Can underexposing and then downsampling improve dynamic range or match a longer exposure?

Asked 5/2/2015

5 views

2 answers

0

In microscopy, sensor binning can improve signal-to-noise ratio by combining neighboring pixels. I’m wondering whether the same idea applies to modern DSLR/CMOS cameras.

If I underexpose by one stop and then downsample the image later (for example, combine 2×2 pixels), will the result approach the image quality of a properly exposed shot at full resolution?

Similarly, how does a shot made at a higher ISO and then downsampled compare with a lower-ISO shot using a longer exposure?

Do practical cameras implement any true sensor-level binning for still photos, or is resolution reduction usually done only after the sensor has already been read out?

Originally by Photography Stack Exchange contributor. Source · Licensed CC BY-SA 4.0

Photography Stack Exchange contributor

11y ago

2 Answers

4

Yes, binning will always increase dynamic range. There's no need to underexpose. Binning (i.e. adding) 4 (presumably 2x2) input pixels to create 1 output pixel gives you 4x signal but only 2x the shot noise. Shot/Poisson noise is proportional to the square root of signal level, so this noise source is only halved not quartered. Other noise sources may also exist, but for a well exposed pixel the shot noise is predominant. So your final signal-to-noise ratio will be doubled after a 4 pixel binning.

Originally by user70246. Source · Licensed CC BY-SA 4.0

user70246

8y ago

0

AI Answer

Generated from our catalog & community — verify before relying on it.

Usually no: with modern DSLRs, underexposing and then downsampling later does not recover the image quality of a properly exposed longer exposure.

Why: true binning helps when charges are combined before readout, because signal adds faster than shot noise. But community answers note that modern DSLRs generally do not do analog binning for stills; they read each pixel first, then reduce resolution afterward. Once read noise has already been added per pixel, downsampling cannot undo the SNR loss from underexposure.

So if you underexpose by 1 stop, you collect fewer photons and lose SNR and dynamic range. Downsampling may make noise look less visible, but it won’t equal the cleaner result of giving the sensor more light in the first place.

Compared with simply underexposing, raising ISO can be slightly better because amplification happens before readout, which can help relative to read noise. But high ISO plus downsampling still is not the same as using a lower ISO with a longer exposure, assuming the longer exposure is practical and doesn’t introduce blur or clipping.

In short: for still photography, more exposure beats underexpose-and-downsample.

UniqueBot

AI

11y ago

Your Answer