Chapter 3. Registry methodology

Sequences of Time Arrested: the Kodachrome Toronto Registry Initiative

In this chapter

The Kodachrome Toronto Registry is a reference tool designed primarily for researchers. This tool offers a structured response for answering two basic questions: what and how.

First: “What do we know exists of Toronto’s history which was also preserved with Kodachrome slides or movies?” Second: “How will these collections be known, described, accessed, indexed, located, and preserved for the benefit of making a stable reference tool?” The Registry is the first step on the way to that what and how. It puts us closer to knowing with certainty.

Owing to how the Registry initiative was undertaken without a disciplinary foundation in the archival or information sciences (or mentorship by specialists in either area), the organizational theory applied for Kodachrome Toronto is kept to the basics. This elementary approach is still adequate for steering how the Registry field research was reviewed, aggregated, and entered into the Registry.

The methodology for developing the Registry was principally informed by two texts: “A first introduction to archival sciences” by Theo Thomassen, and Archival Strategies and Techniques by Michael R. Hill. Both are designed for use by scholars outside the information sciences realm. The former is written for an audience who wants to establish basic protocols for consistency and integrity for describing an archives or collection. The latter is designed to assist the researcher who may call upon an archives for their own investigative work, such as a historian or biographer. By using these two guides, each helps to deliver a bi-directional understanding of the archive, how it is established, appended, and how it will later be accessed.

Records and elements

For the Registry, details on each collection reviewed (or subjects within larger collections) are assigned to an accession record for purpose of collecting evidence on its qualities. The accession record represents the logical representation of the physical object — in this case, a collection — it is describing (Proctor and Cook 2000, 91). As an “agent of accountability and evidence” which functions as a tangible “memory” (Thomassen 2001, 375), the record itself facilitates to enable administrators and researchers to locate metadata relevant to a collection, so to assess the state, scope, and status of the physical parts which that record logically describes.

The different parameters denoted within the record, meanwhile, are its elements. In a database, elements within a record are searchable. They can be indexed along with corresponding elements from other records within the same archives, or records from different databases which share the same element properties (e.g., “title”, “date created”, “abstract”, etc). The structural arrangement of elements contained by a record, meanwhile, helps to define the format of that record. How well the elements in a record interrelate with one another (that is, to provide both a faithful and meaningful account of the physical contents which the logical record describes) can help to inform the quality of data documented by the elements in that record. For example, an element denoting a date value by itself is meaningless unless it is associated with other elements which give that date some kind of useful context.

Finding aid

The Registry is engineered to serve as a research finding aid. A finding aid aggregates individual records in a centralized manner which, in this instance, logically describe physical objects — despite those physical objects being housed neither centrally nor managed by a single party (Hill 1993, 46; Thomassen 2001, 378). A finding aid makes it possible for researchers to survey the entirety of known physical objects associated with the collection without actually being present to locate or review any of those objects.

While a finding aid generally can describe a collection held by a single archives — replete with a repository to house and maintain the structural integrity of that physical collection — the Registry is not an archives in of itself. The efficacy of descriptiveness for each record, meanwhile, is wholly reliant on voluntary updating (usually by an administrator, sometimes by a custodian). These updates denote not only the present location of a physical record, but it also describes the state, scope, and changes (if any) to its contents. For example, there is no viable — or, at least, cost- and time-effective — way to centrally maintain the currency of logical records for a finding aid when there is no central means to manage and house the logical records. Doing so would produce a chaotic system requiring the collection administrator to visit every site with some of the records and then update relevant elements in that part of the finding aid. This is why the invention of the database was so invaluable: it was the first step to centralize logical records strewn across different locations. In addition, a web-based interface, such as planned for the Registry database, further helps with managing the logical contents.

For Kodachrome Toronto, the means to logically bring together knowledge of discrete physical collections is essential. For the inceptive stage of this project — the scope of this SRP — there will not be a centrally managed repository of physical collections. This is mostly due to extensive resource and labour constraints. At a later time, however, should an archives repository be established, this finding aid will already be in place. Updating these records will be a matter of denoting change of media location, date of that record change and, where necessary, change of custodial management.

Scope of the Kodachrome Toronto Registry

The Registry must fulfil three core objectives:

  1. data solicitation and collection
  2. database development
  3. content analysis
Soliciting, promoting, collecting

The greatest challenge for initiating the Registry was to begin locating known collections and repositories where Kodachrome media was known to exist. Achieving a baseline of collections required exploring several avenues — some structured, others more ad hoc:

  • word of mouth
  • social media
  • online message and discussion forums
  • news media
  • institutional and public archives
Word of mouth

Peers in the Toronto civic social community — principally at social networking mixers — were apprised of the Registry plans in 2010 and encouraged to follow updates to the project’s research. This approach helped to yield some of the following opportunities.

Social media

Three core social media tools were used: the Kodachrome Toronto: 1935–2010 blog; @KodachromeTO on Twitter; and the Kodachrome Toronto Flickr group.

In reverse order, the Kodachrome Toronto group pool on Flickr was created in June 2009, well before plans for the Kodachrome Toronto Registry began. Membership was open to any Flickr member with digitized Kodachrome slides to post from their own photo stream. The group was roughly 40 members strong at the time the project was officially launched in January 2011 (61 members as of March 2012).

@KodachromeTO, meanwhile, was prepared to help network with other GTA photographers, as well as with institutional collections and archives from across the world, wishing to follow the Registry project with interest. Twitter was used to announce plans, media links, post featured photos, share incidental facts uncovered by research, and provide live field research findings. Tweets posted by other users which were relevant to Kodachrome film, projects, or found ephemera were also re-tweeted.

Lastly, the Kodachrome Toronto: 1935–2010 blog was designed for posting project updates, observations, selected photos, and overviews too long for Twitter format. The blog features a survey form for contacting the project with leads on possible collections [Appendix A]. This form was used to request for interview once field work got underway.

Message and discussion forums

A listserv and two bulletin board forums were notified of the research. First, the Kodachrome list is an email server whose delivery is by opt-in email subscription.

Second, the Analog Photography Users Group (APUG) is a forum dedicated to enthusiasts of non-digital photographic imaging. Discussions range across every conceivable aspect of analogue or chemistry-based imaging techniques and formats. Topics (or even mention of) digital formats, or the use of digital equipment for archiving analogue media, is verboten. Those discussions, such as how to scan Kodachrome, are re-directed for forum members to register at APUG’s sister site, the Digital Photography Users Group (DPUG). On APUG, amateur, semi-pro, and pro photographers engage regularly with photo industry vendors and photography engineers (some with applied experience of Kodachrome’s research-development process at Eastman Kodak in Rochester, New York). Discussions can sometimes be spirited and active — especially when Kodachrome is of topical interest.

Third, the Kodachrome Project Forums is hosted by Kodachrome enthusiast Daniel Bayer. Bayer, a professional photographer, is renown for being one of the most prolific users of Kodachrome still film during its final few years of availability. Bayer travelled extensively to document culture, one-time events (e.g., the inauguration ceremonies of President Barack Obama in 2009), and distinct, mostly American geographies during the film’s final few years. Bayer was also the next-to-last person to have a roll of Kodachrome film processed on the very last day — 18 January 2011 — the final Kodachrome-equipped lab, at Dwayne’s Photo in Parsons, Kansas, stayed in operation.

News media

While not initially planned as a way to raise awareness, a few reporters solicited interest in the project’s research. The Toronto Star published an interest piece by vintage film enthusiast Eric Veillette in February 2011. This generated buzz which redirected readers back to the project blog. BlogTO’s Derek Flack published an in-depth interview in March 2011. In late April 2011, the editor of UofT Magazine, who expressed interest in this project by a UofT alumna, requested an interview which, while conducted, was ultimately not used for a story.

Public and institutional archives

Several archives — inter alia, Library & Archives Canada (Ottawa); the Archives of Ontario (York University); the City of Toronto Archives; and the CNE Archives (Toronto) — were surveyed for potential leads on Kodachrome media in their holdings. Given time, human resource limitations­, and travel constraints, a visit to Ottawa was not feasible; in addition, few archives records met the criteria to reasonably confirm that photos associated with a fonds were shot with Kodachrome. Two search reviews at the Archives of Ontario were equally inconclusive — which may be less that there are no Kodachrome holdings and more that the Archives finding aid insufficiently described the content formats described by each record. The CNE Archives remains a queued candidate for the second stage of field research (planned to follow the completion of this SRP). The presence of Kodachrome media was confirmed in situ during a December 2011 visit. Subsequent visits to detail those items will be essential. The City of Toronto Archives, while previously known to hold at least a handful of Kodachrome-based media in selected fonds and collections, yielded much more than anticipated. This will be discussed in depth in Chapter 4.

Database development

For the Registry, Filemaker Pro was used.

Filemaker Pro is a proprietary database platform designed to work as a standalone client. This decision was dependent less on the benefits of the platform itself and more on its portability in network-poor environments (i.e., places where wireless internet could not connect to a database located on a remote or cloud server) and its gentle curve for learning and setting up a working database in a minimum of time. The goal at this inceptive stage was less about setting up an extensible or enterprise platform than it was to allow with ease the description labour of manual records entry — that is, “the work involved in transforming objects… into searchable descriptions that will assist subsequent retrieval” (Warner 2010, 41). This is where the physical is rendered into the logical.

While Filemaker Pro’s database format is not as robust an environment as enterprise platforms like MySQL, Postgres, or Oracle are, its conversion to an enterprise database (ideally to an open-source platform, given wide community support for technical concerns) should be fairly uncomplicated for a database administrator to manage. Furthermore, third-party vendors, separate from the database makers, dedicate their resources into developing archival management software for institutional-calibre holdings. These content management solutions, operating as a separate layer atop the database layer (relying on direct communication with the database), provides much greater customizability than what the all-in-one kit (of database and front-end interface) of Filemaker Pro can deliver.

Records architecture

The database was prepared with several elements expected of an archives search tool. These include an accession unit, or accession number (a fixed, unique identifier assigned to a collection) (Douglas 2010, 31); date of record creation/modification; ownership/custodial information; and a textual description of the physical objects described by the record. Further, project-specific elements were added, including options for denoting Toronto’s localized geographies and the Kodachrome media formats used [Appendix B].

Bringing the Registry to fruition means having to collect together known physical repositories Kodachrome media into a finding aid — here, a database. This open-ended process is designed to readily continue after this initial setup and inceptive findings phase is completed. The Registry database is being prepared and designed with extensibility and convertibility in mind: at a future stage, digitized photo and video representations of the original media can be grouped under a single record to allow for a virtual inspection of the original media. Separately, the exporting of logical records to XML )the open-platform markup standard for encoding data strings now used extensively by web and mobile applications) will also make the propagation and distribution of that content a straightforward matter.

Content analysis

The methodology for assessing, reviewing, and documenting new collections is fairly consistent. There are minor variations for identifying whether slide films and movie films are Kodachrome in origin. A very unscientific, extremely complicated, “green-amber-red” flowchart of elimination is used to filter out false leads.


The first step is to pre-assess the scope of content, based on preliminary reference data (provided, if available, by the holder of that content). When the provenance of that collection is a private party, verify in advance that they believe the collection was made using Kodachrome and not another colour-reversal medium. Depending on their experience with Kodachrome or photography more generally, this may only be a marginally useful metric.

When the provenance of a collection either is institutional or corporate, then the approach, based on search result hits, should proceed as follows:

  1. Are “colour photos” or “colour movies” described in the record description?
    • if yes, proceed
    • if unknown, proceed conditionally
    • if no or “b/w”, stop.

  2. What are the estimated years when the collection was created?
    • pre-1935 (lantern slides, Autochrome, carbro, or hand-tinted monochrome), stop
    • pre-1950 (could be Anscocolor/Agfacolor, Ektachrome, Dufaycolor), proceed conditionally
    • post-1985 (spike in widespread adoption of non-Kodachrome stocks), proceed conditionally
    • post-1998 (for City of Toronto, no evidence of post-amalgamation Kodachrome use), stop
    • if originator was known as an “avid photographer”, proceed conditionally
    • all others (Kodachrome dominant between 1950–1985), proceed

  3. Are photos denoted as “slide” or “reversal”?
    • if yes, proceed
    • if not specified and/or listed as formats larger than 35mm denoted, proceed conditionally
    • if “lantern slides”, “negatives”, or “prints”), stop

  4. Are movies denoted as “35mm”?
    • if no, or as either “8mm” or “16mm”, proceed
    • if not specified, proceed conditionally
    • if denoted or inferred as a professional production (probably a reproduced copy), stop
    • if yes (35mm Kodachrome offered only via Technicolor in 1940s for special applications), stop
    • if denoted as “video copy of original”, verify whether original exists. Otherwise, stop

Requesting a review

Request to view the collection (or fonds, if part of an archives). Request an interview with the custodian and/or the owner. If conducting the interview in a location other than where the collection is usually stored, have them bring a sample portion from the entirety of the collection. Conduct a preliminary inspection. Look for telltale signs of provenance:

  • if 35mm, 828 “Bantam”, 126 “Instamatic”, or 110 “Pocket Instamatic”, proceed conditionally
  • if 127 or 620 format film, stop
  • if sheet film, post-1951, stop
  • if 120 format film, pre-1986, stop
  • if packaging is lab-labelled as “Kodachrome”, confirmed
  • if emulsion relief visible, then Kodachrome (or K-12 variant[fn1] from 1960s–1970s), confirmed
  • if film rebate (sprocket area) shows listing as Kodachrome, confirmed
  • all others, stop

Preparing the record

Create a new database record. Populate the elements with as much information as available. To assign a permanent record number:

  1. Assign a unique accession number using the “KTYYYYNNN” convention:
    • KT” prefix for Kodachrome Toronto (n.b., this was arbitrarily chosen);
    • YYYY” for four-digit year when the accession number is assigned;
    • incremental number “NNN”, starting from “001”, for each record (collection).

  2. A way to contact the owner or custodian is required.
  3. Assess what was shot in Toronto, the GTA, or the Golden Horseshoe.

[fn1] 3M’s Dynachrome, Konishiroku’s (Konica) Sakuracolor, and Fujifilm’s Fujichrome were three competitor lines of film designed for Kodachrome’s K-12 chemistry.

Contents ©2012 Astrid Idlewild. Do not excerpt without written permission. A printed version of this SRP is filed with the Blackader-Lauterman Library of Architecture and Art at McGill University. The online version of this manuscript was edited and serialized in 2013.

Series Navigation<< Chapter 2. Finding a medium for posterityChapter 4. Field research findings >>