jarvist
New member
GREYCstoration is a badly named but sophisticated method of removing noise from digital camera shots:
http://cimg.sourceforge.net/greycstoration/
Its free + opensource, with plugins for GIMP & GUI versions for windows etc.
It seems that you need fairly exacting 'recipes' to remove the noise but not destroy the detail, and I was wondering whether anyone has fiddled for this with caving shots, and how far they got?
And are there any other black-magic tricks for noise removal?
Previosuly all I've ever really done was Gaussian Blur to smooth noise, then fiddled with levels + dodged/burned (generally with the 'highlights' choice in gimp).
I've also read of amateur astronomers taking a 'dark' photo for the same exposure as the real one, and then deducting it directly from the genuine shot (pixel by pixel) to get rid of some noise. Has anyone tried this? And if so, does one need to take the 'dark' shot at the same time as the exposure (same sensor temperature, temperament, phase of moon, etc.), or can one tabulate a whole set of different exposures at home? And how much does the noise fluctuate in both the short and the long term? And do digital cameras now-adays already try and do this kind of trickery by themself?
I believe that digital camera 'long exposures' are actually multiple short exposures when are then composed together by the digital camera (including a deconvolution of the expected sensor noise?)...
I've only tried a couple of photos with GREYCstoration so far, but here they are along with settings used for anyone who's interested:
Gaping Gill Main chamber, 15s exposure with (mainly) ambient light from above, Canon A520.
-alpha 1.6 -p 1.2 -dt 100
Orig (Click on photos to see full res):
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/GG_mainchamb_orig.html
Restored:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/GG_mainchamb_GREYCstoration_a16_p12_dt100_gimp_burn.html
Orig:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/GG_mainchamb2_orig.html
Restored:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/GG_mainchamb2_GREYCstoration_a16_p12_dt100_gimp_burn.html
And for a flash photo with lower noise; Avalanche Inlet, flash is Vivtar 283 on Firefly, Canon A520 200ASA(?) 1/60s
-alpha 1.6 -p 1.2 -dt 50
Orig:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/Avalanche_orig.html
Restored:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/Avalanche_GREYCstoration_a16_p12_dt50_gimp_burn.html
Other photos from that trip:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/
http://cimg.sourceforge.net/greycstoration/
Its free + opensource, with plugins for GIMP & GUI versions for windows etc.
It seems that you need fairly exacting 'recipes' to remove the noise but not destroy the detail, and I was wondering whether anyone has fiddled for this with caving shots, and how far they got?
And are there any other black-magic tricks for noise removal?
Previosuly all I've ever really done was Gaussian Blur to smooth noise, then fiddled with levels + dodged/burned (generally with the 'highlights' choice in gimp).
I've also read of amateur astronomers taking a 'dark' photo for the same exposure as the real one, and then deducting it directly from the genuine shot (pixel by pixel) to get rid of some noise. Has anyone tried this? And if so, does one need to take the 'dark' shot at the same time as the exposure (same sensor temperature, temperament, phase of moon, etc.), or can one tabulate a whole set of different exposures at home? And how much does the noise fluctuate in both the short and the long term? And do digital cameras now-adays already try and do this kind of trickery by themself?
I believe that digital camera 'long exposures' are actually multiple short exposures when are then composed together by the digital camera (including a deconvolution of the expected sensor noise?)...
I've only tried a couple of photos with GREYCstoration so far, but here they are along with settings used for anyone who's interested:
Gaping Gill Main chamber, 15s exposure with (mainly) ambient light from above, Canon A520.
-alpha 1.6 -p 1.2 -dt 100
Orig (Click on photos to see full res):
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/GG_mainchamb_orig.html
Restored:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/GG_mainchamb_GREYCstoration_a16_p12_dt100_gimp_burn.html
Orig:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/GG_mainchamb2_orig.html
Restored:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/GG_mainchamb2_GREYCstoration_a16_p12_dt100_gimp_burn.html
And for a flash photo with lower noise; Avalanche Inlet, flash is Vivtar 283 on Firefly, Canon A520 200ASA(?) 1/60s
-alpha 1.6 -p 1.2 -dt 50
Orig:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/Avalanche_orig.html
Restored:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/Avalanche_GREYCstoration_a16_p12_dt50_gimp_burn.html
Other photos from that trip:
http://www.union.ic.ac.uk/rcc/caving/photo_archive/trips/2008-04-25%20-%20avalanche2/
