Autumn Leaf - breaking some "Rules"

[xpatUSA]

Posted bigger than 1600px
Not shot with a CaNikony
Shot with a cheap Sigma 17-50mm zoom
Close-up shot at f/2.8
Shot under a domestic cool white LED in full-spectrum i.e no UV-IR blocking filter
Incandescent in-camera White Balance
Embedded JPEG processed in FastStone Viewer
Not color-balanced in post
No dodging or burning
Object dead-center

Voila:


Comments/critiques welcome.

[Geoff F]

Interesting idea. Maybe I would clone out the holes in the leaf and a few small spots on the cloth background. Probably better without that crease in the background?

[xpatUSA]

Interesting idea. Maybe I would clone out the holes in the leaf and a few small spots on the cloth background. Probably better without that crease in the background?

Thanks Geoff. I agree about the spots on the cloth and the crease - it was a quick shot that came out well enough to post as-is with no detail adjustments.

Cloning out the spots and the crease might not be so easy if one wants to keep the Foveon-rendered weft and warp aligned ...

I used the cloth because it was in a complementary color and readily to hand. Might be better on a plain paper background with adjusted HSL.

I'll be keeping the leaf's imperfections though.

[Wavelength]

I think a Natural diffused light would have been better

[xpatUSA]

I think a Natural diffused light would have been better

Natural diffused light from say an overcast sky would have reduced the shadow tones and also the contrast but please remember that I shot with a full-spectrum camera so infra-red would have had an effect, too.

[Manfred M]

How much IR does a white LED produce? I thought that they were generally biased towards the blue end of the spectrum, not the red end.

[xpatUSA]

How much IR does a white LED produce? I thought that they were generally biased towards the blue end of the spectrum, not the red end.

There are many many varieties of "a White LED", Manfred.

Here's my 3500K LED studio lights, FWIW:



As you can see - very little NIR emission but a lot of orange/red ...

[Manfred M]

Thanks Ted. Surprisingly large drop off in the blues but certainly a decent bit of IR in the 700 - 725nm range.

[xpatUSA]

Thanks Ted. Surprisingly large drop off in the blues

The reason for that is that the bare photo-emitter gives out a 450nm peak with approximately a normal distribution each side - perhaps from 400 to 500nm. All the other output upwards is from phosphor coatings starting from, say, 470nm.

but certainly a decent bit of IR in the 700 - 725nm range.

That range is somewhat debatable, Manfred, (terminology). Some might just call it "really red".

Many sources quote "visible" as going up to 770 or 780nm and IR is beyond that. But, as I'm sure you know, "visible" light doesn't stop dead at 700nm. So, for example, 800nm of near IR at sufficient power could be visible even to the average human - might hurt though.

Pardon my pedantry ...

[Manfred M]

The reason for that is that the bare photo-emitter gives out a 450nm peak with approximately a normal distribution each side - perhaps from 400 to 500nm. All the other output upwards is from phosphor coatings starting from, say, 470nm.



That range is somewhat debatable, Manfred, (terminology). Some might just call it "really red".

Many sources quote "visible" as going up to 770 or 780nm and IR is beyond that. But, as I'm sure you know, "visible" light doesn't stop dead at 700nm. So, for example, 800nm of near IR at sufficient power could be visible even to the average human - might hurt though.

Pardon my pedantry ...

Thanks Ted - I'm aware of the arbitrary definition of the boundaries of the visible spectrum and have read that there are some people (seems to be primarily females) that can see into wavelengths that are nominally into the near IR and near UV. Us older guys, less so...

I had also read that phosphors are used in some white LEDs to produce "white". The ~450nm spike on the graph obviously corresponds to the actual emissions of the LED. Nice to see that the output does not have the colour spikes we see with fluorescent light sources. I expect that is related to being able to fine tune the phosphors as so little is required in an LED versus "affordable" materials used in fluorescent lights.

[xpatUSA]

I had also read that phosphors are used in some white LEDs to produce "white". The ~450nm spike on the graph obviously corresponds to the actual emissions of the LED. Nice to see that the output does not have the colour spikes we see with fluorescent light sources. I expect that is related to being able to fine tune the phosphors as so little is required in an LED versus "affordable" materials used in fluorescent lights.

Quite so. Out of interest, I started with these 5000K LED floods but got tired of coping with the obvious blue cast:



They are now doing well as yard lights.