Part of our excavation strategy at Tall Dhiban is to “float” a sample of dirt that we excavate. So up on the site we collect about 30L of dirt out of the context we are currently digging up and send it to the lab house, where Alan (and company) diligently processes it through a flotation tank. The flotation tank separates out what we call the light fraction and the heavy fraction–basically particulate matter that either floats or sinks, and tries to get rid of all of the dirt around it. Then he sorts the heavy fraction (usually lots of rocks and small pot sherds) and looks at the light fraction under a microscope. Then he identifies the different seeds and such (this is where I get a bit hazy–I’m not a paleobotanist, though I know the mechanics of it) and is able to talk about the paleoclimate, what people might have been eating, seasonality, that sort of thing. He then takes microscopic photos of the more interesting bits and uses these photographs for publication. I was asked to “clean up” these photographs for publication.
So all of this is a large introduction to a fairly basic blog post. I have edited artifact photographs for publication before, but I upgraded from photoshop CS3 to CS4 this Fall and my settings were all wonky and I forgot some of the basics. I’m blogging about it so that I can show folks what I do to clean up these photos and to get feedback on better ways. Back when I posted about photo scales, Jason popped up and was able to give me much better scales that I use all the time.
So, anyway, here is one of the photos I started with:
Perfectly acceptable and lots of publications might use it as-is. I quite like it as it shows the curve of the microscope and emphasizes the shape of the seed with the shadows. But some of that detail isn’t as pertinent in black and white publication, so let’s simplify it.
I usually begin by making another duplicate layer in photoshop so that I have a back-up, something I can use for reference later on. Then I create two new transparent layers — one for the background that I “fill” with white, and another that I’ll use to paste the seed onto after I cut it out from the photo.
I tried a couple of different ways to erase the background–people have different preferences in this respect. I usually just rely on the pen tool to trace around the object, moving the points closer in to the object or further away. People have success with the masking tools in photoshop as well, but I don’t feel like I can control the paintbrush quite as well as the pen tool.
Once you extract the image with the pen tool, you can paste it on one of the transparent layers. Then I like to hide the original background images so that I can see how well I did. This seed was a bit hard because the aperture made the edges fuzzy and hard to disambiguate from the shadows.
But, you say, what about the scale?! Yes, archaeological photographs have scales in them so that you know that this seed is not the size of a bowling ball.
I like to make digital scales, and they’re pretty easy–either have some ready made and scale them into the photo, or quick-make them. Either way it’s about the same about of time, unless you do something fancy. I usually just make them from the reference image out of a couple of filled rectangle boxes and the font tool.
So the digital scale is just a transparent layer on top of the original image and I can move it around however I like. I was thinking about omitting the “1” in front of the millimeter as it seems obvious.
I also left in most of the screen on this shot so you can see all the different layers. I usually merge the scale layers together so I can move it around more easily.
The final step is filling one of the background layers with white, then making the original photographs invisible, saving a photoshop copy (with layers) if I want to mess with it later, then collapsing the layers into a tif file. This can be cropped (if need be) and send to the publisher:
The background erase on this one is a little wonky, but it gets the job done. I also didn’t adjust the light levels or anything on the seed itself because I wasn’t sure what would be the most helpful to the other paleobots looking at this thing. Some people like to add drop shadows, but I didn’t bother. It occurs to me that I have no training in microscope photography of any kind, how disturbing! I’ll have to remedy that.
Anyway, after you do a few dozen of artifact photo fix-ups (it’s even worse when you have to put multiple artifacts into one figure that have been taken at different scales and with different backgrounds) you begin to understand the value of taking a good shot with a good scale and nice light. I recently saw a photo of an artifact on a blue towel–the horror.
Anyway, I’d be happy to hear any tips from people who have more experience in the matter and are willing to share!
My apologies if this was partially inarticulate–I just lectured on Ethnography, Hitchcock & film and Berger/Sontag/Barthes for three hours and I’m feeling a bit wiped out!
3 thoughts on “Photoshop for Archaeological Publication – Seeds”
Great post. I have done a ton of this with pots and sherds and you are so right that it really makes you appreciate a good initial shot with good lighting and contrast between the artifact and background.
The quick selection tool (added with CS3 and located in the toolbar with the magic wand) can really speed things up if the original shot is good. I sometimes use the various lasso tools to refine my selection.
A great tip you might try, if you are not doing this already, is once you get your selection defined pretty well and before you copy out to a new layer, right click and select ‘refine selection’. This will let you slightly expand or contract the entire selection to get those rascally fuzzy edges, and also feather your selection to help cut down on the hard lines you sometimes get between the subject layer and the new background.
The second figure on this page shows some examples of pots I have done, pretty similar to what you are doing:
The gradient backgrounds are nice for web or other full color content, but generally not so good for publication.
It was good to meet you on the Big Island. Hope we meet again soon!
I was going to say some of what he said, but I’ve learnt a tip or two from both post and comment. So that mainly leaves me adding sympathy with the blue towel problem. I do a lot of scanning and cleaning-up of coins—largely ancient or medieval ones without clean machined edges—and it’s amazing how many people think they should photograph them either on their hand or on a rich rumpled cloth background…
Very cool. Now I know the true story behind my seed photos!