archaeometallurgy

There are a number of Linux software solutions that do tethered shooting, aka remote capture and they do it well. In this article I am describing the difficulties of setting up darktable for tethered shooting with a Canon EOS 600D digital camera on Ubuntu 12.04. Plenty of people managed to do this and I meant to do the same. I was quite surprised when it did not work the way it should have worked.

I am using a Canon EOS 600D digital camera as a microscope camera, because I can put its liveview functionality to very good use. Either to teach or to produce good quality micrographs. If you are using Mac OS X or Microsoft Windows you can use the Canon EOS utility to use tethering to control a digital camera with your PC. However if  you like to – or have to – use a brand of Linux you will find that the Canon software does not run with Wine. Of course you could run Windows or OS X on a virtual machine, such as Virtualbox, but that became too cumbersome for me.

The image shows the darktable interface for tethered shooting. With Canon Liveview cameras the computer may be used to control the camera and take pictures. The image shows a section of an etched copper roof from the 18th century.

My enthusiam for tethered shooting on a linux machine was beginning to wane when I connected my camera to the USB port. All I got was a generic error message:

Unable to mount Canon Digital Camera Error initializing camera: -1: Unspecified error

The camera was not  mounted. I could not even get it to work with digiKam to download images. I did check for SD card corruption, updates for all sorts of related software libraries, tried to access the camera with  gphoto2. …and nothing did work. Eventually it turned out that the USB 3.0 ports did not work well with the camera. After figuring this out I did find this post which seems to confirm my findings. Plugging the camera into the USB 2.0 port  the cameras was mounted as a storage device, however now Ubuntu issued the following error message:

Error (-60: 'Could not lock the device')

After checking for this error I found this post, which solved the problem:

• make sure your camera is connected to a USB 2.0 port
• unmount the camera with your file browser
• start darktable, switch to the “Tethering” tab
• the software should be able to find your camera

Addendum 9.11.2014

If you remove the SD card from your camera darktable will work out of the box as long as you start after you have plugged in your camera. Where there is no storage device the gvfs cannot mount it. Hence the camera is not blocked. Give it a try. 1. If If you are looking for more permanent solutions check this post: Delete or rather rename the following files, so you can undo things if they go wrong:

/usr/share/dbus-1/services/org.gtk.Private.GPhoto2VolumeMonitor.service
/usr/share/gvfs/mounts/gphoto2.mount
/usr/share/gvfs/remote-volume-monitors/gphoto2.monitor

and optionally:

/usr/lib/gvfs/gvfs-gphoto2-volume-monitor

2. Try this solution: remove the execute rights of the following file:

1. open terminal
2.  command: sudo killall gvfs-gphoto2-volume-monitor
3. command: sudo chmod -x /usr/lib/gvfs/gvfs-gphoto2-volume-monitor
4. close terminal and restart.

These two solutions have the same effect, because they disable the same feature: It prevents gvfs from mounting your cameras SD card as a storage device. So unless you are happy with the fact the camera will not appear in your file browser any more, I would recommend the other solutions. Photo managing software should still be able to find your camera, though.

Well, I hope this helps. Let me know if you found any other solutions.

Do you work with a lot of Exif data or other meta data, such as XMP or IPTC? Storing Exif data in your images is a good way of keeping information about your photos together with your photos or micrographs. You do know the problem: all EXIF data is suddenly lost, the micrograph is now almost useless without it. This is especially true if you use ImageJ to measure features in your micrographs. However if you were working on copies instead of the originals, as you should always do, all is not lost.

The all mighty Exiftool provides a simple solution:

exiftool -tagsFromFile originalimage.jpg anotherimage.jpg

The  -tagsFromFile  option copies all metadata fom the original file to another image file.

If you are using some sort of linux installation of exiftool is done with your package management system. For OS X there is a dmg package, for WIN there is an executable available on Phil Harveys website.

Don’t get me wrong, ImageJ is a great piece of software and I use it on a daily basis and will continue to do so. I do hope that at some stage in the  future this Exif data bug (or is it a feature?) will be fixed and make this current workaround obsolete.

Sand moulding, or rather the casting in naturally bonded sand can be traced back to 17th century AD originating in France. Modern casting industry’s demand of constant quality of casting sands led to a decline in the use of  sand which is bonded naturally.  A few art foundries in Germany, though, are still knowledgeable therein and use it for casting purposes products. Due to the high expenditure of human labour it is only a question of time until all art foundries will work with the lost-wax process.
The main difference to modern casting sands are the utilized binders. In naturally bonded sand this is usually a percentage of clay. In industrial sand they are Bentonite (mainly a clay mineral called Montmorillonite), Furans, Phenolic resin or similar chemical compounds. Since this page is concerned with archaeometallurgy we will dismiss all the explanations on modern casting technologies.

The stop motion animation shows the fabrication of a green sand mould with naturally bonded sand. It is a reconstruction of a Zeiss Junior microscope foot. This mould was made for aluminum casting.

Moulding with naturally bonded sand

Since sand on its own is not enough to produce a sand mould, a frame for holding the sand in place is needed. In foundry terms this is referred to as a flask. Flasks are usually rigid frames without fixed bottom or top and consist of two parts: a lower one also known as drag and an upper one which is called cope. So called flask pins at the sides of the cope and drag ensure a proper alignment of upper and lower half during moulding and later, once the pattern is withdrawn, during pouring.
The sand is moist or green when it is compressed into the flask, and has to be dried in a kiln before pouring. Does sand really stay in the mould!? Two processes are responsible for the stability of the sand once it is compressed. (a) The dilatant nature. (b) The sand has 8-15% clay as binder and up to 5% of water contents.
A simple test is to compress a handful of sand and throw this about half a meter into the air, and if you can subsequently catch it as a whole, you might just have a good moulding sand. There are other important criteria to be matched: Permeability, refractoriness and a fine grain for surface smoothness. Fine quartz sand (or rather coarse silt) fulfills all conditions perfectly, if there is a clay contents between 8-15 wt%. Moulding with naturally bonded sand is easily one of the most fascianting techniques, because of its simplicity and its apparent defiance to obey the law of gravity. Once you have seen a flask -a metallic frame that has no bottom and no lid- being turned without the sand dropping out of it you will surely believe me.

Moulds, patterns & co

Best suited for sand casting are simple patterns without too many undercuts. Because the pattern needs to be withdrawn to create the mould cavity, the flask consists of two halves. To mould a pattern you first have to decide were the parting of the mould is. The parting is that line, where the mould cuts the cavity in two halves. It is evident that neither side can have undercuts, or you cannot withdraw your pattern.
First we create a false half, which holds the pattern in place: The pattern is buried up to the parting line of the pattern in sand. The sand is compacted and smoothed to produce a nice parting surface arounf the pattern. Then you powder this surface with a parting agent, to prevent the green sand sticking to the model or the parting surface. Put on the cope, put finely sieved sand on top of pattern and parting surface and compact it carefully with your hands. Put more sand into the cope and compress it with tampers and mallets. The whole flask is now turned, and the false half (drag) removed. It can be discarded and is shaken out. Put the drag back on, and repeat the steps previously described for the cope. After the drag has been made it is removed again, however, this time carefully, as to avoid damage to the mould cavity (You might want to lift the drag by pushing wedges into the four corners of the flask and thus to lever the drag slowly and in a controlled way from the cope. After that you tap the pattern and withdraw it carefully. Now you have produced the mould cavity. Downsprue, sprues and vents have to be cut into the sand, so the metal melt can reach the cavity. After drying the mould in the kiln it is ready for pouring. Cope and drag are reunited, clamped and metal is poured in. After casting shake the sand: your first cast!
What happens with more complicated patterns, e.g. sculptures? Can you cast them in sand? Yes, you can; if I did not manage to confuse you as yet, you may read further on the Stückform process here.