Timeline for Why does a bigger sensor have better dynamic range?
Current License: CC BY-SA 3.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 22, 2017 at 18:43 | comment | added | Jody Bruchon | @erotsppa The surface area used to calculate each final pixel is all that matters. If you have 4 tiny photosites and 1 large photosize of the same size, the average of the 4 will be almost exactly the same as the large single photosite. Having more photosites in the same surface area just increases the sampling frequency; each will have worse characteristics by itself but if the photo is scaled down to the same pixel density as if they were all just one big photosite it would look no different. This is why cell phone cameras look fine on phone screens and poor on computer screens. | |
| May 15, 2013 at 16:04 | comment | added | erotsppa | Am I correct in saying that, given the same read noise (technology etc) the larger sensor with bigger pixels will have less photon noise and thus better dynamic range because of increase latitude between the lowest charge and saturation point? | |
| May 15, 2013 at 14:04 | comment | added | jrista | Per-pixel CDS was eliminated. I don't even think each pixel has an amplifier, as with near-noiseless output, all ISO settings can be achieved with digital amplification. Analog pixel charge is converted immediately to a digital unit, and from that point on error-corrected information transfer of digital information is utilized throughout the rest of the image processing pipeline. That eliminates the chance for contamination of the analog signal by interference of any kind as early as possible. | |
| May 15, 2013 at 14:03 | comment | added | jrista | The benefit of Sony Exmor is that it is digital readout. The signal is only analog so long as the charge remains in the pixels. Upon read, on-die column-parallel ADC immediately converts the analog charge of each pixel into a digital unit. High frequency components (clock, pll, etc.) are placed elsewhere on the die to eliminate high frequency noise introduction in the ADC circuitry. The image signal is not read twice...a "reset read" is performed when the sensor is reset to accumulate digital CDS as a "negative result", which is then applied to the "image read" upon exposure. | |
| May 15, 2013 at 8:00 | history | answered | Matt Grum | CC BY-SA 3.0 |