Orchids, neon-flourescent fringing, and & F707

[Robert63289]

Last week Ed looked at my F707 orchid pictures and called my attention to the magenta/plastic/flourescent/neon rim at the intense red edges of my orchid lips. I realize now I did not understand the problem--and still don't.

I invite discussion of this 'magenta' effect, most recently evidenced by this F707 picture of my blooming mini-Cattleya, (Blc. Jane Helton x Slc. Vallezac) 'Kudo's Golden Dragon' x Slc. Kauai Star Bright 'Walter':

http://www.pbase.com/image/450557/large

This fringe is not on the plant.

In the "neon" lip edging, the histogram of red values is crowded to the top, but it does does not look cutoff at 255.

However, the green and blue are obviously under exposed in the histogram.

Is that the problem? Perhaps an exposure does not capture enough blue-green information before the red channel tops out and the exposure ends? If so, is a blue filter indicated? I could imagine there is a calibrated approach: take an exposure through a bluish filter (with "indoor" white balance selected, in my case), then, on the computer, do some Photoshop operation to restore the correct balance of red, green, and blue.

Is this this reasonable? What do you recommend?

===Robert

 

[Ed (;¬ >)]

I find it interesting that you chose this image. Of all the ones showing this magenta cast it shows the least. Some of the other deeper reds don't show it at all, and one of the plants (Phal.equestris) shows bright white with just a hint of purple.

The images that caught my eye were, in particular, Slc. Pumpkin Festival 'Fong Yuen' and also C. Walkeriana alba 'Pendentive' AM/AOS x Lc. Mari's Song 'Orchidlibrary'.

Offhand I would think that a filter might be bludgeoning the problem. It might be as simple as reducing the back lighting. Or maybe just a Phaotoshop tweak. I haven't time tonight to play with it in PS, but have you tried?

Just off the top of my head I think its the strong backlight vs the key light. It is possible that the flower is actually flourescing. Perhaps a UV filter would have an effect.

-Ed

Last week Ed looked at my F707 orchid pictures and called my
attention to the magenta/plastic/flourescent/neon rim at the
intense red edges of my orchid lips. I realize now I did not
understand the problem--and still don't.

I invite discussion of this 'magenta' effect, most recently
evidenced by this F707 picture of my blooming mini-Cattleya, (Blc.
Jane Helton x Slc. Vallezac) 'Kudo's Golden Dragon' x Slc. Kauai
Star Bright 'Walter':

http://www.pbase.com/image/450557/large

This fringe is not on the plant.

In the "neon" lip edging, the histogram of red values is crowded to
the top, but it does does not look cutoff at 255.

However, the green and blue are obviously under exposed in the
histogram.

Is that the problem? Perhaps an exposure does not capture enough
blue-green information before the red channel tops out and the
exposure ends? If so, is a blue filter indicated? I could imagine
there is a calibrated approach: take an exposure through a bluish
filter (with "indoor" white balance selected, in my case), then, on
the computer, do some Photoshop operation to restore the correct
balance of red, green, and blue.

Is this this reasonable? What do you recommend?

===Robert

 

[David J. Littleboy]

Last week Ed looked at my F707 orchid pictures and called my
attention to the magenta/plastic/flourescent/neon rim at the
intense red edges of my orchid lips. I realize now I did not
understand the problem--and still don't.

Just a guess, but it might be sharpening artifacts. Try setting sharpening to -2, and look at the result before additional sharping.

David J. Littleboy
Tokyo, Japan

 

[Robert63289]

Ed, I chose that flower because it was the only imperfection I could find in it. The other pictures have multiple issues if they have this one.

===Robert

The images that caught my eye were, in particular, Slc. Pumpkin
Festival 'Fong Yuen' and also C. Walkeriana alba 'Pendentive'
AM/AOS x Lc. Mari's Song 'Orchidlibrary'.

Offhand I would think that a filter might be bludgeoning the
problem. It might be as simple as reducing the back lighting. Or
maybe just a Phaotoshop tweak. I haven't time tonight to play with
it in PS, but have you tried?

Just off the top of my head I think its the strong backlight vs the
key light. It is possible that the flower is actually flourescing.
Perhaps a UV filter would have an effect.

-Ed

Last week Ed looked at my F707 orchid pictures and called my
attention to the magenta/plastic/flourescent/neon rim at the
intense red edges of my orchid lips. I realize now I did not
understand the problem--and still don't.

I invite discussion of this 'magenta' effect, most recently
evidenced by this F707 picture of my blooming mini-Cattleya, (Blc.
Jane Helton x Slc. Vallezac) 'Kudo's Golden Dragon' x Slc. Kauai
Star Bright 'Walter':

http://www.pbase.com/image/450557/large

This fringe is not on the plant.

In the "neon" lip edging, the histogram of red values is crowded to
the top, but it does does not look cutoff at 255.

However, the green and blue are obviously under exposed in the
histogram.

Is that the problem? Perhaps an exposure does not capture enough
blue-green information before the red channel tops out and the
exposure ends? If so, is a blue filter indicated? I could imagine
there is a calibrated approach: take an exposure through a bluish
filter (with "indoor" white balance selected, in my case), then, on
the computer, do some Photoshop operation to restore the correct
balance of red, green, and blue.

Is this this reasonable? What do you recommend?

===Robert

 

[Perry]

I think you guys are getting a little too picky. I have seen worse problems with just about all the digital cameras. These orchid pictures look almost perfect if there is such a thing as perfect. Digital is not a 100% representation of the real thing and neither is film. Accept microscopic imperfections. If I were the owner of this camera I would be tickled pink with these photos. You got this guy thinking he bought a 1 M pixel camera from walmart with a 3x digital zoom. I am relatively new to digital and I think its pretty darn good compared to 35mm, otherwise I would not be spending my money on digital cameras.

Perry

 

[Mike Fitzgerald]

Perry --

I suspect that Robert is already "tickled pink" with the capabilities of his new camera, while harbouring no delusions about what he should be expecting for the price. He is evidently an old hand with Photoshop, and is no doubt ready to enjoy the result analyses and tweaking as much as the setting up.

We need to be very careful to distinguish "criticism" -- in the truest sense of the word -- from the directionless denigration of particular cameras that is, sadly, so often encountered in these forums and does nothing but chew up bandwidth.

The pure form of criticism we are seeing here can be both productive and enjoyable for participant and onlooker alike. Let's hope we can keep it at that level.

Mike

I think you guys are getting a little too picky. I have seen worse
problems with just about all the digital cameras. These orchid
pictures look almost perfect if there is such a thing as perfect.
Digital is not a 100% representation of the real thing and neither
is film. Accept microscopic imperfections. If I were the owner of
this camera I would be tickled pink with these photos. You got this
guy thinking he bought a 1 M pixel camera from walmart with a 3x
digital zoom. I am relatively new to digital and I think its pretty
darn good compared to 35mm, otherwise I would not be spending my
money on digital cameras.

Perry

 

[Mike Fitzgerald]

Hi Robert --

An interesting theory, and probably one that can be explored by overexposing -- forget the red subleties for the purpose of the exercise, let it clip so as to to permit the others to "normalise" to some extent, and see what happens.

Without knowing what's going on inside the 707 I'd still be more suspicious of a combination of other things: high background contrast (even with the actual background colour you mentioned) is an interesting one. Adjacency problems ("blooming") are well recognised with the high charge levels from extreme highlights, but what happens with moderate highlights against extreme lowlights? (rhetorical question -- I haven't a clue personally). There's also the possibility of in-camera sharpening and JPEG artefacts. These, at least, can be experimented with directly.

I'd also be trying a lighter background for shooting. With the colours in some of these subjects you've got Buckley's of using a general chroma-key approach, but a uniform mid-grey one should be easy enough to isolate in PS, at least for the exercise.

Mike

Last week Ed looked at my F707 orchid pictures and called my
attention to the magenta/plastic/flourescent/neon rim at the
intense red edges of my orchid lips. I realize now I did not
understand the problem--and still don't.

I invite discussion of this 'magenta' effect, most recently
evidenced by this F707 picture of my blooming mini-Cattleya, (Blc.
Jane Helton x Slc. Vallezac) 'Kudo's Golden Dragon' x Slc. Kauai
Star Bright 'Walter':

http://www.pbase.com/image/450557/large

This fringe is not on the plant.

In the "neon" lip edging, the histogram of red values is crowded to
the top, but it does does not look cutoff at 255.

However, the green and blue are obviously under exposed in the
histogram.

Is that the problem? Perhaps an exposure does not capture enough
blue-green information before the red channel tops out and the
exposure ends? If so, is a blue filter indicated? I could imagine
there is a calibrated approach: take an exposure through a bluish
filter (with "indoor" white balance selected, in my case), then, on
the computer, do some Photoshop operation to restore the correct
balance of red, green, and blue.

Is this this reasonable? What do you recommend?

===Robert

 

[Robert63289]

Thanks for all the great comments.\! I'll report the results of further experimentation.

And yes, I am delighted with the F707. Being a physicist and and astronomer, I simply want to understand the light-camera-computer-printer combination so as to do the best I can capturing nature--in this case, orchids.

Robert

 

[Errol Gibbs]

Last week Ed looked at my F707 orchid pictures and called my
attention to the magenta/plastic/flourescent/neon rim at the
intense red edges of my orchid lips. I realize now I did not
understand the problem--and still don't.

I invite discussion of this 'magenta' effect, most recently
evidenced by this F707 picture of my blooming mini-Cattleya, (Blc.
Jane Helton x Slc. Vallezac) 'Kudo's Golden Dragon' x Slc. Kauai
Star Bright 'Walter':

http://www.pbase.com/image/450557/large

This fringe is not on the plant.

In the "neon" lip edging, the histogram of red values is crowded to
the top, but it does does not look cutoff at 255.

However, the green and blue are obviously under exposed in the
histogram.

Is that the problem? Perhaps an exposure does not capture enough
blue-green information before the red channel tops out and the
exposure ends? If so, is a blue filter indicated? I could imagine
there is a calibrated approach: take an exposure through a bluish
filter (with "indoor" white balance selected, in my case), then, on
the computer, do some Photoshop operation to restore the correct
balance of red, green, and blue.

Is this this reasonable? What do you recommend?

What is wrong here?
These pictures are fantastic!
Muey Bueno! Outa Site Man!
.................Hey they are good....er great!
PREEN yourself man you deserve it!
Errol

 

[Gregory Alter]

Really an amazing thread in terms of the technical sophistication of the comments. It takes a lot of picture taking to get to the point of even suggesting all these possible reasons for color aberations, if we are to call them that. Thanks to all...

greg

Last week Ed looked at my F707 orchid pictures and called my
attention to the magenta/plastic/flourescent/neon rim at the
intense red edges of my orchid lips. I realize now I did not
understand the problem--and still don't.

I invite discussion of this 'magenta' effect, most recently
evidenced by this F707 picture of my blooming mini-Cattleya, (Blc.
Jane Helton x Slc. Vallezac) 'Kudo's Golden Dragon' x Slc. Kauai
Star Bright 'Walter':

http://www.pbase.com/image/450557/large

This fringe is not on the plant.

In the "neon" lip edging, the histogram of red values is crowded to
the top, but it does does not look cutoff at 255.

However, the green and blue are obviously under exposed in the
histogram.

Is that the problem? Perhaps an exposure does not capture enough
blue-green information before the red channel tops out and the
exposure ends? If so, is a blue filter indicated? I could imagine
there is a calibrated approach: take an exposure through a bluish
filter (with "indoor" white balance selected, in my case), then, on
the computer, do some Photoshop operation to restore the correct
balance of red, green, and blue.

Is this this reasonable? What do you recommend?

What is wrong here?
These pictures are fantastic!
Muey Bueno! Outa Site Man!
.................Hey they are good....er great!
PREEN yourself man you deserve it!
Errol

 

[Perry]

Here is another example. It seems to be a red saturation problem. For some reason when the red saturates it turns magenta.

Perry

 

[Robert63289]

My investigations show it has nothing to do with camera sharpness setting or JPG. It is present with S=-2 and format = TIFF.

For those of you admiring my orchids, here are some new findings:

TIFF format does show the benefits of higher information content than JPG. I can see more detail on them at high sharpness.

I find in-camera sharpening superior to unsharp mask in PhotoShop.

I am currently using S=2, shooting in TIFF, and doing no postprocessing whatsoever. Latest result:

http://www.pbase.com/image/461188

===Robert

 

[Perry]

Try bracketing the exposure? I bet you will see a dependence there. Great orchid photos. You must be a proud papa of the orchids and the camera.

Perry

 

[Frank65645]

called attention to the magenta/plastic/flourescent/neon rim at the
intense red edges of my orchid lips.

The problem is due to the limitations of how we are representing color, Sony's choice to give you saturated colors, and the white balance program in the camera. The regions that go magenta are somewhat brighter, and perhaps a little less saturated than the darker red areas next to them. These are already maxed out in the red, but the blue green values are down around 0 to 15 - almost non-existant. The only way a brighter area can be shown as brighter is by bringing up the blue and green - and by comparison, this goes away from red to magenta! You can duplicate the effect with a paint program, put a "lighter red" - say 255 66 66 - next to a dark red - your image was about 250 0 16 - and the light red looks magenta.
Next to that red, a gray looks blue.

Frank

 

[Robert63289]

Frank, is there a remedy? Could a bluish filter be used, holding back the red a bit, which could then be restored in PhotoShop, playing with the RGB color channels to remap the linear RGB data from the camera to a truer representation.

Are such calibrated, filter-software combinations available?

===Robert

 

[Mike Fitzgerald]

Yes, the standard PS unsharp mask has its limitations, and there are better ways to drive Photoshop to this end.

But this is an extraordinarily good result you're getting, through exploring uncharted territory. The general feeling has been that "Standard" (0) sharpening in the F707 is excessive, with people grizzling about visible halo. So much so that I don't think anyone's bothered to even try maximum in-camera sharpening with it before. (Well, not posted or discussed the results, anyway.)

However, there's no obvious suggestion of artificial sharpening at all in this image, so I wonder if what people have been seeing in this regard has been a combination of sharpening and JPEG artefacts. Certainly the greens have a somewhat variegated appearance in places, but I'm assuming this is natural. Am I correct?

I'm sure there are a few of us who'd welcome a look at the JPEG of the same shot, if you still have it. BTW, what are the approximate dimensions of this image, and the shooting distance?

Thanks for your input, Robert. You're doing much to unleash the true capabilities of this camera.

Mike

My investigations show it has nothing to do with camera sharpness
setting or JPG. It is present with S=-2 and format = TIFF.

For those of you admiring my orchids, here are some new findings:

TIFF format does show the benefits of higher information content
than JPG. I can see more detail on them at high sharpness.

I find in-camera sharpening superior to unsharp mask in PhotoShop.

I am currently using S=2, shooting in TIFF, and doing no
postprocessing whatsoever. Latest result:

http://www.pbase.com/image/461188

===Robert

 

[Robert63289]

Mike,

I only have a low-res 640x480 jpg file, because I was being way too clever. Someone had posed earlier that you could save MS space if you chose low resolution for the jpeg file when all you cared about is the tiff. Penny wise, pound foolish. Anyway, here it is:

http://www.pbase.com/image/462188

The orchids are 8" wingtip to wingtip.

Specs: ISO100, manual exposure, 1/13 sec, f8.0, focal distance 10mm, exposure correction zero.

===Robert

 

[Ed (;¬ >)]

Robert, you continue to make my jaw drop. The detail of these images is considerable. I am amazed that the cilia on the leaves which are black on black are for the most part visible.

If you are using the undifferentiated Unsharp Mask option in Photoshop, I'm not surprised that you prefer the in camera option. If you continue to use TIFF I suppose there's no harem done. If you decide to use JPEG, I would urge turning off (-2) in camera sharpening.

In Photoshop I recommend using the Unsharp Mask Dialog so that you can adjust the parameters for each picture.

-Ed

My investigations show it has nothing to do with camera sharpness
setting or JPG. It is present with S=-2 and format = TIFF.

For those of you admiring my orchids, here are some new findings:

TIFF format does show the benefits of higher information content
than JPG. I can see more detail on them at high sharpness.

I find in-camera sharpening superior to unsharp mask in PhotoShop.

I am currently using S=2, shooting in TIFF, and doing no
postprocessing whatsoever. Latest result:

http://www.pbase.com/image/461188

===Robert

 

[Mike Fitzgerald]

Robert --

Thanks for the info. Certainly nothing obvious in terms of sharpening artefacts on the JPEG either, even though it is low res. S+2 is looking far more useful than we've been crediting.

For your general information: For much closer macro work the 707 produces a lot of CA at the edges, as well as some barrel distortion. In case you haven't been looking at related threads, the Canon 500D supplementary lens is expected to work very well on it. Just something to keep in mind if you have the need...

Mike

Mike,

I only have a low-res 640x480 jpg file, because I was being way too
clever. Someone had posed earlier that you could save MS space if
you chose low resolution for the jpeg file when all you cared about
is the tiff. Penny wise, pound foolish. Anyway, here it is:

http://www.pbase.com/image/462188

The orchids are 8" wingtip to wingtip.

Specs: ISO100, manual exposure, 1/13 sec, f8.0, focal distance
10mm, exposure correction zero.

===Robert

 

[Robert63289]

Thanks Ed.

My experiments with the unsharp mask used only the verson with dialogue box to set amount, radius, and threshold. Except for Paph Ianthe Stage, my orchid shots were shot with s0 and sharpened with USM settings 125/3/0. In my experments yesterday, I rambled through A/R/T parameter space on pictures taken with s-2 and never could approach the detail evident in identical shots taken with s+2.

===Robert