Building Preventative Conservation & Documentation into a Private Digital Art Collection

Kate Weinstein
Electronic Media Review, Volume Seven: 2021-2022

The Carl & Marilynn Thoma Foundation has a Digital & Media Art collection of 308 time-based media artworks and counting. Of these, 149 works integrate software or video. Drawing from the extensive resources provided by larger institutions with existing registration and preservation strategies, as well as our close relationships with time-based media dealers and artists, we have adapted many policy and procedure best practices to better suit our smaller staff. This has enabled us to achieve quicker timelines for acquisitions, exhibitions, and loans, as well as an entrepreneurial spirit toward collecting and preservation collaboration.

This article outlines the decision-making process and preventative conservation progress we have made over four years at the Foundation for our fast-growing time-based media collection. This article describes our internal learning curve and outlines not only the policies we have implemented for acquisitions, loans, and internal documentation, but also those we have discarded or have had to rethink as our planned procedures did not meet real-world needs. Finally, this article provides practical examples in implementing a multiphased approach to time-based media policies, procedures, and documentation for smaller institutions.

Introduction

The Carl & Marilynn Thoma Foundation averages 30 time-based media (TBM) artwork acquisitions a year. With a small staff (and without any conservators on staff), it falls upon only a few people to process each acquisition from registration, testing, and inspection to installation and preservation planning (creating archival and managed files, verifying checksums, etc.). Beyond our active acquisition program, we also proactively lend our collection to museums throughout the United States, both as individual loans and as packaged exhibitions. Many of our borrowers do not have their own TBM collections nor their own TBM conservator to oversee the exhibition planning, installation, and maintenance while on loan.

By the end of 2016, our collection of TBM artworks surpassed 100 objects. These objects included software-based artworks, as well as video-based, LED-based, and custom circuit board-based artworks. This milestone was the motivation to create a policy for acquiring, lending, and documenting TBM artworks. Over the next four years (2017–2021), we overhauled our documentation policies and procedures, staffing, databases, servers, and physical spaces for inspection. From top to bottom, we completely reimagined our pre-acquisition, acquisition, post-acquisition, loan, exhibition, storage, and shipping procedures to better document, preserve, and care for these artworks.

Phase One: Assessment_Knowledge Gathering (2017)

At the beginning of 2017, I began a review of our existing policies, procedures, and internal cataloging standards for the entire collection, but with a particular focus on the TBM works. From this review, I noted several consistent hurdles. The first was our geographic locations. The artworks were often stored in our Santa Fe location, but most of the cataloging was occurring in our Chicago location. Additionally, sometimes the Chicago location would house the certificates of authenticity (COAs) and main artist files or physical backups, sometimes not. Because the artwork frequently traveled between the two locations (as well as on loans), there was inconsistency as to which location would be the primary holder for the physical artworks, main artist files, COAs, physical or digital archival versions, or exhibition files.

Second was that our existing collection management system (CMS) was underutilized, and the artworks were minimally cataloged. Additionally, the software had limitations to its customization and could not support many of the fields required to properly document TBM artworks. Because of these shortcomings, we relied heavily on personal computers and a shared Dropbox account to store documentation, files, images, and archival files of video artworks. The Dropbox folders did not have standardized organization or file-naming practices, making this information difficult to find.

Third, many of our existing procedure forms, particularly the bills of sale and condition reports, were built for traditional media and did not account for the specific needs of TBM.

The year 2017 was primarily one of research and knowledge gathering of other institutions’ best practices. These conversations gave us the resources to begin implementing policy and procedure and documentation standards, as well as gave us insight into where I could push in our own requirements for deliverables from dealers and artists. Because the Foundation was a young institution (three years old at the time) and still heavily straddled the line between a private collector and small institution, many of my initial requests required negotiation with the dealers and artists to see what they would be willing to provide us. By the end of this year, I implemented an initial acquisition process, a TBM bill of sale, loan policies and procedures, and updated registrarial cataloging.

Acquisitions and Loans
For our new acquisition policy, I implemented a bill of sale aimed toward TBM. New language added both requested and required deliverables and I provided the bill of sale to any new dealers or artists. However, I did not always provide a bill of sale with existing dealers. This included language for the following:

1. Requested permission to publicly exhibit and loan the artwork
2. Requested main artwork files on a separate portable drive (not just exhibition files embedded within the artwork)
3. Required artwork on dedicated, new equipment (instead of used or refurbished equipment)
4. Required install and operating manuals
5. Requested artist-approved high-resolution images and video clips (if applicable).

I also implemented a loan policy and procedure for all our artworks and created condition reports specifically for TBM artworks that would list all the provided components, installation, and maintenance notes. Along with this, I created an internal outgoing loan form to account for testing workflows, all the physical components sent to the lender, and their compiled user manuals. This form was completed prior to releasing artworks for loan so that we could begin to troubleshoot any outgoing loans remotely instead of having our borrowers reach out to the artists whenever an issue popped up with the equipment.

Registrarial Documentation
I implemented registrarial standards for documentation and created custom fields in our CMS to record duration, a full list of components tied to the artwork, and cleaning information (fig. 1). Of note, my colleagues had implemented cleaning information because previous borrowers would try to clean the artworks like regular office equipment, often not realizing that we were providing the original, artist-provided equipment and art elements. These three fields contained information already recorded elsewhere by my colleagues. Finally, I created standardized naming practices and subfolders in our Dropbox to make the information not stored in our CMS and the archival video files easier to navigate and find.

By the end of 2017, I knew where our policies and procedures should be heading to meet best practices, and made great strides into the preparation required to centralize the information we already had on each artwork.

Phase Two: Initial Implementation (2018)

In 2018, I began to revise our acquisition and loan policies and procedures with additional requirements. In addition, our CMS released a components module so that there was a place to begin documenting artwork components.

Acquisitions and Loans
In the first revision of our acquisition policy, I updated our bills of sale by changing any requested deliverables to required deliverables and included the following:

1. Required permission to publicly exhibit and loan the artwork
2. Required testing periods for all artworks—20 business days for traditional media, 30 business days for videos, and 60 business days for software-based artworks
3. Required artist to review the main and exhibition files and assure the files were free of unintended technical defects
   • Required the artist or dealer to provide replacements of included components if any defects were discovered later (upon reasonable notice)
4. Required artwork to not contain viruses that could compromise our software, hardware, data, and so forth.
5. Required the artist to authorize us to make file copies and migrate media to new formats for preservation and exhibition
6. Requested main artwork files on a separate portable drive, including additional deliverables:
   • Source code/project files
   • Wiring diagrams
   • Backup specialty or artist-modified components (same-batch LEDs, etc.)
7. Required physical COAs signed by the artists
8. Required artist-approved high-resolution images and video clips (if applicable).

The bill of sale was also now mandatory for all purchases, without any exceptions. Our loan policy and procedure changed to stop providing the main artist files with any video loans. Instead, we would make an exhibition file to lend or to exhibit in-house.

Registrarial Documentation and Storage
In 2018, I flushed out much of our registrarial documentation and maxed out our CSM’s customization with new fields based on information recorded elsewhere by my colleagues. The goal was to provide a centralized location to store and revise the information and minimize data duplication and conflicting information across other files. This included special install and display instructions, safety information, operating summaries, behavior summaries, technical narratives, on-off procedures, and video links (fig. 2). Additionally, with the CMS’s new component module, I began to document any physical and digital components provided by the artist. This often was a stop-gap procedure, as the component module did not have many of the fields needed and was aimed toward more traditional media. As such, I had to use some fields creatively to record as much as possible within the CMS and there was still a lot of information without a place to store within the CMS. In particular, the digital components were only minimally recorded using HTML lists within the physical component record. Even this minimal recordkeeping was time intensive as I was providing a narrative-like description of the file contents (fig. 3).

I also implemented guidelines for storage. This included implementing electrostatic discharge (ESD) shielding polyethylene bags for many components, instead of regular polyethylene bags, and requiring all cables or any removable components to be secured with Velcro straps instead of zip ties. I also made Chicago the primary location for all main and archive physical drives and Santa Fe the primary location for all physical artworks. This was to provide us with two geographic locations for our data while we did not have a full repository for our digital data.

Phase Three: Revision_Edit (2019)

Beginning in 2019, I revised our acquisition, loan, and documentation policies again to implement newly learned lessons from our 2018 policies as well as pushing our best practices to be more in line with larger institutions.

Acquisitions and Loans
Within our acquisition process, I implemented optional artist interviews and artist questionnaires for major or complex artworks. During this year, the artist questionnaires went through three revisions based on the responses we received from the artists, and they were eventually broken into three questionnaires, based on media type: software, video, and custom electronics/LED works. For our loan program, I began providing standardized loan packets with each artwork. These packets included normal operating behavior videos, on/off videos, installation and operating manuals with wiring diagrams, condition/iteration reports, equipment user manuals, and combined condition/iteration reports which included behavior summaries, technical narratives, daily operating summaries, and daily on-off procedures. My intention with the loan packets was to provide enough information for the borrowing registrars, exhibition managers, and installers to properly install and assess the artworks on their own, as well as minimizing our courier requirements. Additionally, it was to provide information for guards or other staff who would be turning the artworks on and off daily during the exhibition to minimize calls for troubleshooting. We continued to provide exhibition files for all video artworks and began implementing exhibition equipment for software-based works where the original artist’s computer was not required for the artwork to run or for works where the original artist’s computer would be at risk for failure during the exhibition.

Registrarial Documentation and Storage
In 2019, the Foundation hired a collections assistant to help with the additional workload the new documentation, acquisition, and loan processes required. While the physical components documentation remained relatively the same, this additional position gave us bandwidth to develop the documentation standards of the digital components. All components were now input into the system, including Foundation-purchased equipment. However, all physical and digital components, regardless of type, were recorded the same due to the limitations of the CMS (fig. 4). At this time, I was able to streamline the digital recordkeeping using command line and disk utility tools (primarily tree and mediainfo) to copy the information I needed into a Word table, which I HTML-pasted into the system. This worked with moderate success. However, I could not attach multiple images or files to the component record, so I would have to keep additional images, videos, and metadata text in Dropbox. Policies and procedures were put in place to record the needed video documentation for new acquisitions and the loan packets.

By the end of the year, I moved all our archival video files out of Dropbox and onto their own dedicated RAID, keeping the same file organization previously implemented in Dropbox. Dropbox then only held documentary information, not actual artwork files.

Phase Four: Revision_Edit_Final (2020)

The year 2020 brought major changes to our documentation process and physical space to process artworks.

Acquisitions and Loans
The year 2020 also brought on a few more revisions to our acquisition and loan process, including another revision to the bill of sale. The bill of sale now also:

1. Required artist questionnaire
2. Requested artist interview
3. Required additional deliverables for all software-based artworks
   • Remote login information
   • User and admin login information
   • License keys.

For loans, I began working with borrowing institutions to see what equipment from their pool could be used for video artworks instead of always sending our equipment along with each artwork. This was to streamline our loan process and to make the installation easier for the borrowing institutions, as they would have familiarity with their own equipment.

Registrarial Documentation and Storage
Large changes to our processing storage, hardware, and software were made in 2020. I began a migration to a new CMS, Argus, which was customized for our collection and includes multiple object designations, component designations, and fields to properly document all our TBM artworks. This system gave me the tools to begin implementing my final goal—all documentation in a single system. A copy of the Dropbox and old CMS system was archived. Now as we move data into Argus, we are cross-referencing the existing Dropbox files and images and deleting any redundancies in Dropbox and leaving only working files. I also began creating standards for the new database to properly record our artist-provided components and the equipment pool. With the new system, I was able to implement new component identification numbers that provided more context for their purpose and have the system generate all condition, identity, and iteration reports.

Along with Argus, we also built out a new TBM processing workstation at our Chicago location. The workstation included a Mac Pro, a FRED, a mobile imaging table, a media rack for older media players, and a large server to house our archival files for both video- and software-based artworks. We also purchased two large cabinets to hold in-process artworks and a final storage system for TBM portable drives, flash drives, and edition boxes.

Phase Five: Revision_Final_Final (2021)

Acquisitions and Loans
Finally, at this point, we did not have to make any yearly revisions to our acquisitions and loan policy, except to have our bills of sale and loan agreements autogenerate by Argus.

Registrarial Documentation and Storage
In 2021, most of the work had been done to our registrarial documentation procedures and working with Argus to implement and adjust our fields to record TBM artworks. The goal was to minimize staff time and data duplication while maximizing documentation records within Argus. With the new workstation, we began creating disk images of all our computer-based artworks and began to properly test all TBM artworks. At our Santa Fe location, we built out a secondary workstation, aimed toward documenting the physical components and video works.

Instead of relying solely on the artists to assist with migration and replacement equipment, we had now begun to migrate media and replace equipment in-house. To illustrate our current progress, next we present a condensed case study.

Michal Rovner “Duhai”
In 2018, the Foundation purchased Michal Rovner’s Duhai (fig. 5). The work projects animated figures on a found stone, displayed within an artist-made, museum-style vitrine. When purchased, it was still playing from a DVD in a portable DVD player with a small projector, all original and from 2004. The original equipment had additional fans and adapters added over the years to keep the increasingly overworked equipment from overheating while on display. Since the work combined TBM equipment as well as artist-made, dedicated art elements, we hired a conservator to make repairs to the vitrine and add pads to the found stone’s mount to stop it from scratching the vitrine when installed. However, in-house, we made significant changes to the internal equipment and media files. This included:

• Updating the DVD player to a BrightSign media player
• Disk-imaging the DVD and archived ISO
• Replacing the projector
• Migrating media from the DVD to an mp4 file on a microSD card
• Remapping the video file to project on stone with new equipment
• Replacing glass panes for comparable acrylic sheets for safety in transit, install, and display
• Removing outdated equipment (IR sensors, adapters, switching power supplies, cable keepers, and fans)
• Replacing fluorescent lights with LEDs (comparable lumens, CRI, etc.)
• Streamlining the powering system
• Adding a surge protector
• Moving the power cable exit on the vitrine so that the vitrine no longer rested on the power cable
• Changing the install procedure and manual to match new equipment requirements.

In the end, we removed 31 pieces of equipment and parts and replaced them with 12, streamlining the data and power connections, as well as minimizing the equipment held within the vitrine, decreasing the likelihood of overheating (fig. 6).

Figure 7 also provides examples of the final appearance of Rovner’s Dahui registrarial and component recordkeeping in Argus and how we document our TBM artworks in Argus currently.

Conclusions

In those four years, I learned how best practices, tools, and software capabilities all determine our ability to document TBM artworks. It is my hope that our learning curve provides insight and inspiration to other smaller institutions struggling with the same hurdles in processing documentary-heavy artworks with a small staff while maintaining acquisition and loan schedules. Our forthcoming goals include automated fixity checks and networking our server so that our other locations can remote-in to process artworks outside of Chicago. Larger goals include maintaining pace with new acquisitions while completing our backlog of older artworks. As we move forward, I am sure that there will be new opportunities to streamline and increase useful information in our system and look forward to all the possibilities.

ACKNOWLEDGMENTS

Thank you to the Thoma Foundation and my colleagues for their patience, support, legwork, and feedback in building our time-based media policies and procedures. Additionally, thank you to everyone in the Electronic Media Group and New York University’s It’s About Time! workshops for letting me pick your brains over those four years. It has been a pleasure sharing everything you have taught me.

AUTHOR

Kate Weinstein
Collections Manager & Registrar
Thoma Foundation
Chicago, IL
kate@thomafoundation.org