Discussion
Weird. I've seen that on my laptop and images from digital projectors but that's just due to a limitation on the range of colours it can display.
Could it be a form of over exposure/hitting the boundary of the colour space ? That you've hit the upper limit of what can be captured so the gentle gradation in the sky has to be approximated to in 1 bit steps that cover a wide area. The fact that there's one dominant colour making it show up particularly clearly.
I'm guessing here, but could it be that if you're using sRGB instead of Adobe format, or in 8-bit rather than 16-bit capture this might be more likely to happen ?
Could it be a form of over exposure/hitting the boundary of the colour space ? That you've hit the upper limit of what can be captured so the gentle gradation in the sky has to be approximated to in 1 bit steps that cover a wide area. The fact that there's one dominant colour making it show up particularly clearly.
I'm guessing here, but could it be that if you're using sRGB instead of Adobe format, or in 8-bit rather than 16-bit capture this might be more likely to happen ?
I've got something similar from sunrise shots. I can only put it down to the effects of severely overexposing, so that the "bleed" effect starts to take place from one sensor to the next. You then lose the expected graduation of intensity from one pixel to the next.
Like here http://image42.webshots.com/43/2/8/31/348020831fwVAUO_ph.jpg
Like here http://image42.webshots.com/43/2/8/31/348020831fwVAUO_ph.jpg
I'd say this is just down to one of the limitations of digital - they're just not as good as film for shooting into the sun.
Basically, a certain change in light levels causes more of a change on a digital sensor than it does on film, so it appears to be more stepped, and this will be exaggerated if you shoot jpg/sRGB/8-bit as opposed to RAW/Adobe RGB/16-bit.
Even my 1Ds II produces c
p sunrise/sunset pics, so that's when I'll dig out some of my trusty Velvia 
some website said:
More of an issue is that digital has no “shoulder” – that is, the light sensitivity is more of a straight line than an s-curve. With film, the amount of detail in blown highlights or dark shadows tails off smoothly; with digital, highlights just blow with no gentle tail-off.
Basically, a certain change in light levels causes more of a change on a digital sensor than it does on film, so it appears to be more stepped, and this will be exaggerated if you shoot jpg/sRGB/8-bit as opposed to RAW/Adobe RGB/16-bit.
Even my 1Ds II produces c
p sunrise/sunset pics, so that's when I'll dig out some of my trusty Velvia 
V6GTO said:
What filter did you have on the lens? Looks to me like it's bending the light (I presume it's a good lens)?
Canon 100-400L IS @135mm with no filters.
1/100 @ f8
Martin.
Hmm.
It is quite interesting (though I am not sure what it means) to have a look at the Levels of the different colour channels in the 100 crop vs the original. The differences in the ranges for each channel for the sky part of the shot become much more obvious.
Einion Yrth said:
What filter did you have on the lens? Looks to me like it's bending the light (I presume it's a good lens)?
[ pedant ] If it weren't bending the light it wouldn't be a lens, it would be a tube [ /pedant ]
point taken
...but I meant.. if there was a UV with imperfections (you know how cheap these Noble drivers are) stuck to the front of the lens - with the sun at that angle, then even the best of lenses will get refracted light which might have caused the problem for the 'oh so dumb' CCD.
Gassing Station | Photography & Video | Top of Page | What's New | My Stuff