AP projects 2015
Posts tagged archaeological objects
While still being occupied with the processing of the CT-scans, the first impression was that it wasn’t as easy and accurate as we expected it to be. Maaike had reserved the Artec Spider for Friday for another project. We wanted to see if this scanning technique gave a better or different result then the CT-scanners. We were lucky that this was possible.
The Artec Spider is a scanner for small objects that captures really complex details and color. It has an accuracy up to 0.05 mm.
We scanned three objects:
- The comb, because this object is the only object that was scanned by the macro and micro scanner. Scanning this object with the Artec Spider will show the difference between these three scanners.
- Hermione, to see if the accuracy of details is better with the Artec Spider.
- Harry, to see if with this technique the break lines and repaired break lines will be visible
In the files from the CT-scanner the repaired break lines weren’t visible and the break lines were given an arched shape.
The objects and the scanner had to be moved/operated by hand, since the objects were so small. For the comb this wasn’t a problem because we put it in a paper cup and turned that around. The scanner has problems with really small or big gaps. The small gaps in the comb were too small, in the model there was a relief instead of gaps.
The problem with Hermione was the big gap on one side. Because of the gap the model had a lot of failures. We tried multiple ways of scanning. Eventually we scanned only the handle, by making multiple scans from different angles and combining them together.
Because of the problems we had with Hermione we decided to fill Harry with a napkin, so that no problems would arise from the difference in distance. This went ok, only afterwards the napkin had to be deleted.
The results of processed 3D images will be posted later.
We had to make some choices since there were a lot of bowls, plates, hair brushes and other ceramics. To distinguish them we gave them names. We joked about ‘Harry Pottery’ and decided to choose the names from Harry Potter characters. We based our choices at differences of advantages and interesting characteristics. Since we didn’t know how much time it would take to digitalise the cups and plates, we numbered them in order of importance.
We took in account the different types of advantages.
To make the cups useful again, we need to fix the holes.
Some cups like Harmione and Hagrid have some nice details. Since the CT scanner is not that accurate we have to find other solutions for bringing those details back.
- Fixing techniques
There are different ways used for fixing the objects. How can we translate those methods in our new design?
We decided to exclude the hair brushes in our process, because it has nothing to do with tableware.
Here the advantage is to replace the three separate shards. Diameter of +/- 18 centimeters.
Small plate. Missing a piece and two repaired cracks. Diameter of +/- 15 centimeters.
Kind of fruit bowl with lid. It has nails and glue as fixing methods. The cracks are really fragile, but as good as complete. Diameter of +/- 25 centimeters.
We liked this small cup because of its small hole. Diameter of 5 centimeters.
This one is familiar to Hagrid, but much smaller. The one defect is the bottom which is fixed. The cup is complete. Diameter +/- 15 centimeters.
As the project slowly went into motion we had the first digitizing session in the laboratory of Geosciences&Engineering. Our group was provided with the luxury to first hand observe both micro- and macro-CT scanners in working. Both with their advantages and limitations, they gave us a new perspective of how to order and process given archaeological objects.
When Maaike came in with boxes full of ceramics from the Archaeological archive of Amsterdam, we understood that it was neither efficient, nor possible to scan them all. At this point selection was crucial. At first sight we had three main groups of objects: lice combs (highest level of detail), broken colored ceramics bound with metal strings (necessity to make more detailed scans to understand the technique) and sets of white ceramic tableware lacking multiple shards.
The latter seemed to be the closest to the issues visible in the goal of the project. Yet the other two gave us interesting side paths which would improve overall understanding of the methods and possibilities of 3D scanning. Based on this, we made a queue sorted by importance, which would lead to at least one object of a group scanned.
After the first inspection of the digitized forms we were rather amazed that the precision of 0,3mm was not sufficient for some of the fine-detailed specimens. E.g. the combs lost their teeth, metal bindings were muffled, crack lines barely visible. Consequently we were offered to work with much finer machinery (micro-CT scanner) mostly used for small scale material research. Yet the time and money needed for this method led to only two specimens scanned: the finest ivory comb and a detail of a metal connection. In total we got 13 scans, excluding identical scans in higher precision. The notes and conclusions after this are as follow:
1. There are 2 CT-scanners in the Geoscience&Engineering laboratory:
- Macro-scanner can be used to scan rather big objects, but the fine details are almost completely neglected; object is stationary, thus there is a small chance of damage. Precision 0,3mm.
- Micro-scanner is very slow (1h per object) and has very limited object size: till 100-120mm in diameter; object is rotating, thus it needs to either be glued or fixed, which requires extra attention not to damage the object. Precision 0,03mm.
- Both scan only the form and not color; they can detect cavities, but not slight changes in the material density
2. The digitized forms are saved as 2D images of section cuts in .dcm or .ima file format, which need multiple steps to be converted into editable 3D objects. Even though we were informed that it is a very quick procedure, to gain fine details it is necessary to have a powerful computer(16GB RAM) and correct software (which is usually paid).