Those who don’t care about the minutia of photography should probably back away now…

I’ve been doing all these fruit pictures and have been concerned that I’m loosing a lot of detail in tonality and color range. If that doesn’t make sense, basically, rather than the color looking smooth when transitioning from one tone to another, it looks rather blocky. This becomes a huge issue in these brightly colored subjects.

I think this is a problem of bit depth. My D3 is 14 bit, and even with those 4398046511104 (that’s 14 bits for each color, R, G, B, or 2^14*2^14*2^14) possible colors/tones, in intensely colored areas I’m loosing a lot of data. How do you fix this? Well, my first thought was to upgrade to a medium format camera. That bumps you up to 16 bit acquisition and 2.8 × 10^14 possible colors/tones. That’s a big number.

There are many reasons that I’d like to move up to medium format, including better quality pixels and the lack of an anti-aliasing filter (which is total lameness, the AA filter, not the lack of one). What’s the problem? Mostly money. I don’t have the bux. So the question is, until I can scrape together enough money to upgrade, how can I fix this problem?

I’ve played with high dynamic range (HDR) photography, but I’m not really a fan in general. Even subtly HDR’d pictures look a bit fake. One thing HDR does give you, is a much wider tonal range. The raw HDR actually gives you 32 bits!!! Wholly crap! That’s 7.9 × 10^28! Exclamation point! That’s probably excessive, but the compressed HDR image (which is a topic I won’t go into now, but is interesting in and of itself) is 16 bit, which is what I would like to have. Now if only I can do this without making it look like a cartoon.

So, many moons ago I bought a software package called Photomatix Pro (www.hdrsoft.com). Photoshop can assemble HDR images, but photomatix is the go to software for not making HDR that looks like crap. Awesomely, the registration code that I have from eight-hundred versions ago still works. So hey, I’ll use it. Thanks HDRsoft.

As an OCD scientist, I felt the need to test this and make sure that I’m not just making things up, and that these extra 2-bits are actually full of something, and not just extra space. So here’s what I did:

I set up my usual home made light box, but modified to be dark-field (I’ll explain that another time) because it’s cooler, with a slice of orange as my subject. My camera is on a big-ass carbon-fiber tripod with a monstro really-right stuff (RRS) ball-head, plus an RRS focusing rail, all pointed down towards the subject. The camera is leveled, and the sample stage (a piece of mostly clean glass) is leveled as well (since my floor is horizontally challenged). I used f16 to make sure that I don’t get too much diffraction softening, and then I took 4 exposures 2/3 stop apart.* Normally, you’d expose an HDR image with 2 stops between exposures, but I don’t need the dynamic range. All I want is slight changes in tone to fill in all those extra bits.

A quick caveat: If you want to take pictures of people, or anything that moves in any way, and you want awesome tonal range, buy a medium format camera.

Above you’ll see a figure I made (sorry, I’m a scientist, I make figures… ) showing one of the properly exposed single exposures (top left) and a compressed 4 exposure HDR (top right). What were the settings for the HDR compression you ask? It’s not very interesting. The short answer is I played with sliders until it looked like there was a lot of detail, and didn’t look fake. My guess is that these settings will be different for every image, so if you try it, my settings won’t help you anyway.

On the bottom of that figure are 100% zooms of the above shots. What I see here is that the HDR’d image has quite a bit more detail, even though the dynamic range of the image isn’t substantially different. That’s awesome! Go me! The intensely orange areas on the left image are rather blocky, the ones on the right are much smoother, and have much more detail to my eye. An added bonus is that the image looks much sharper to boot. That’s probably because the blockies make it look unsharp.

In summary, rock on! It’s kind of a pain in the ass, and requires a lot of photoshoppery, especially when you consider that some of these fruit images are made up of 40 or 50 individual shots (3 exposure HDR * 40 frames = ass ton of pictures for a single image). So it goes. It’s the price of awesomeness I guess.

If you got to the end, I give you my undying gratitude and a seal of approval.

*Extra bonus details:

Flash from below – Profoto D1
Camera – Nikon D3
Lens – Nikon 105 2.8 macro (the older version, not the new hotness)
Black background – Black velvet (sans Elvis and poker playing dogs) from Amazon

“These days, China is lavishing money on Mr. Science. But without the checks and balances provided by Mr. Democracy, the corruption plaguing the rest of the system is infecting the reputation of Chinese science.”

I review many articles from China and “text reuse” is a common theme I’ve seen. I think it’s partially due to the language barrier. It’s much easier to copy text then re-synthesize and translate new.

The amount of money being spent on science in China currently must also be placing huge pressure on scientists to perform. That generally doesn’t end well either. I hope they can figure out the regulatory part soon. All this money in science is somewhat rare these days and it would be a shame to see it wasted.

At any rate, an interesting read:

http://www.npr.org/2011/08/03/138937778/plagiarism-plague-hinders-chinas-scientific-ambition?ft=1&f=1007

last

• 
• 
• 
• 
• 
• 
• 
• 
• 

next

I’m working on a new project these days. I’ve been thinking for a while about how to properly show this on the intarwebs, because it isn’t the kind of thing that can be fully understood by looking at low res images in a browser. What I’m doing is making extremely high resolution images of cross-sections of foods. The first image in the gallery above is of a persian cucumber. Each subsequent image is a zoom in on that original image.

I’m trying to give a feel for how the printed image will look. The idea is that the images will be printed very large, so from across the room the overall image will be clear, but by coming closer fine details of the food’s structure will be revealed.

I like to think of it as combination of my two favorite things, photography and science.

More on these to come.