11  Pottery from the Americas

Under the NMHC ARP subaward, Otto proposed generating an online exhibit and finding guide for Pottery from the Americas permanent exhibit. Continuing from Otto’s work, Craig developed an exhibit website and finding guide (Figure 11.1). Harper was the graduate student who did most of the data entry and cleaning for this project. Harper also selected the objects to feature, and Craig worked with this list of 39 objects when building the exhibit website and finding guide.

Figure 11.1: Pottery from the Americas online exhibit and finding guide

11.1 How It Works Under the Hood

Craig’s solution is based on Quarto’s document listing feature with a grid layout. The site presently features 39 objects and is scaleable to a degree (but see Section 11.2). New photographs were taken of the ethnographic vessels. The existing archaeological photographs should be re-captured.

11.1.1 Photographic Challenges

Craig and Harper worked together to generate new photographs of the ethnographic vessels. Some of these objects presented fascinating photographic challenges. Reflective vessels are difficult to light for a photograph because light sources easily appear as distracting highlights on the vessel. Even with light “bounced” off of a ceiling, highlights are difficult to contend with on dark colored but reflective objects. Putting such an object into a light-box causes the entire vessel to glow as one massive highlight. To get the softest most diffuse light possible, Craig used a light diffuser on the source and bounced the light off the ceiling. Doing this means there is low light. Low light means long exposure times. However, that is not the sum of it.

As focal length become shorter, depth of field shortens. In photographing small objects like pottery vessels, one is often shotting macro which means that depth of field is very truncated due to short focal lengths. While bokeh or selective focus can be desirable in portraiture, it is not recommended for documenting museum objects. Rather, unless there is an overwhelming reason otherwise, it is generally preferable to have the entire object in focus. High lens aperture (A.K.K “stopping down”) can be used to mitigate an abrogated depth of field. However, with higher apertures there is less light hitting the sensor which attenuates shutter speeds.

Between the need for diffuse light to mitigate distracting highlights, narrow depth of field due to short focal lengths, and the use of high aperture to extend depth of field–one may find themselves facing long exposure times. Given the generally controlled conditions, Craig opted for soft low lighting and high aperture with very slow shutter speeds. Some images were captured at f/20 or higher with shutter speeds between 8-12 seconds. In such contexts, a solid tripod and camera head are essential. It is also recommended to shoot with a timer delayed shutter or remote shutter release. Craig used the camera’s built in time delayed shutter. It would be really interesting to learn how others approach capturing images of reflective objects.

11.1.2 Building the Exhibit Web Interface

Using the Object ID, each item was assigned to a page of the same value. Likewise, object photographs are named according to the Object ID and stored in an images directory. Attributes and descriptions are pulled from a PastPerfect database export file and displayed on the object’s interface page. Similarly, using the Object ID corresponding images are pulled and displayed on the object’s page via the pixture library (Francis 2022).

YAML front matter on individual pages was used to link Object ID with the PastPerfect record and images. For example:

---
params:
  id: "1967.06.209"
title: "Object ID: 1967.06.209"
image: "./../images/UM_1967.06.209_1.jpg"
categories: archaeological
description: "Unknown Plainware Minature Bowl"
---

11.2 What Could be Improved

Museums have large numbers of objects and may want to change which items are featured at any given moment. Therefore, a useful system must be able to generate pages quickly or change them easily as the need arises. While Craig’s PfA exhibit is scaleable, doing requires hand creating individual pages for each featured object. This is fine when working with dozens of images, but it could quickly get difficult with samples in the hundreds let alone thousands of objects. Ideally, pages would be rendered by a script that passes variable values to a template that generates a new page for each record of the table. This lies in the domain of parameterized reports. The trick of this approach is iterating over records of a table and passing those values to parameterized reports. While Craig has yet to implement this on data exported from PastPerfect, Craig was able to demonstrate proof of concept based on sample data (live demo and GitHub repository). Moving forward, it makes sense to design around the ability to automate the production of large numbers of pages based on a script and a template.