Darks with scales, flash vs. continuous

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

johan
Posts: 1005
Joined: Tue Sep 06, 2011 7:39 am
Contact:

Darks with scales, flash vs. continuous

Post by johan »

I have been trying to image a very beautiful Papilio Ulysses but I'm running into dynamic range issues and I'm wondering if I'm on a road to nowhere, so wanted to ask some advice. Papilio Ulysses is a beautiful blue specimen with brown patches in between, comprised of elongated scales. This is my limited experience and I wonder if I should just give up on trying this flash lighting:

With flashes on 1/32 and flash shutter speed if I expose nicely for the blue scales I lose more or less all detail in the browns. If I expose nicely for the browns the diffuser reflection in the highlight is blown out. This is using two flashes permutated in various ways, both headon and obliquely.

The weird thing is if I use continuous lighting I seem to be able to do much better in terms of preserving the range. The browns oranges and yellows in the darks show up but the blues still have that dusty blue that I was looking for the and highs arn't blown out. So a much flatter outcome, on about 2s of lighting.

I can post the output if you like but I didn't think it worth posting :)

So I wonder if this is just one of those learning experiences about light sources and if this is something that's well known. What is it about continuous light that makes it so much flatter/able to show that range. Is it something about its accumulative nature?

Many thanks
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.

johan
Posts: 1005
Joined: Tue Sep 06, 2011 7:39 am
Contact:

Post by johan »

Forgot to add, I'm using JPEG for this. So I suppose I could try doing RAWs and making 2 outputs from a RAW set, one for the darks and one for the highs... hmmm...
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.

johan
Posts: 1005
Joined: Tue Sep 06, 2011 7:39 am
Contact:

Post by johan »

Could I also ask, is there a sensible way people know of to mount butterfly scales vertically?

Thanks
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

I'm running into dynamic range issues
This is not a well known problem, and it's not anything that I've studied personally. However, given your observations I'm willing to speculate on the causes.

The sensor in your camera may be balanced so that its RGB photosites give roughly equal values for a gray subject illuminated by some color temperature in the middle of the usual range from incandescent to daylight.

But now you're asking it to image a brilliant blue subject illuminated by light that is also heavy in the blue. If you're really unlucky, there's also a metamerism effect in which some blue peak in the illumination spectrum happens to line up with a blue peak in the reflection spectrum so the sensor sees BLUE! Anyway, my guess is that if you looked at the raw sensor data, you'd see that when the brightest blue photosites are just about to saturate, the red and green photosites are still way down in the noise. Of course the RGB channels get balanced out again by scaling during the raw conversion process, but there's only so much you can do to repair data that starts off degraded.

One way to explore this issue would be to actually look at the raw data using dcraw with the -D option.

In terms of fixing your problem, I definitely would recommend shooting raw and converting it on your computer where you have a lot more control over the options and where the software has a lot more time to get it right. It's been several years since I looked closely at the tradeoffs, and on only one camera, and I never wrote up the results, but if memory serves correctly the main difference was that in-camera JPEGs (Canon 300D) had lousy color rendition in the dark areas compared to what Photoshop's converter would give from a CRW shot at the same time. Even when saved to the same size JPEG file in the end, Photoshop's version from raw had strikingly better color in dark shadows. I had to apply a strong levels adjust to see the colors at all, but when I did that, the camera's colors were streaky, noisy, and all over the map, while Photoshop's colors were fine.

A second thing you could do is filter the flash and/or camera down to something closer to incandescent. If you don't have proper photographic filters available, you might try using colored anything as either diffusion or reflectors. If you use custom color balance, the raw conversion process can compensate for some pretty bizarre spectra in terms of rendering a gray target. It can't do much about metamerism, of course.

Let us know what happens, please.

--Rik

johan
Posts: 1005
Joined: Tue Sep 06, 2011 7:39 am
Contact:

Post by johan »

Thank you Rik

I had to Google some of the terms and software as I wasn't familiar with them (new word of the day metamerism - http://en.wikipedia.org/wiki/Metamerism_(color) )but I have to say I'm bemused by my current 'scales' explorations. I thought I had an ok grasp of technique but this is throwing up some unfamilar challenges. Possibly it's also a consequence of setting a higher standard for myself too. That can only make for improvement but I havn't been able to explain some of it yet :). Though, maybe my challenges arn't so odd, what you're saying about excess blue makes perfect sense - I had to set my cam WB at 5800degK and very far into green/orange to make the blues look anything like normal on my 4800degK flash, very unusual.

I do have some small flash-sized gels around somewhere I think, I'll have a tinker with that. The other options I thought of was using some bracketing to create various JPEG sets at different ISOs, or some extended bracketing with different high/low key settings, doing RAWs and generating different sets from them, or exposing for the lows and masking off the diffuser to make the highs managable. I usually use CS4 anf the RAW processor is nice, I do prefer it to the one shipped with the cam (silkypix). I've just always tried to avoid going into RAW or TIFF files more for processing time reasons than anything else. But I'll probably have to bow to the inevitable.
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Some while ago it occurred to me that those startling blues might coincide with the response of our eyes' cones. They don't!
We think of "blue" as about 450-495nm. P.ulysses is slap in the middle at 475, but the relevant cone's sensitivity peaks at 440nm, so we must be using some Green (~540nm) to evaluate it.
I don't know if there are standardised vales for camera sensors these days but the article below alludes to sensor differences, and more in the display monitor's primary colours, as well as in the algorithms which can be used to process the numbers. (Getting the hues from the chromaticities, if I have the terminology right.)

Plenty of scope for seeing things differently. My eyes are very slightly different from each other, too. In case of lingering doubt, ask your wife... :)

Chromatic Adaptation performance of different RGB sensors
Anatomically diverse butterfly scales all produce ...

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

ChrisR wrote:In case of lingering doubt, ask your wife... :)
Heh -- indeed! http://www.huffingtonpost.com/2012/06/2 ... 31480.html

--Rik

johan
Posts: 1005
Joined: Tue Sep 06, 2011 7:39 am
Contact:

Post by johan »

:smt119 You guys have completely lost me. Except for the fact that my wife is always right, which I knew anyway :)
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

johan wrote:You guys have completely lost me.
Sorry...sometimes we get too playful.

Quickly summarizing the papers linked by ChrisR...

"Chromatic Adaptation Performance of Different RGB Sensors" evaluates the ability to convert colors captured under one illumination to colors as they would appear with a different illumination. It's a very formal paper that presents a lot of data, but offhand I don't see how to apply it to the current problem.

"Anatomically diverse butterfly scales all produce structural colours by coherent scattering" basically argues that in butterflies all structural colors are produced in the same way, "by coherent scattering, i.e. differential interference and reinforcement of scattered, visible wavelengths". For the current problem, it just confirms that the color coming out was extracted from the source illumination without changing its wavelength. That being the case, it doesn't matter whether you stick a filter over the flash or the camera.

The "Tetrachromatic 'Super Vision'" article notes that some humans actually have four different pigments in their eyes. In theory this possibly could provide those people with color discrimination that is fundamentally more powerful than we three-pigment folks can do. There's an interesting history here. In 2001, there was a paper titled "Richer color experience in observers with multiple photopigment opsin genes" (Jameson and Highnote) which experimentally demonstrated that "Women with four-photopigment genotypes are found to perceive significantly more chromatic appearances in comparison with either male or female trichromat controls." However, the finding was based on the number of chromatic bands that each subject labeled within a conventional spectrum. Unfortunately enthusiasm for placing labels is not the same as demonstrating the ability to discriminate colors on more than three axes, and direct experimental evidence of that latter capability has been surprisingly difficult to find. Over 10 years later, one subject has finally been identified who definitely seems to have it. It remains to be seen how many other people can do the same thing, and to what extent, and what the practical significance might be. With respect to the current problem, it just means that our wives might be even less happy than we are with the weird colors.

--Rik

johan
Posts: 1005
Joined: Tue Sep 06, 2011 7:39 am
Contact:

Post by johan »

Thanks Rik, appreciated! Phew

I'm kind of glad I got stuck on this. It made me have another look at my setup and I've found that by removing one element in the chain I seem to be able to remove my vibration problems. Which may mean, fingers crossed, that with an additional adaptation I can start using continuous for the first time, which would make me very happy!
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

The refererences given were just that - sources for detail I quoted. Only part of each paper was relevant to what I did state I was commenting on - the particular blue of the butterfly in question. If it had been exactly the same wavelength that a camera sensor blue channel sees, that could have been part of the exposure problem. "Camera sees "BLUE" ".

One reference provides the wavelength for the particular butterfly and the other mentions some sensor and monitor R, G and B peak spectral wavelengths. Its main subject is how different methods can be used, which can lead to different hues, from measuring the three primary colours. This gives scope for a given system to come up with a different result than your eye - leading to a metamerism issue.

It's usually from reading around a subject that I make progress, rather than the rare case of finding an article that gives a direct answer!

Charles Krebs
Posts: 5865
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

I always like to see some visual examples of the problem. I would like to see how the lighting was set up for flash versus continuous light. What continuous light sources were used? What is their color characteristics? Also, I find it much easier to avoid problems by adjusting the position/reflectors/diffusers with a relatively tight continuous light source compared to trying the same with a flash unit. I would like to eliminate that possibility before getting too involved in a continuous light/flash dynamic range quest. (Although brilliant blue iridescent scales with a high color temp electronic flash does raise a few questions).

With that butterfly, if a shot is composed with the iridescent blues scales (illuminated to "show their stuff") along with some dark non-iridescent scales it seems entirely conceivable the dynamic range of the camera is being sorely tested. (Just do a Goole search for "Papilio Ulysses" and take a quick look at the "images").

Chris S.
Site Admin
Posts: 4042
Joined: Sun Apr 05, 2009 9:55 pm
Location: Ohio, USA

Post by Chris S. »

johan wrote:Forgot to add, I'm using JPEG for this. So I suppose I could try doing RAWs and making 2 outputs from a RAW set, one for the darks and one for the highs... hmmm...
Johan, in my experience with other prismatic-scaled butterflies on a Nikon D200 (which has less dynamic range than most of the more recent cameras), the use of raw files has been quite necessary for getting both prismatic scales and very dark scales properly exposed in the same image. But I wouldn’t think you’d need two outputs. What has worked for me is to shoot the image a little bit “hot”—that is, with highlights a bit off the right side of the histogram. Then in the raw converter (Nikon Capture NX2, in my case), pull the highlights down and boost the dark areas up, so that all the information is represented within a single TIF file. Not sure I’m explaining this clearly, but my point is to consider acquiring your images a two-step process: When shooting, work for maximum information capture in the raw file; when processing the raw files, work for maximum representation of that information in a TIF file.

It may be worth mentioning that I often make stacks with JPEGs, and such files can work well for a lot of subjects. But for subjects where the dynamic range exceeds the limits of a JPEG file (and a butterfly wing with both shiny and very dark scales is great example of this), a careful, raw-based workflow is (in my experience) necessary to capture and convey all the values from dark to light.
Could I also ask, is there a sensible way people know of to mount butterfly scales vertically?
If it’s a butterfly wing, I tape it to a microscope slide or a piece of aluminum. And I have a homemade gizmo for holding the slide upright in front of my horizontal rig. (Aluminum piece for the base, two strips of brass epoxied to it, bent at right angles to point straight upward; bobby pins (not sure if they are called this worldwide—bobby pins are inexpensive hair clips) epoxied to the brass to clip the slide into.

For individual butterfly scales, you can do something I tried just last night: Pluck a few scales off the butterfly with tweezers, drop into alcohol, agitate to get them separated and suspended. Use an eyedropper to drip some scale-laden alcohol onto a slide, and let it dry. Some scales will adhere nicely. Then mount the slide in the gizmo I described above, and shoot.

--Chris

johan
Posts: 1005
Joined: Tue Sep 06, 2011 7:39 am
Contact:

Post by johan »

Thanks gents, appreciate you all took the time to pop down your comments, I'll keep testing and post a shot when it's vaguely worth posting. On reflection maybe the sequence I shot was a little too under - I was trying to prevent the highs from being completely blown out. And this maybe killed the lows a bit too much.

Great little gizmo! I'm going to copy it :)
Chris S. wrote:I have a homemade gizmo for holding the slide upright in front of my horizontal rig. (Aluminum piece for the base, two strips of brass epoxied to it, bent at right angles to point straight upward; bobby pins (not sure if they are called this worldwide—bobby pins are inexpensive hair clips) epoxied to the brass to clip the slide into.
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.

johan
Posts: 1005
Joined: Tue Sep 06, 2011 7:39 am
Contact:

Post by johan »

Quite pleased with the scales on this one in terms of the light. Lighting the tips really makes scales shots pop for me. First reasonable effort with continuous light... but at ISO400 so hopefully better when I use 80 (lowest on my cam).

Image
Papilio ulysses by Johan J.Ingles-Le Nobel, on Flickr
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.

Post Reply Previous topicNext topic