Y'CbCr is a perceptually-motivated color space in the sense that it is designed based on how humans perceive color differences and is often used in digital video formats like JPEG as it is useful in compression of data while maintaining high quality of image/ video. The Y' component represents the brightness of the image and since human eyes are more perceptive of variations in luminance than color, this is the most important part of the video signal and most data is required here. Cb and Cr (chroma) correspond to the blue0yellow and red-green differences from grey and are often subsampled as the human eye is less sensitive to chroma differences. About nonlinear gamma encoding, gamma encoding is a technique used to map brightness levels of images to a scale that is aligned with human vision. Our eyes perceive changes in brightness in a nonlinear manner so they are more sensitive to variations in darker tones relative to bright tones. I.e. if we increase the brightness of a scene by double, our eyes perceive this increase on a nonlinear scale. The gamma curve is applied to the linear brightness values to account for this. This occurs in the Y' channel. The linear light intensity information is gamma-encoded to match how human eyes perceive light. Before the image is dispalyed, the gamma-encoded Y' values are decoded to ensure that the displayed brightness accurately represents the original scene. Overall, gamma encoding in the Y' channel maps the image's brightness to a scale that matches the human visual system more accurately and compresses the data into a form that maintains details required for digital representation by storing the data required to maintain differences detectable to our eyes.
ElShroomster
It's crazy to think that putting those three images results in something with as much quality as the one on the left. I know we are normally used to doing projects in the RGB color space, but I think it would be very cool to do projects in the Y'CbCr color space since we could make cool shaders and lighting (this overall just seems cooler than RGB...)
yangbright-2001
Y'Cb'Cr color space is an innovative design, it builds a new type of chroma channel by combining the traditional blue-yellow and red-green, it also contains emphasis on human eye's sensitiveness on illuminance
JunoLee128
It's interesting how this combines numerical compression theory (Shannon theorem? etc) with human perception science (which seems more "subjective" but offers better results). I like how we can combine these concepts into such an oft-used algorithm
Y'CbCr is a perceptually-motivated color space in the sense that it is designed based on how humans perceive color differences and is often used in digital video formats like JPEG as it is useful in compression of data while maintaining high quality of image/ video. The Y' component represents the brightness of the image and since human eyes are more perceptive of variations in luminance than color, this is the most important part of the video signal and most data is required here. Cb and Cr (chroma) correspond to the blue0yellow and red-green differences from grey and are often subsampled as the human eye is less sensitive to chroma differences. About nonlinear gamma encoding, gamma encoding is a technique used to map brightness levels of images to a scale that is aligned with human vision. Our eyes perceive changes in brightness in a nonlinear manner so they are more sensitive to variations in darker tones relative to bright tones. I.e. if we increase the brightness of a scene by double, our eyes perceive this increase on a nonlinear scale. The gamma curve is applied to the linear brightness values to account for this. This occurs in the Y' channel. The linear light intensity information is gamma-encoded to match how human eyes perceive light. Before the image is dispalyed, the gamma-encoded Y' values are decoded to ensure that the displayed brightness accurately represents the original scene. Overall, gamma encoding in the Y' channel maps the image's brightness to a scale that matches the human visual system more accurately and compresses the data into a form that maintains details required for digital representation by storing the data required to maintain differences detectable to our eyes.
It's crazy to think that putting those three images results in something with as much quality as the one on the left. I know we are normally used to doing projects in the RGB color space, but I think it would be very cool to do projects in the Y'CbCr color space since we could make cool shaders and lighting (this overall just seems cooler than RGB...)
Y'Cb'Cr color space is an innovative design, it builds a new type of chroma channel by combining the traditional blue-yellow and red-green, it also contains emphasis on human eye's sensitiveness on illuminance
It's interesting how this combines numerical compression theory (Shannon theorem? etc) with human perception science (which seems more "subjective" but offers better results). I like how we can combine these concepts into such an oft-used algorithm