Here's the setup I've been using to photograph identification details on live spiders without harming them. It's an evolution of the palp setup
I posted before. It's easy to do and doesn't require anything expensive.
The trick is to safely immobilize the spider. I put it on several layers of tissue paper and cover it wholly or partially with a transparent restraint. For small spiders, I use a microscope slide. The clear half of a CD case works decently for larger ones. I also keep a sheet of flexible plastic, such as a cutting from a bag, directly under the spider. It's smooth so the spider won't get its leg spines caught as it might with only the tissue. See the next three images for more info.
My lighting setup changes often, but right now I'm using a flash from below coupled with a reflector atop the restraint (the cardboard square pictured above). The reflector secondarily blocks excess light that would otherwise wash out the image. There's a small hole in it which I shoot through, positioned over the desired area.
Changing reflector materials will affect how the surface of whatever detail you're photographing appears. For palps I like the cardboard square because it's very weak and doesn't produce any harsh highlights. The back side of photo printing paper works nicely for most epigyna.
The choice of camera and lens isn't really important so long as you get plenty of magnification! I've had luck with reversed lenses on DSLRs. Telephoto lenses with close-up attachments work too. I'm currently using a superzoom (Canon SX40 HS) with an old 50mm prime reversed in front. Depending on the size of spider you're working with, the macro mode on a point and shoot might even do the trick.
Here are some examples of what you might be able to do with the spider in different positions.
On its back, facing belly up, the epigynum is the most obvious thing to go for.
Ventral views of the palps are also best done from here. You may need to straighten the palp with a wire (see here
The distal palp view is also useful for some spiders. Linyphantes in Linyphiidae, for example. This is very easy to do since in most cases the spider's palps will be pointing straight up when it's on its back anyway.
Lots of times the chelicerae will be in the right position to get a good look at the teeth on the posterior margin.
If you've only restrained the cephalothorax (like this
), posterior epigynum shots will also be possible.
Or the anterior cheliceral teeth if at least part of the cephalothorax is free.
With the spider on its side, you can often get retrolateral palp shots without any extra work.
Same with the epigynum.
Also leg details like spines, trichobothria, or tarsal claws.
You can simplify things by using only direct light, such as an on-camera flash. It may take some fiddling to figure out where to position it to avoid reflections from the restraint.
I've run into some special cases that require tweaks to this setup, but not many. I'll add examples and workarounds as time permits.