JohnR wrote:
Can a pixel vary in color across its dimensions or must each pixel be a single discrete color/shade?
Yes one pixel can have varied colors hitting it across it's dimension. But, since each pixel only has one pair of wires leaving to report it's output voltage to the processor only one color/brightness can be reported.
terryMc
Loc: Arizona's White Mountains
CHG_CANON wrote:
Please ... there's a single BRIGHT LINE through this supposed maze ...
Capture in RAW with a full-frame mirrorless Canon. Protect your valuable lenses with B+W XS-Pro filters (coupled with the lens hood). Focus with your BBF assigned button in Continuous / AI-Servo mode. Edit your RAW files in LR Classic on a Windows-based machine. Export your sRGB JPEGs, and post and print to achieve the Photography Glory you first imagined when deciding to own a digital camera.
You're just so precious...
R.G. wrote:
I'd like to join in but I keep losing count
.
Maybe that's a good thing.
---
Longshadow wrote:
Thanks!
I forgot to add mirrorless vs. mirrored!
And Nikon vs Canon vs Sony etc.
kymarto
Loc: Portland OR and Milan Italy
Delderby wrote:
But I am correct in thinking that there will always be a trade-off between IQ and detail when comparing pixel numbers and pixel size in any given sensor? I'm thinking that less pixels means larger pixels, which logically means more room for more detail?
Each pixel has a single luminance value and a single chrominance value. Other things being equal, the only advantage to larger photosites is better low light response. Things like color bit depth and dynamic range depend more on sensor design.
kymarto wrote:
Each pixel has a single luminance value and a single chrominance value. Other things being equal, the only advantage to larger photosites is better low light response. Things like color bit depth and dynamic range depend more on sensor design.
Actually, each photosite only has a luminance value. It is only luminance values that are read during the sensor readout. During Demosaicing, the process knows what color filter is over each photosite and then mathematically determines a Red, a Green and a Blue value for each pixel in the output file by averaging some number of photosites together, where it only averages red sites, then only green sites and then only blue sites..
Please note that the Demosaicing process follows rules and is determined by the camera manufacturer and may vary from one model to another model. And it may vary between different Software suppliers. In fact, RawTherapee provides about 10 different Demosaicing algorithms to choose between.
Also, when a new camera is released, the camera manufacturer is responsible for releasing Demosaicing rules for the Software manufacturers to follow. This is why you see a new camera work in JPEG only at first, and then later, RAW files can be used after the manufacture releases the rules and the Software people incorporate it into their product.
This means that demosaicing is also not an exact science. Colors may vary from one process to another process.
Also, fine detail can be impacted too when Demosaicing uses a greater number of pixels to average to get better color. This is a trade off that has to be considered by the manufacturer. Also considered is how it senses edges since the averaging may have to be adjusted. The process is not simple.
If you want to reply, then
register here. Registration is free and your account is created instantly, so you can post right away.