Graphics and Archaeology: Interpreting the Past

Snowbird, Utah
20-23 May 2000

 

Presentation:

From Trowel to Computer Graphics: Ten Years of Experimenting with the Past

Presenters:

  • Philippe Martinez

Summary:

Philippe's presentation covered over ten years of his career as an egyptologist and archaeologist faced with applying various computer techniques to the analysis and presentation of their work.

The first project discussed was a study of some 12,000 decorated stones discovered reused inside of another structure at Karnak. The task involved deciphering the pieces and their glyphs in order to reconstruct their original context. Through a series of iterations in computer representation (first text, then glyph fragments, then a homogenized combination of data), a successful reconstruction was possible. This was accomplished in collaboration with a sponsoring department of the company Electricit de France (EdF), one of many such successful projects with this organization.

The next project was a visualization of the Karnak temple throughout its evolutionary history. A precursor to this project involved a complex wooden model that showed the site in its final form but was not flexible enough to represent a broader history. Working with engineers of EdF and the PDMS software system from CAD Centre, they created a simple, text-based vocabulary of primitives that could be interpreted and compiled into a larger visualization of the temple from any angle and stage. From this powerful model, it was possible to test many hypotheses and views of its lifetime.

Philippe showed a recent reconstruction of the monumental work (1986) as visualized in 1996 with Maya software to achieve the original vision of a walkthrough.

From this work, he joined a project to reconstruct the temples at Delphi from the remains there. Again, the approach used was a synthesized text vocabulary this time with Greek primitives to build up larger structures. The complexity created was powerful but also unusable. Vast numbers of interior primitives/facets went unused from external views, thus burdening the rendering system.

To add to this data, decorative sculptures were scanned with 3D scanning techniques (much like those used for dinosaurs!) As a side note, Philippe showed a wonderful video of the 3D scanning technology used on the interior of a prehistoric cave only accessible by divers. The work was done for a French TV program and many viewers did not realize the results were synthetic. Anyway, the scanner was brought to Greece and Philippe showed how various gargoyles could be scanned, meshed, and added to the growing visualization of the Delphi temples.

"We have also used 3D scanning to scan huge stone fragments and artifacts." This allowed them to effortless reorient the fragments in space as to estimate original pitch and angles of columns and roofs. This allowed the French to determine (actually prove) that the columns were much higher/longer than claimed by the English (something disputed for over 50 years). Refining work has continued to add decoration and color as details were discovered.

Another reconstruction performed was the Colosseum using the same previous techniques. By building one complete segment of the ampitheatre, it was discovered that the original structure was built on 6 different axes, even though it was possible to repeat large segments with repeated segments.

Philippe concluded his presentation showing a visualization of pyramids done to show them on the grassy savanna and with trees and animals as were the climatic conditions of the time. He also showed work done post-EdF. This included visualizing a plot of land with simple 3D StudioMax models - this allowed them to realize they were working on a fortress with walls and towers (even though they had been working on the site for some 5 years!)

Conclusions / Issues:

  • They built 3D reconstructions of the Delphi temples, but at some point, you must stop because you don't have sufficient data/facts to continue your estimation
    • the result leads to models that were not possible in time
    • you end up with unreal views of the structure
  • They also plotted the models and colored them with watercolours by an architect to produce artistic estimations of the data
    • In 1996, for a brief time, this work even appeared on French postage stamps as part of a 150 year anniversary
  • 3D StudioMax visualizations can be just as effective for gathering quick insights as massive, expensive undertakings with larger systems
  • Care must be taken with computer representations - showing a Photoshop image - it is possible to do too complete of an estimation that could be proven wrong later (he showed a courtyard done by an artist that dramatically completed information not available in the original view)
  • It was noted that sometimes computer representations lead to totally inaccurate representations whereas hand-drawn images provided simpler, greater accuracy of the same subject
    • "Do not feel humbled if you are expressing your work with conventional methods [especially if they are more appropriate and accurate]"
  • Question - should we use estimations or should we be doing things to preserve entire sites?
    • yes, you need to make decisions based upon your intent - at Delphes, we were going to reconstruct an estimated view, rather than working directly from every block that was on the ground
    • they only had 10 days to complete their scanning work - far too short to scan every block at the site
  • Question - PDMS is an expensive piece of CAD software (complete with design database) - did you require time in the CADCentre to complete this work?
    • unfortunately, we've had to decimate our original datasets in order to even interact with them in a VR environment
    • Philippe only has access to 3D StudioMax now
    • Lon did not that some large institutions occasionally become interested in large projects (such as Taisei in the 1980's)
  • Question - do you feel comfortable doing such reconstructions with such small amounts of data / fragments?
    • yes, I'm comfortable - I know the level of detail that I have and what I'm trying to achieve
    • I must do harder work to show what I do know and what I don't know
    • I cannot present everything as truth - and I must help my audience to understand that

 

[ Previous ] [ Index ] [ Next ]