Iridescence Artifacts

[Andrea B.]

The conical cells on petal surfaces of some flowers cause diffraction effects (iridescence and/or almost-specular reflections) easily seen in UV photographs as bright, "sparkling" patches on the petals which shift as the direction of the illuminating light is changed. Sometimes you can see little pixels of colour in iridescent areas different from the usual false-blue, false-yellow or neutral colours.

 

In the case of this Snowdrop flower (a Galanthus species), the very fine surface detail revealed by the shorter UV wavelengths combined with the diffraction from the conical cells has caused maze-like artifacts, probably a kind of moiré interference pattern.

 

Nikon D610 + Zeiss UV-Planar 60/4.0 + Convoy UV-Led

^{f/11 for 1/1.3" @ ISO-100}

^{This test photograph has not been color profiled. And my watermark is incorrect. Geez.}

 

Look at the crazy patterns here!

 

When the photo size is reduced, the artifacts are more difficult to see.

 

 

For this next photo, the diffraction effects were a bit more stable and didn't form too many artifacts.

Nikon D610 + Zeiss UV-Planar 60/4.0 + SB-14 UV-Flash (mod)

^{f/11 for 13" @ ISO-100, two flashes fired.}

^{This test photograph has not been color profiled.}

 

 

[Andy Perrin]

Love that, Andrea. Reminds me of the cinnamon sticks I imaged back here:

http://www.ultravioletphotography.com/content/index.php/topic/2471-ir-visual-uv-ug1-animation/page__view__findpost__p__18130

[Cadmium]

The 365nm Nichia LED light from a Convoy S2+ illuminates with a fairly monochromatic light, from my experience of it, the colors in your top example using the Convoy look more broad band UV to me.

See the stem of the flower at the left, how it is more yellow, a color I would not expect to see with the Convoy/365nm Nichia.

You shot that top shot with only two Convoy's for illumination?

 

The D610 has a weak or no AA filter (anti-aliasing filter, or low-pass filter), so I wonder if this 'iridescence' could be moiré.

Try this same shot with some other camera that has similar resolution, but uses a low-pass filter, keep the post processing minimal with both comparison shots,

and I wonder if you would be seeing the same kind of 'crazy patterns' of colors in close up crops.

D610's have some people complaining about moiré, and I wonder how such anomalies project into the realm of UV/IR converted cameras, maybe the problems become even worse.

Maybe not, but the subject may deserve closer examination given that all the newer and higher resolution Nikon cameras have no low-pass filters, and their certainly are complaints.

I keep reading posts from people who are complaining about moiré from their non AA Nikon cameras. I rather doubt this problem 'goes away' with full spectrum converted cameras,

in fact it may exacerbate the problem for all I know. Something to think about and look into,

because I for one am a bit hesitant and skeptical about converting these newer non AA Nikon cameras to UV/IR given the borderline problems with moiré.

Do I really want to convert a 3 grand camera to full spectrum that may have even worse moiré problems when used in full spectrum?

Don't such moiré problems defeat the the reason for having the higher sensor resolution that attract us to these newer and 'better' cameras?

Perhaps the resolution vs moiré factor was jumped on a bit premature of the resolution point at which moiré would not show up,

but I wonder if the idea of moiré will always be a factor when cropping at 100%.

Don't get me wrong, I am probably the first person that wants to jump on the D610/D810/D850 full spectrum converted non-AA band wagon, but show me that my fears are wrong,

show me that moiré is not even a worse problem with a UV/IR converted camera. Which is more a of a gimmick, a non-AA camera, or a camera with an AA filter to prevent moiré?

 

https://www.dpreview.../thread/3602255

[otoien]

Does not conversion to full spectrum (or UV/IR only) remove the AA filter anyway, as it is integrated with the UVIR cut filter? Also high resolution should mitigate the problem - high sampling frequency means that a higher signal frequency can be sampled before aliasing occurs.

[Andy Perrin]

Steve, if she is correct about that being iridescence, then it would split the light up the same way your diffraction grating experiments do. When you shine the convoy through the diffraction grating, what range of false colors do you image?

[Andrea B.]

You shot that top shot with only two Convoy's for illumination?

 

I only used one because I don't yet have any holders for them. ( I used to have some flash holders. Don't know where they walked off to.)

 

************

 

The D610 has a weak or no AA filter (anti-aliasing filter, or low-pass filter), so I wonder if this 'iridescence' could be moiré.

 

This is a converted D610 so it has no AA filter. You probably missed the part where I wrote that the patterns were probably moiré. :D As I write this, I'd say there is no probably about it. That is moiré in the first Snowdrop set above.

 

Whether a camera is converted or not -- or whether it has an AA filter or not -- we can get moiré "mazes" under certain combinations of high-frequency fine detail, diffraction gratings and illumination. The first Snowdrop set above shows you how petal conical cells can work to create moiré.

 

In my little humble op, moiré isn't found all that often. And when it is, there are work-arounds. Re-shoot from a diffferent angle, change the illumination, apply about a .25-.50 radius worth of gaussian blur before other edits, and so forth.

 

Iridescence, per se, is not moiré, but it contributed to the colourful effect above and likely made the moiré look "worse".

[Andrea B.]

Here is a nice visible shot from today. It shows the straight line (more-or-less) striations of the conical cells and some sparkles of iridescence. But there is no moiré because (I think) the longer visible wavelengths do not reveal the surface detail as much.

 

D610-bb + Zeiss UV-Planar 60/4.0 + Baader UVIR-Cut Filter + Flashlight

^{f/11 for 1/5" @ ISO-100}

^{This test shot was not color profiled.}

 

Unresized crop. (BTW, such unresized crops never look sharp, only detailed.)

 

[Andrea B.]

Here is the UV companion to the preceding. Again, quite a lot of iridescence in the bulbous area where the petals attach and some moiré on the petal surface.

D610-bb + Zeiss UV-Planar 60/4.0 + Baader UVIR-Cut Filter + Nikon SB14-UVmod

^{f/11 for 20" @ ISO-100 with 3 flashes}

^{This shot was, surprisingly, color profiled.}

 

 

[Cadmium]

Does not conversion to full spectrum (or UV/IR only) remove the AA filter anyway, as it is integrated with the UVIR cut filter? Also high resolution should mitigate the problem - high sampling frequency means that a higher signal frequency can be sampled before aliasing occurs.

 

The D610 and many other newer Nikon camera have no AA filter to remove, that is the point.

The AA filter is used in traditional cameras as a moiré preventative. Some newer models have no AA filters as a way of achieving sharper resolution.

So converting the D610 or not still has no AA filter.

Many people have complained about the absence of the AA filter in stock D610's and other such non AA models.

 

So the problem or tendency may not be worse when converted, but the problem is there to begin with, and so would remain if converted.

https://petapixel.co...with-no-brakes/

[Cadmium]

This is a converted D610 so it has no AA filter.

 

The D610 never had an AA filter. That is the point.

[Cadmium]

Hmm, perhaps I didn't understand your point.

So are are you saying that all full spectrum converted cameras have their AA filters removed when the camera is converted by default?

Thus all full spectrum cameras would be just as prone to moiré?

I didn't know that, I learn something every day I guess.

[Andrea B.]

OK, here's the thing --- even WITH an AA filter you can still get moiré. The AA filter just provides a bit of blur which can help prevent moiré most of the time but not always. The fears about moiré with AA-filterless cameras are somewhat overwrought in the typical photo forum.

 

And, yes, during conversion the filter pack is removed.

[Cadmium]

I feel so stupid. I didn't know the AA was removed with all conversions OK, never mind then. .

[Andrea B.]

oh now. It's no biggie. Besides which - if we knew everything, then we would be perfect. You don't want that do you?? :D