR) images are designed to capture a wide range of luminance values, from the darkest shadows to the brightest highlights, in order to reproduce a scene with more accuracy and detail. The bit depth of an image refers to the number of bits used to represent each pixel’s color information. When comparing 32-bit float and 16-bit float HDR images, the drop in accuracy primarily relates to the precision of the color information.
A 32-bit float HDR image offers a higher level of precision compared to a 16-bit float HDR image. In a 32-bit float format, each color channel (red, green, and blue) is represented by 32 bits, allowing for a larger range of values to be stored. This increased precision enables the image to retain more details and subtleties in color and luminance.
On the other hand, a 16-bit float HDR image utilizes 16 bits per color channel, resulting in a reduced range of values that can be represented. This lower precision leads to a loss of fine details and color nuances, especially in highly contrasted areas of the image where there are significant differences in luminance.
The drop in accuracy between 32-bit and 16-bit float HDR images becomes more noticeable as the exposure range of the scene increases. Exposure range refers to the span between the darkest and brightest areas of an image. In scenes with a limited exposure range, where the luminance differences are relatively small, the loss of accuracy may not be as prominent or perceptible. These images usually are around 8-10 exposure levels.
However, in scenes with a wide exposure range, such as a landscape with deep shadows and bright highlights, the reduced precision of a 16-bit float HDR image can result in visible artifacts like color banding, posterization, and loss of detail in both shadows and highlights. The image may exhibit abrupt transitions between tones or colors, which can appear unnatural and less realistic.
To provide a rough estimate, it is often observed that exposure values beyond approximately ±6 to ±8 stops from the middle gray (18% reflectance) may be more prone to accuracy issues in a 16-bit float format. This range may vary depending on the specific implementation and encoding scheme used.
To summarize, the drop in accuracy between 32-bit and 16-bit float HDR images is mainly related to the reduced precision of color information. This decrease in precision becomes more apparent in scenes with a wide exposure range, affecting the representation of fine details and leading to visible artifacts in the image.
In practice, this means that exposure values beyond a certain range will experience a loss of accuracy and detail when stored in a 16-bit float format. The exact range at which this loss occurs depends on the encoding scheme and the specific implementation. However, in general, extremely bright or extremely dark values that fall outside the representable range may be subject to quantization errors, resulting in loss of detail, banding, or other artifacts.
HDRs used for lighting purposes are usually slightly convolved to improve on sampling speed and removing specular artefacts. To that extent, 16 bit float HDRIs tend to me most used in CG cycles.
-
Colormaxxing – What if I told you that rgb(255, 0, 0) is not actually the reddest red you can have in your browser?
Read more: Colormaxxing – What if I told you that rgb(255, 0, 0) is not actually the reddest red you can have in your browser?https://karuna.dev/colormaxxing
https://webkit.org/blog-files/color-gamut/comparison.html
https://oklch.com/#70,0.1,197,100
-
What is a Gamut or Color Space and why do I need to know about CIE
Read more: What is a Gamut or Color Space and why do I need to know about CIE
http://www.xdcam-user.com/2014/05/what-is-a-gamut-or-color-space-and-why-do-i-need-to-know-about-it/
In video terms gamut is normally related to as the full range of colours and brightness that can be either captured or displayed.
Generally speaking all color gamuts recommendations are trying to define a reasonable level of color representation based on available technology and hardware. REC-601 represents the old TVs. REC-709 is currently the most distributed solution. P3 is mainly available in movie theaters and is now being adopted in some of the best new 4K HDR TVs. Rec2020 (a wider space than P3 that improves on visibke color representation) and ACES (the full coverage of visible color) are other common standards which see major hardware development these days.
To compare and visualize different solution (across video and printing solutions), most developers use the CIE color model chart as a reference.
The CIE color model is a color space model created by the International Commission on Illumination known as the Commission Internationale de l’Elcairage (CIE) in 1931. It is also known as the CIE XYZ color space or the CIE 1931 XYZ color space.
This chart represents the first defined quantitative link between distributions of wavelengths in the electromagnetic visible spectrum, and physiologically perceived colors in human color vision. Or basically, the range of color a typical human eye can perceive through visible light.
Note that while the human perception is quite wide, and generally speaking biased towards greens (we are apes after all), the amount of colors available through nature, generated through light reflection, tend to be a much smaller section. This is defined by the Pointer’s Chart.
In short. Color gamut is a representation of color coverage, used to describe data stored in images against available hardware and viewer technologies.
Camera color encoding from
https://www.slideshare.net/hpduiker/acescg-a-common-color-encoding-for-visual-effects-applications
CIE 1976
http://bernardsmith.eu/computatrum/scan_and_restore_archive_and_print/scanning/
https://store.yujiintl.com/blogs/high-cri-led/understanding-cie1931-and-cie-1976
The CIE 1931 standard has been replaced by a CIE 1976 standard. Below we can see the significance of this.
People have observed that the biggest issue with CIE 1931 is the lack of uniformity with chromaticity, the three dimension color space in rectangular coordinates is not visually uniformed.
The CIE 1976 (also called CIELUV) was created by the CIE in 1976. It was put forward in an attempt to provide a more uniform color spacing than CIE 1931 for colors at approximately the same luminance
The CIE 1976 standard colour space is more linear and variations in perceived colour between different people has also been reduced. The disproportionately large green-turquoise area in CIE 1931, which cannot be generated with existing computer screens, has been reduced.
If we move from CIE 1931 to the CIE 1976 standard colour space we can see that the improvements made in the gamut for the “new” iPad screen (as compared to the “old” iPad 2) are more evident in the CIE 1976 colour space than in the CIE 1931 colour space, particularly in the blues from aqua to deep blue.
https://dot-color.com/2012/08/14/color-space-confusion/
Despite its age, CIE 1931, named for the year of its adoption, remains a well-worn and familiar shorthand throughout the display industry. CIE 1931 is the primary language of customers. When a customer says that their current display “can do 72% of NTSC,” they implicitly mean 72% of NTSC 1953 color gamut as mapped against CIE 1931.
-
Photography Basics : Spectral Sensitivity Estimation Without a Camera
Read more: Photography Basics : Spectral Sensitivity Estimation Without a Camerahttps://color-lab-eilat.github.io/Spectral-sensitivity-estimation-web/
A number of problems in computer vision and related fields would be mitigated if camera spectral sensitivities were known. As consumer cameras are not designed for high-precision visual tasks, manufacturers do not disclose spectral sensitivities. Their estimation requires a costly optical setup, which triggered researchers to come up with numerous indirect methods that aim to lower cost and complexity by using color targets. However, the use of color targets gives rise to new complications that make the estimation more difficult, and consequently, there currently exists no simple, low-cost, robust go-to method for spectral sensitivity estimation that non-specialized research labs can adopt. Furthermore, even if not limited by hardware or cost, researchers frequently work with imagery from multiple cameras that they do not have in their possession.
To provide a practical solution to this problem, we propose a framework for spectral sensitivity estimation that not only does not require any hardware (including a color target), but also does not require physical access to the camera itself. Similar to other work, we formulate an optimization problem that minimizes a two-term objective function: a camera-specific term from a system of equations, and a universal term that bounds the solution space.
Different than other work, we utilize publicly available high-quality calibration data to construct both terms. We use the colorimetric mapping matrices provided by the Adobe DNG Converter to formulate the camera-specific system of equations, and constrain the solutions using an autoencoder trained on a database of ground-truth curves. On average, we achieve reconstruction errors as low as those that can arise due to manufacturing imperfections between two copies of the same camera. We provide predicted sensitivities for more than 1,000 cameras that the Adobe DNG Converter currently supports, and discuss which tasks can become trivial when camera responses are available.
-
Björn Ottosson – OKHSV and OKHSL – Two new color spaces for color picking
Read more: Björn Ottosson – OKHSV and OKHSL – Two new color spaces for color pickinghttps://bottosson.github.io/misc/colorpicker
https://bottosson.github.io/posts/colorpicker/
https://www.smashingmagazine.com/2024/10/interview-bjorn-ottosson-creator-oklab-color-space/
One problem with sRGB is that in a gradient between blue and white, it becomes a bit purple in the middle of the transition. That’s because sRGB really isn’t created to mimic how the eye sees colors; rather, it is based on how CRT monitors work. That means it works with certain frequencies of red, green, and blue, and also the non-linear coding called gamma. It’s a miracle it works as well as it does, but it’s not connected to color perception. When using those tools, you sometimes get surprising results, like purple in the gradient.
There were also attempts to create simple models matching human perception based on XYZ, but as it turned out, it’s not possible to model all color vision that way. Perception of color is incredibly complex and depends, among other things, on whether it is dark or light in the room and the background color it is against. When you look at a photograph, it also depends on what you think the color of the light source is. The dress is a typical example of color vision being very context-dependent. It is almost impossible to model this perfectly.
I based Oklab on two other color spaces, CIECAM16 and IPT. I used the lightness and saturation prediction from CIECAM16, which is a color appearance model, as a target. I actually wanted to use the datasets used to create CIECAM16, but I couldn’t find them.
IPT was designed to have better hue uniformity. In experiments, they asked people to match light and dark colors, saturated and unsaturated colors, which resulted in a dataset for which colors, subjectively, have the same hue. IPT has a few other issues but is the basis for hue in Oklab.
In the Munsell color system, colors are described with three parameters, designed to match the perceived appearance of colors: Hue, Chroma and Value. The parameters are designed to be independent and each have a uniform scale. This results in a color solid with an irregular shape. The parameters are designed to be independent and each have a uniform scale. This results in a color solid with an irregular shape. Modern color spaces and models, such as CIELAB, Cam16 and Björn Ottosson own Oklab, are very similar in their construction.
By far the most used color spaces today for color picking are HSL and HSV, two representations introduced in the classic 1978 paper “Color Spaces for Computer Graphics”. HSL and HSV designed to roughly correlate with perceptual color properties while being very simple and cheap to compute.
Today HSL and HSV are most commonly used together with the sRGB color space.
One of the main advantages of HSL and HSV over the different Lab color spaces is that they map the sRGB gamut to a cylinder. This makes them easy to use since all parameters can be changed independently, without the risk of creating colors outside of the target gamut.
The main drawback on the other hand is that their properties don’t match human perception particularly well.
Reconciling these conflicting goals perfectly isn’t possible, but given that HSV and HSL don’t use anything derived from experiments relating to human perception, creating something that makes a better tradeoff does not seem unreasonable.
With this new lightness estimate, we are ready to look into the construction of Okhsv and Okhsl.
LIGHTING
-
Bella – Fast Spectral Rendering
Read more: Bella – Fast Spectral Rendering
Bella works in spectral space, allowing effects such as BSDF wavelength dependency, diffraction, or atmosphere to be modeled far more accurately than in color space.
https://superrendersfarm.com/blog/uncategorized/bella-a-new-spectral-physically-based-renderer/
-
HDRI Median Cut plugin
Read more: HDRI Median Cut pluginwww.hdrlabs.com/picturenaut/plugins.html
Note. The Median Cut algorithm is typically used for color quantization, which involves reducing the number of colors in an image while preserving its visual quality. It doesn’t directly provide a way to identify the brightest areas in an image. However, if you’re interested in identifying the brightest areas, you might want to look into other methods like thresholding, histogram analysis, or edge detection, through openCV for example.
Here is an openCV example:
# bottom left coordinates = 0,0 import numpy as np import cv2 # Load the HDR or EXR image image = cv2.imread('your_image_path.exr', cv2.IMREAD_UNCHANGED) # Load as-is without modification # Calculate the luminance from the HDR channels (assuming RGB format) luminance = np.dot(image[..., :3], [0.299, 0.587, 0.114]) # Set a threshold value based on estimated EV threshold_value = 2.4 # Estimated threshold value based on 4.8 EV # Apply the threshold to identify bright areas # Theluminancearray contains the calculated luminance values for each pixel in the image. # Thethreshold_valueis a user-defined value that represents a cutoff point, separating "bright" and "dark" areas in terms of perceived luminance.thresholded = (luminance > threshold_value) * 255 # Convert the thresholded image to uint8 for contour detection thresholded = thresholded.astype(np.uint8) # Find contours of the bright areas contours, _ = cv2.findContours(thresholded, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE) # Create a list to store the bounding boxes of bright areas bright_areas = [] # Iterate through contours and extract bounding boxes for contour in contours: x, y, w, h = cv2.boundingRect(contour) # Adjust y-coordinate based on bottom-left origin y_bottom_left_origin = image.shape[0] - (y + h) bright_areas.append((x, y_bottom_left_origin, x + w, y_bottom_left_origin + h)) # Store as (x1, y1, x2, y2) # Print the identified bright areas print("Bright Areas (x1, y1, x2, y2):") for area in bright_areas: print(area)More details
Luminance and Exposure in an EXR Image:
- An EXR (Extended Dynamic Range) image format is often used to store high dynamic range (HDR) images that contain a wide range of luminance values, capturing both dark and bright areas.
- Luminance refers to the perceived brightness of a pixel in an image. In an RGB image, luminance is often calculated using a weighted sum of the red, green, and blue channels, where different weights are assigned to each channel to account for human perception.
- In an EXR image, the pixel values can represent radiometrically accurate scene values, including actual radiance or irradiance levels. These values are directly related to the amount of light emitted or reflected by objects in the scene.
The luminance line is calculating the luminance of each pixel in the image using a weighted sum of the red, green, and blue channels. The three float values [0.299, 0.587, 0.114] are the weights used to perform this calculation.
These weights are based on the concept of luminosity, which aims to approximate the perceived brightness of a color by taking into account the human eye’s sensitivity to different colors. The values are often derived from the NTSC (National Television System Committee) standard, which is used in various color image processing operations.
Here’s the breakdown of the float values:
- 0.299: Weight for the red channel.
- 0.587: Weight for the green channel.
- 0.114: Weight for the blue channel.
The weighted sum of these channels helps create a grayscale image where the pixel values represent the perceived brightness. This technique is often used when converting a color image to grayscale or when calculating luminance for certain operations, as it takes into account the human eye’s sensitivity to different colors.
For the threshold, remember that the exact relationship between EV values and pixel values can depend on the tone-mapping or normalization applied to the HDR image, as well as the dynamic range of the image itself.
To establish a relationship between exposure and the threshold value, you can consider the relationship between linear and logarithmic scales:
- Linear and Logarithmic Scales:
- Exposure values in an EXR image are often represented in logarithmic scales, such as EV (exposure value). Each increment in EV represents a doubling or halving of the amount of light captured.
- Threshold values for luminance thresholding are usually linear, representing an actual luminance level.
- Conversion Between Scales:
- To establish a mathematical relationship, you need to convert between the logarithmic exposure scale and the linear threshold scale.
- One common method is to use a power function. For instance, you can use a power function to convert EV to a linear intensity value.
threshold_value = base_value * (2 ** EV)Here,
EVis the exposure value,base_valueis a scaling factor that determines the relationship between EV and threshold_value, and2 ** EVis used to convert the logarithmic EV to a linear intensity value. - Choosing the Base Value:
- The
base_valuefactor should be determined based on the dynamic range of your EXR image and the specific luminance values you are dealing with. - You may need to experiment with different values of
base_valueto achieve the desired separation of bright areas from the rest of the image.
- The
Let’s say you have an EXR image with a dynamic range of 12 EV, which is a common range for many high dynamic range images. In this case, you want to set a threshold value that corresponds to a certain number of EV above the middle gray level (which is often considered to be around 0.18).
Here’s an example of how you might determine a
base_valueto achieve this:# Define the dynamic range of the image in EV dynamic_range = 12 # Choose the desired number of EV above middle gray for thresholding desired_ev_above_middle_gray = 2 # Calculate the threshold value based on the desired EV above middle gray threshold_value = 0.18 * (2 ** (desired_ev_above_middle_gray / dynamic_range)) print("Threshold Value:", threshold_value) -
Convert between light exposure and intensity
Read more: Convert between light exposure and intensityimport math,sys def Exposure2Intensity(exposure): exp = float(exposure) result = math.pow(2,exp) print(result) Exposure2Intensity(0) def Intensity2Exposure(intensity): inarg = float(intensity) if inarg == 0: print("Exposure of zero intensity is undefined.") return if inarg < 1e-323: inarg = max(inarg, 1e-323) print("Exposure of negative intensities is undefined. Clamping to a very small value instead (1e-323)") result = math.log(inarg, 2) print(result) Intensity2Exposure(0.1) -
Neural Microfacet Fields for Inverse Rendering
Read more: Neural Microfacet Fields for Inverse Renderinghttps://half-potato.gitlab.io/posts/nmf/
-
Light and Matter : The 2018 theory of Physically-Based Rendering and Shading by Allegorithmic
Read more: Light and Matter : The 2018 theory of Physically-Based Rendering and Shading by Allegorithmicacademy.substance3d.com/courses/the-pbr-guide-part-1
academy.substance3d.com/courses/the-pbr-guide-part-2
Local copy:
-
Rendering – BRDF – Bidirectional reflectance distribution function
Read more: Rendering – BRDF – Bidirectional reflectance distribution functionhttp://en.wikipedia.org/wiki/Bidirectional_reflectance_distribution_function
The bidirectional reflectance distribution function is a four-dimensional function that defines how light is reflected at an opaque surface
http://www.cs.ucla.edu/~zhu/tutorial/An_Introduction_to_BRDF-Based_Lighting.pdf
In general, when light interacts with matter, a complicated light-matter dynamic occurs. This interaction depends on the physical characteristics of the light as well as the physical composition and characteristics of the matter.
That is, some of the incident light is reflected, some of the light is transmitted, and another portion of the light is absorbed by the medium itself.
A BRDF describes how much light is reflected when light makes contact with a certain material. Similarly, a BTDF (Bi-directional Transmission Distribution Function) describes how much light is transmitted when light makes contact with a certain material
http://www.cs.princeton.edu/~smr/cs348c-97/surveypaper.html
It is difficult to establish exactly how far one should go in elaborating the surface model. A truly complete representation of the reflective behavior of a surface might take into account such phenomena as polarization, scattering, fluorescence, and phosphorescence, all of which might vary with position on the surface. Therefore, the variables in this complete function would be:
incoming and outgoing angle incoming and outgoing wavelength incoming and outgoing polarization (both linear and circular) incoming and outgoing position (which might differ due to subsurface scattering) time delay between the incoming and outgoing light ray
-
ICLight – Krea and ComfyUI light editing
Read more: ICLight – Krea and ComfyUI light editing
https://drive.google.com/drive/folders/16Aq1mqZKP-h8vApaN4FX5at3acidqPUv
https://github.com/lllyasviel/IC-Light
https://generativematte.blogspot.com/2025/03/comfyui-ic-light-relighting-exploration.html
Workflow Local copy
COLLECTIONS
| Featured AI
| Design And Composition
| Explore posts
POPULAR SEARCHES
unreal | pipeline | virtual production | free | learn | photoshop | 360 | macro | google | nvidia | resolution | open source | hdri | real-time | photography basics | nuke
FEATURED POSTS
Social Links
DISCLAIMER – Links and images on this website may be protected by the respective owners’ copyright. All data submitted by users through this site shall be treated as freely available to share.
