• 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Forums > Back > How could image stacking increase dynamic range ??
#1
Was just reading this tutorial on Dpreview 

https://www.dpreview.com/articles/0727694641/here-s-how-to-pixel-shift-with-any-camera


they noted "Pentax K-1 II's new hand-held Pixel Shift mode which combines four images to create a 'super resolution' file in-camera with better detail, dynamic range and lower noise. "
it seems stacking multiple photos can increase dynamic range but how ? it they are taken with the same exposure settings ?
  Reply
#2
Because you capture 4x the light, that is why.
Or (iow): DR gets determined by establishing a noise floor, and then seeing when that noise floor gets reached. By taking 4 samples and averaging them, you also average the noise per pixel, and the result is a higher DR.
  Reply
#3
(04-27-2018, 09:36 AM)Brightcolours Wrote: Because you capture 4x the light, that is why.
Or (iow): DR gets determined by establishing a noise floor, and then seeing when that noise floor gets reached. By taking 4 samples and averaging them, you also average the noise per pixel, and the result is a higher DR.

didn't get it, too complicated to me can you explain more please ?
  Reply
#4
BC's explanantion is absoultely correct and rather straightforward to understand. If you have problems with it, try for yourself. 

DPReview gave a good recipe to do the same what Pentax K1 could do with one push of the release button, but occassional artefacts with 4 pushs, more post-work but no artefacts. Just don't fprget to enlarge the file to 200%.
  Reply
#5
Never said BC was not correct, just I didn't get it, Xan someone explain that in a simple way I could understand?
  Reply
#6
Try this video Toni!

https://www.youtube.com/watch?v=E-GbuXU8YMQ
  Reply
#7
(04-27-2018, 03:35 PM)toni-a Wrote: Never said BC was not correct, just I didn't get it, Xan someone explain that in a simple way I could understand?

well, to make it simple, take 4 times the same picture.    Now, every pixel has a bit of noise. In dark areas, the light signal is so week, it's difficult to see what is the image and what is noise. 
let's take a single pixel.   It should have, for example, a grey value of 12 ideally, but with noise in one image it's 15, in another it's 11, then 9, last 13.    Now you average the 4 pixels, the noise will be less because you average over 4 frames. So, totally, you decrease the noise in the image, which increases the DR, since you have now more good information in the dark regions.
  Reply


Forum Jump:


Users browsing this thread: 2 Guest(s)