DigitalRockArt Overview

Synopsis

DigitalRockArt is an experimental web-based application for rock art recording and research. It is designed for use by public land managers (BLM, NFS, NM, State and Local Parks), volunteer rock art recorders, and rock art researchers. The goal of the DigitalRockArt system is to create a multiple-use rock art inventory by facilitating the rapid recording of large numbers of rock art sites using digital images. Field recording is made more efficient by moving about half of the recording work to the home or office computer. By using a standardized recording and classification methodology, a variety of database query functions are available to visually compare and contrast rock art elements across multiple sites. Use of the DigitalRockArt system will enable public land managers to create and maintain an accurate, detailed inventory of archaeological resources by using teams of volunteers to record the rock art on their lands.

Introduction

Prehistoric rock art is both a legacy given to us by our predecessors and one that we owe to our successors. It arouses our imaginations and creates interest, excitement and a desire to learn more about those who lived here long ago. But while we wonder about the people who created these mysterious symbols, our prehistoric rock art is disappearing. In too many cases we will never even notice the loss because we have no recordings that enable comparison of the past with the present.

While all archaeological resources are fragile, rock art is especially vulnerable to damage by nature and vandalism. If we are to preserve this cultural heritage, we must face the challenge of recording rock art sites before they disappear forever. If we require more funds for protection, we must be able to prioritize our requirements and provide ample evidence to show what needs protection.

The primary obstacle that has prevented more rock art from being recorded is the amount of time and equipment required in the field. The strategy of the DigitalRockArt system is to minimize the man-hours and equipment required in the field and to simplify the field recording tasks so more people will want to participate.

Once the field recording work is completed, the next obstacle in traditional rock art recording is organizing and publishing the accumulated data. Few people have the skill and resources necessary to analyze a site and create a publication summarizing their findings. The DigitalRockArt application eliminates this obstacle because sites added to the database are automatically published to a small and controlled community of rock art devotees. The application provides a variety of reports and database searching and selection functions to facilitate the analysis of rock art elements from multiple sites. CD-ROMs and paper reports can be readily produced from the recorded data.

Multiple teams of rock art recorders can contribute toward building a digital inventory of rock art. As the digital inventory grows, it can be used for multiple purposes including research and site management. Over time the digital inventory can be updated to reflect both the deterioration of the sites and any advances in the classification and interpretation of rock art.

The DigitalRockArt application does not resolve the methodological issue of how to classify or interpret rock art. The application focuses on the issues of how should the data be collected, how should the data be organized and stored, and how should the data be presented. The application does force all users of a single database to compromise on one standard scheme for the classification of rock art. This classification scheme may be easily changed from time to time, but each change may require some minor rework of the inventory database to update previously recorded elements to the new standard.

Rock Art Recording Goals

To date, rock art recording has been primarily driven by the interest of academics and volunteer rock art enthusiasts. The processes used are very labor intensive, and only a small percentage of sites have been recorded. Due to the difficulty of analyzing and organizing the recorded data, publication in any form is frequently neglected for years. Comparing multiple sites is made difficult by the wide variety of formats used by the recorders. The published data frequently lacks sufficient detail to positively identify subsequent site vandalism or natural deterioration.

While the traditional approach is still valid, the design of the DigitalRockArt application anticipates that the large land managers (National Forest, BLM) will coordinate most future recording activities. The driving factor is based upon the same laws that have been in place for many years. Public land managers are encouraged to maintain an inventory of archaeological resources. Digital technology will make the recording of rock art more efficient and enjoyable, thereby making it easier to attract volunteer recorders.

When considering whether to record a rock art site, land managers are confronted with a dilemma. Having the documentation is good, but the resultant publicity may expose the site to more damage from vandalism or theft.

While all concerns can never be completely eliminated, the design of the DigitalRockArt system does address this dilemma.

1. A site's location need only be shared with a small recording team. No site location or access information is present on the database. Site names and numbers may be viewed only by users selected by the land managers.
2. With small teams and minimal field recording time, the impact to a site is reduced and there is less activity to attract the general public.
3. The inventory reports produced by the recording process can be used to better prioritize the allocation of resources for site protection and identify subsequent vandalism.
4. Should there be a theft after a site is recorded and stolen rock art subsequently recovered, the searching functions of the application can provide a means of linking the recovered property to missing artifacts.

There are several reasons why land managers need to organize the rock art recording effort. Reason enough is that the land managers control access through the permitting process. In addition, land managers know the site locations and are in the best position to establish priorities. Land managers are best able to identify those sites that should not be recorded because of their sensitivity and significance to Native American religious beliefs or for other reasons. Land managers are more likely to value recording large numbers of sites using the same standards and producing uniform documentation than academics or other individuals pursuing separate goals.

The goal should be to record all of the rock art, with the exception of those few sites deemed too sensitive. To do so, we need to rethink the process of recording rock art and rearrange the tasks and rewards such that adequate numbers of volunteers will want to record rock art. Field recording with the DigitalRockArt application requires no specialized equipment. The primary skills required are photographic and mastery of a compass and tape measure. The application will provide volunteer recorders with rapid visual feedback and enable them to compare the results of their work with the recordings completed by other teams.

Simplifying the Field Recording Process

To record rock art more efficiently, the first obstacle that must be overcome is to reduce the amount of field time required. The number of man-hours required in the field is viewed as the primary constraint that prevents the recording of more rock art. Field recording is a daytime activity, it requires good weather, there is usually a significant amount of travel time, and it requires that members of the recording teams coordinate and agree upon a scheduled date.

One way to reduce the man-hours required in the field is to change the field recording procedures and processes to take advantage of the features of digital images.

1. It is easy to add textual data to digital images later using a home or office computer. The time consuming process of setting up mug boards can be eliminated (see Exhibit 1).

2. Digital cameras have internal clocks and store information about the date, time, camera model, speed, focal length, etc. within every digital image. This duplicates some of the data recorded in the traditional Photo Log forms and provides additional information. The Photo Log form can be eliminated.
3. The cost of taking additional digital images is near zero. By eliminating the work associated with changing mug boards and completing the Photo Log, hundreds of images can be taken throughout the day in different lighting conditions. The best images can be selected later and the remainder can be archived or deleted.

4. Images can be zoomed or cropped to show closeup detail of a panel or element. This makes it possible to record data related to panels and elements by viewing the digital images in the home or office. The field recording forms can be greatly simplified to collect only the data that must be gathered in the field (see Exhibit 2).

Another way to reduce the man-hours required in the field is to simplify the data structure. Unlike traditional film-based recording methods that record details related to boulders, galleries, and datums, the DigitalRockArt system identifies and measures only one type of object --– a panel. Multiple panels in close physical proximity or on the same boulder are grouped into clusters. This simple cluster-panel measuring and numbering scheme works the same way for boulder-strewn fields, scree, cliff faces, and caves. Clusters are only used for grouping panels and are not measured. Exhibit 2 referenced above reflects this simplified data structure.

Because the DigitalRockArt application can renumber clusters and panels as part of the computer recording process, less time and care is required to number cluster and panels in the field. There is no need to waste time or delay recording until every panel is located and numbered.

Using the DigitalRockArt system, a typical day's output for a two person team recording 50 panels will be four pages of recording forms and 400 digital images.

Overview of Field Recording Procedures

Field recording requires very few tools: a digital camera, compass, five meter steel measuring tape, 30 to 50 meter measuring tape or electronic measuring device, clipboard, forms, pen or pencil, and zip-lock bags with prenumbered index cards. Optional tools include sun shade and reflector, polarizing lens for the camera, tripod, GPS and binoculars to assist in locating distant panels.

The first step at a site is to choose a dominant cluster-panel or other permanent feature as a starting datum point. If the site is too big to record in one day, it is best to concentrate on recording a portion of the site. Once the area to be recorded is determined, the team splits into two. The photographer and possibly a helper photograph the site, the ground surveyor and possibly a helper do everything else.

Even though the photography and ground survey are separate tasks, it is important that there is sufficient communication to ensure that all panels being measured are photographed and vice versa. For large panels that require multiple photographs to capture sufficient detail, the thumbnail sketches on the recording forms should note extra measurements to coincide with each close-up photograph. A small team working in a small area makes it easy to communicate this information.

Digital Photography

The photographer begins to photograph panels by first attending to those panels best viewed in the current lighting conditions. The objective of the photographer is to capture each panel in the most favorable light. Because it is not known when that time will occur, many digital photographs of each panel will be taken throughout the day in various lighting conditions. Other than shooting each panel once with a pre-numbered marker as an aid to be used in sorting the photos later, the photographer does not use mug boards, measuring sticks, or a photo log. The required data will be associated with the digital images later in the recording process.

In addition to the panel images, the photographer should shoot cluster and site overviews. The site overviews should include the datum points and as many panels as practical. A good practice is to take a photo with the cluster markers visible and then take a second photo with the markers hidden or removed.

Ground Survey

The ground surveyor begins by locating the rock art clusters and placing prenumbered markers so they may be easily seen. The markers are placed in an orderly sequence to facilitate field recording. Only clusters are normally marked, not individual panels. Most panels belonging to a cluster will be easily located, and the discovery of those few obscure panels will be communicated to all other team members. Grinding slicks, metates, tool marks, historic writing, and vandalism are always recorded in the same manner as primitive rock art.

Once the markers are dropped, the surveyor works on the recording forms. The recording forms will be used to notate up to eight measurements and a thumbnail sketch for each panel. The origin or datum point, bearing, distance, and elevation measurements are usually recorded only for the dominant A panels of each cluster. Facing, inclination, width and height are recorded for all panels. Recording this data is easiest when two passes are made over the site. On the first pass, the quick sketches are made and the panel facing, inclination, width and height are recorded by following the sequence of the prenumbered markers.

When convenient, each recording form is shared with the photographer. The photographer may use the recording form as a checklist to ensure each panel has been photographed and reviews the quick sketches of the panels to insure that each panel has been photographed in a manner consistent with the panel facing shown on the sketch. Any disagreements with regard to panel boundaries, number of panels and panel facings must be resolved by the team members and the effected panels rephotographed or remeasured.

Prior to the second pass, the primary datum point is reevaluated and, if required, several secondary datum points are chosen. The primary objective in choosing datum points is to create topography charts that will facilitate future site audits or inspections. Datum points should be easy to locate, easy to access and have a direct line of sight to many clusters. The interconnection of secondary datum points should suggest a convenient route of travel through the site by going around natural obstacles.

After the datum points are chosen, the remaining topographic measurements (origin, bearing, distance, and elevation change) are made. The datum points, rather than the prenumbered markers will determine the sequence of measurements during the second pass.

As the day proceeds and the light changes, new clusters and panels will usually be discovered. Should this occur, a surplus prenumbered marker is dropped. Measuring and photographing the newly discovered cluster out of sequence is a minor inconvenience.

Organizing the Digital Images

Taking multiple images of the rock art panels based upon the lighting conditions results in hundreds of images in a near-random order. These must be sorted by panel, cluster overview, site overview, etc.

The photographer must be familiar with at least one software utility needed to accomplish this task. The first step is to create a folder or directory structure suitable for sorting digital images. Create a folder for the survey, a subfolder to hold all the images uploaded from the digital camera, a subfolder to hold the selected images to be used for the recording, subfolders for each cluster (or each cluster-panel), and additional subfolders for cluster overviews, site overviews, etc.

The digital images are uploaded to the starting subfolder. Using the recording forms and a utility such as MS Windows Explorer or Xplorer2 that will display thumbnails of the images, move each image to the appropriate cluster (or other) folder. Assuming the photographer was careful to shoot each panel with the pre-numbered cluster marker and the surveyor was careful to capture unique panel features within each quick sketch, this process should take about an hour for 400 images.

After all the images are sorted, the best image for each panel is chosen, renamed, and moved to the recording folder. The new file names must use a consistent format to facilitate sorting. Images of panels must include both the cluster number and panel letter in the file name. Cluster overview images and site overview images must also be review, renamed, and moved.

In most cases, the folder containing the final images to be used for recording are copied to CD-ROM and mailed to the web-site administrator. For very small surveys, it is possible for the administrator to create and empty survey. In this case, the recording team must upload the images directly to the DigitalRockArt web server.

DigitalRockArt Web Site Administrator

Upon receipt of the images (or request) from the recording team, the web site administrator verifies the concurrence of the land manager/owner to add the survey to the DigitalRockArt database. On large projects with several recording teams, the administrator may have already received a schedule of the sites to be recorded, the assigned teams, and the agreed upon site names, site numbers, and other data to be entered.

If concurrence is given, an Add Survey function is run to enter the pertinent site data and assign update authority to the designated member of the recording team. If the images were received on a CD-ROM, the images are moved into a folder on the server. An email message is sent to inform the recording team that the survey is ready for processing.

Recording with the DigitalRockArt Application

Computerized recording of rock art with the DigitalRockArt application is a multi-step process. The first step is to categorize the images using the CategorizeImages function. Each image is presented to the recorder in file name sequence. The recorder indicates the appropriate category (rock art, site overview, structure, historical artifact, etc.) by clicking a radio button. If rock art is selected as the category, a cluster number and panel letter must be entered. This is made easy because the image file name created by the photographer contains the cluster number and panel letter. If there is a measured length to indicate the image scale, the length is entered and two points on the image are clicked to denote the end points. Panel attributes are entered by viewing the panel image and clicking the appropriate check boxes. The panel data is entered from the field recording form.

The second recording step identifies rock art elements on those images categorized as rock art using the IdentifyRockArtElements function. Images are presented to the recorder in cluster-panel sequence. If there are multiple images for a panel, all the images will be presented together. Any number of the images may be used to identify elements. As each element is identified, the assigned element number will overlay the image as an indication of the recording progress.

The recorder identifies each element by selecting a rock art class (anthropomorphic, zoomorphic, etc.) and then clicking on an image to identify two opposite corners of the cropped element. After clicking two opposite corners, the recorder may click the image a third time to indicate a pecked area, a fourth time to indicate a point of natural patina, and a fifth time to indicate a repecked area. These additional points can be used in lieu of taking Munsell readings in the field. There is functionality within the application to compare and analyze the readings for variations in element patina. As is true of the Munsell readings, there are many problems related to the application of this technique.

The third recording step is used to fine-tune the element and assign detailed attributes using the DefineRockArtElementAttributes function. Each element identified in the previous step is presented to the recorder in cluster-panel-element sequence. The cropping corners or element class may be changed. The pecked, natural and repecked areas may be reselected. Element comments and actual element measurements may be entered. Detailed attributes are assigned to each element by viewing the cropped element image and clicking the appropriate check boxes. Detailed attributes include those that describe the condition of the element (easily visible, difficult to see, etc.) and various aspects of the element design (arms up, arms down, etc.).

Attributes that describe the condition of the element are called general attributes and are common to all elements. Class attributes are specific to each element class. The attributes used to describe anthropomorphic elements will be different from those attributes used to describe geometric elements. Anthropomorphic attributes could describe details such as the presence of fingers, torso shapes or head ornamentation. Attributes for geometric elements will include things like type of object (cross, circle, rectangle, triangle, spiral) and various enhancements such as center dot, tailed, bisected, or rayed.

If needed, there are optional functions for specialized tasks related to correcting, reorganizing, and refining data. The CheckSurveyData function allows images to be deleted or requeued back into the step one processing queue. Individual elements may be deleted or corrected.

The CheckAttributeSet function may be used to fine tune the attribute choices for a single attribute set. For example, if an attribute set for zoomorphics includes choices for length of tail as short, medium, or long, all zoomorphic elements can be displayed sorted by the selection for length of tail. One or many elements can be corrected to achieve a greater consistency.

The CustomizeTopoCharts may be used to fine tune the topo charts automatically generated by the DigitalRockArt application. Large sites may be split into multiple charts. A few clusters that lie far from the main site can be shrunk in scale to produce a more usable chart.

The AnnotateImages function may be used to overlay text or hyperlinked cluster-panel numbers on overview images. The original images are never modified so annotations may be changed or removed.

The RenumberPanelsOrElements function will renumber cluster, panels, elements, and rename files. Renumbering clusters is very easy –-- print out the topography chart generated by the application, notate the new numbers on the chart, and key in the revised numbers. The computer will create a log showing the prior and current panel numbers. Renumbering will update annotated images but will not update user entered comments on images or elements.

A short report highlighting any unusual or significant aspects of the site may be created (an example is Badger Springs). Survey reports can easily be created on the DigitalRockArt web site using the wiki software and can incorporate images or cropped elements from the database to add interest or illustrate a point. Wiki macros may be used to add tables of data and graphs directly from the database. All team members can participate in writing the report by using the wiki tools for collaborative authoring.

Standard Reports

There are currently six standard reports generated by the DigitalRockArt application for each survey. These reports may be viewed via the internet or from a Rock Art inventory CD-ROM or DVD generated by the application. All of the reports feature multiple hyperlink functions that link to additional images or data.

• Show Images by Category - thumbnails of all images sorted by image category.
• Show Elements by Class - all elements sorted by element class.
• Show Topo Charts - panel placement generated from distance and bearing measurements.
• Panel Reports - five report summarizing panel data.
• Element Reports - four reports sumarizing element data.
• Survey Recap - detailed report with thumbnails of all panel images and data and all element images and data.

Research Functions

The research oriented functions work across all surveys.

• Search by Attributes - finds elements with selected attributes.
• Search for Element Coincidence - finds panels or clusters containing two to five elements having selected attributes.
• Search Element Comments - finds elements with comments containing a selected word or initials.
• Compare Survey Stats - compares surveys by numbers and percentages of elements by class.
• Compare Surveys - visual comparison of elements from multiple surveys.
• Compare Surveys by Attributes - enumerates common and unique element attributes in multiple surveys.
• Show Elements by Patina - sorts elements based upon patina.
• Show Notebook - enables users to create collections of elements or images from multiple surveys.

Exhibit 1

The image below shows that digital images may be easily modified to include captions. Captions may include site name and number (both blurred below), cluster number and panel letter, date and time of the photo, facing, inclination, comments and a metric scale. The pointed logo indicates the orientation of the photo.

Exhibit 2

This is the only form used for field recording. Additional panel and element data is collected at the home or office of the recorder by keying the data directly into a computer while viewing digital images of the panels and elements. Note the Elevation Change field is useful when recording on steep hillsides or scree and is usually left blank when recording on flat terrain. The dots on the panel edges of each thumbnail sketch indicate the termination points for measuring panel width and height -- clicking these points on the panel images enables the program to calculate the metric scale shown above.