Canon IXUS 240 HS
Canon PowerShot SX240 HS
Nikon D800
Samsung DV300F
Olympus OM-D E-M5
Nikon Coolpix P510
Canon PowerShot SX260 HS
Panasonic Lumix DMC-TZ30
Nikon Coolpix S9300
Understanding Colour Fringing
Ever wondered why some of your digital photographs have coloured fringes along the edges of your subject? PhotoReview.com.au explains that this is caused by one of two things; chromatic aberrations and blooming, and suggests a couple of software packages that will remove the unwanted effects (found via Digicamera.net).
“If the picture was captured on film, deciding what caused the coloured fringing is easy: only chromatic aberration can be blamed. It’s more difficult if the shot was taken with a digital camera because both blooming and chromatic aberration can create coloured fringes. Teasing out which one is relatively easy once you understand how the fringes are produced.”
Website: PhotoReview.com.au - Colour fringing explained
Canon IXUS 240 HS Review
Canon PowerShot SX240 HS Review
Nikon D800 Review
Samsung DV300F Review
Olympus OM-D E-M5 Review
Nikon Coolpix P510
Panasonic Lumix DMC-TZ30
Canon PowerShot SX260 HS
Fujifilm X-Pro1
Sony Cyber-shot DSC-HX200V
#1 Olivier_G
I have a different explanation about the usual purple fringing seen in digicams:
- Longitudinal Chromatic Aberration (with achromatic lenses)
AND
- Saturation in the Sensor (not Blooming)
The explanation is here: http://forums.dpreview.com/forums/read.asp?forum=1000&message=9701481
Image of a Simulated model:
http://olivier.gallen.mageos.com/Temp/purple-fringing.jpg
(last line)
In the article you linked, I am particularly sceptical about the following: "When data from a group of pixels affected by IR blooming is interpolated, the resulting pixels will have higher signal levels from red and blue than from the green channel"
IR => Red and Blue... huh??? Why?
Olivier
1:00 pm - Thursday, January 27, 2005
#2 Mark Goldstein
I equally have no idea why the green channel would be different to the other two. Any colour scientists out there who can explain why this might be true (and if it actually is)?
5:23 pm - Thursday, January 27, 2005