About

The Gevaert Paper Project: unlocking a reference collection of photographic paper, packaging and documentation (2020-2024) is a project of FOMU – Fotomuseum Antwerpen (Photomuseum Antwerp) and was set up after the museum acquired the historical collection Agfa-Gevaert in 2017. The goal of the project is to preserve the rapidly disappearing knowledge about analogue photography and to give the general public access to a wealth of information concerning photographic paper, photo documentation, packaging and sample books from the period 1894-1964.

Because of the size of the historical collection, the Gevaert Paper Project focusses solely on the photographic papers, packaging and documentation. The 67 samples of photographic paper included in the historical collection, supplement the 1110 samples of photographic paper from the FOMU’s own collection. In the end, 1300 photo paper packaging and 70 hanging folders were digitised and registered as part of the project.

The Company Agfa-Gevaert: a brief history

In 1890, then 22-year-old Lieven Gevaert (1868-1935) opened a photography shop in Antwerp. However, he quickly discovered that he was too dependent on the import of (expansive) photo paper and decided to start producing his own photo paper. The result was Calciumpaper in 1894.

The success of Gevaert’s Calciumpaper was enormous, both domestic and abroad, and allowed Gevaert to build a completely new factory in Mortsel in 1904-1905, thus strengthening his leading position in the business. Gevaert guessed correctly that photo paper – and not photographic plates – would become the future go-to material for photography. The reason for this was that paper had several advantages: production was no longer manual (and was, as a result, very cost-efficient), the shelf-life of the products was guaranteed, the necessary equipment for the photographer was simple, and sales could be done through a distribution network.

Along with the success grew the number of innovative papers: daylight paper Blue Star in 1895, Gevaert Mat in 1903, daylight paper Ronix in 1904, chloro-bromo-silver paper Ridax in 1905, enlarger-developer paper Ortho-Brom in 1905, gaslight-developer paper Artos in 1911, and Gevaluxe luxury paper (with a velvety surface) in 1933. Gevaert introduced roll film in 1923, followed by the first X-ray films and sound cine films. In 1935, the year of Lieven Gevaert’s death, the production of papers for document reproduction began. The only thing Gevaert did not succeed at in those first decades was the development of its own color film, because the company was not fully committed to structured technical or fundamental research (unlike competitor Kodak Research Laboratories).

Because of the fierce competition with international players such as Kodak, Agfa, Fuji and Ilford, Gevaert began to engage intensively in marketing and corporate identity from the 1950s onward, with a particular focus on the foreign market. The company’s ‘Reclaamafdeling’ took many initiatives, such as publishing its own annual calendars and handbooks (with explicit mention of the negative material used), or an active presence at world exhibitions. Furthermore, the house magazine Gevaert Foto-Dienst (1938-1951) was launched, followed by Fotorama (1952-58), which featured practical photography tips for professionals and amateurs. Gevaert also produced demonstration folders for potential customers and purchased shots from internationally renowned photographers in order to give the photo papers the necessary allure. The result of these marketing efforts was that Gevaert, like the various other international photography players, started to develop a clear corporate identity. This was mainly determined by a main color: Kodak was yellow, Fuji green and Gevaert orange-red.

Gevaert experienced its absolute peak in 1964, when the company had as many as 9,000 employees (including 111 researchers). In that same year, the merger between Agfa AG and Gevaert took place. In other words, the peak of the Gevaert company immediately marked the end of a hitherto purely Belgian history. The new imaging company was initially called Gevaert-Agfa and from 1971 Agfa-Gevaert. It was fifty percent owned by the Gevaert holding company and fifty percent owned by Bayer. In 1981 Bayer became Agfa-Gevaert’s sole shareholder when it bought out Gevaert. For Agfa-Gevaert, the 1980s and 1990s were dominated by digitization and image processing in the graphics and healthcare industries. Today, the Mortsel site still employs some 3,000 people.

Historical Collection Agfa-Gevaert

The Historical Collection consists of much more than just the samples and technical files. The collection started in 1963, when (former) employee Laurent Roosens (°1923) began collecting ‘Gevaertiana,’ mostly documents and objects related to the company. Soon after, however, Roosens acquired part of Lieven Gevaert’s personal archives as well as the photo-archives of the publications branch of the company. The Historical Collection continued to grow as it acquired company files, the files of ex-employees, newspaper clippings, etc. Eventually, the Historical Collection was divided into subcollections, of which the most important were the ‘Assortiment’ (Assortment, detailing nearly every product that was made by Gevaert), ‘Documentheek’ (Document Library), and ‘Fototheek’ (Photo Library, a collection of many proof prints made by company photographers testing new papers). The ‘Documentheek’ was further categorized with 999 keywords, called ‘D-numbers,’ relating to any and all subjects concerning the company Gevaert. The bulk of the collection focuses on the Gevaert period. There is information available on the merger with Agfa and the period Agfa-Gevaert, but this information is much scarcer than the pre-merger period.

In 1984 the collection was radically re-ordered and structured on the occasion of the exhibition Lieven Gevaert 1868-1935. That same year it became clear that the Sterckshof site had become too small and that a new home for the historical collection had to be found. This time the choice fell on the Varenthof in Mortsel, a small castle bought by Agfa-Gevaert and where the collection was housed in 1988 – coincidentally also the year Roosens retired. Nevertheless, this pioneer remained active until 2014 as archivist and custodian of the historical collection, which he himself regularly called ‘the most important photographic archive in the world’.

Database

‘Reading’ photographic papers provides a lot of information and data about the language of photographers and photographic manufacturers, such as Gevaert. Elements such as texture, gloss, color and thickness, among others, make each paper unique. If you can identify these, you can name the manufacturer and learn a lot about the technical and artistic methods of photographers. However, this pivotal information is not typical ‘museum’ data that can be captured in a standard collection management software which is why FOMU looked at a different type of database to store the data: Wikibase.

A Wikibase is a semantic repository that operates akin to Wikidata, with the main difference being that only users authorized by the FOMU can add or remove data. This solution – rather than simply uploading the data to Wikidata – was preferred in order to have control over the nature of the consensus. In addition to describing and collecting the data, the project aims to make the data optimally available to other cultural heritage institutions by mapping it to a common ontology within the sector. The importance of using an ontology is that the information will be interpreted as the same by machines (and humans) everywhere. In this way, the concept of ‘thickness’ will have a clear definition, referring to a shared ontology, so that other institutions can adopt this information knowing exactly what it means. For the purpose of this project, a CIDOC-CRM ontology was chosen because Wikibase uses a knowledge graph and not a relational database.

Data model

The Gevaert data model distinguishes between 3 levels: ‘product line, ‘product’ and ‘product instance’. In order to avoid repetition, each level contains information that is true for all underlying levels. What follows is a hypothetical example to explain the information structure of the data model. A photo paper was launched under a particular product name, for instance Gevarto. Gevarto’s user target group was the artistic photographer, and the characteristics of the photo paper were tailored accordingly. Gevarto is the ‘product line’ in this example.

The specific characteristics of a photo paper were cast into product codes for easy recognition, such as ‘Gevarto K8XZ’, and which we describe on the level of “product”. In that code, the ‘K’ stands for a cardboard backing, ‘8’ for a particular texture and reflection, and ‘XZ’ for extra soft contrast. We describe that combination of properties within a product code in the record ‘product’. For example, there are records for Gevarto 7N, Gevarto 7V, Gevarto 47N, and so forth.

Finally, we describe the features that make a package unique compared to other packages: there is a handwritten inventory number on it, inscriptions may have been made by previous owners of the package, we note whether the package is complete (how many photo papers have been taken out), and what the physical condition of the package is. This information is described on the level of the ‘product instance’. In other words, a ‘product instance’ is the physical representation of a ‘product’ within the collection. In some cases, the database may contain multiple ‘product instances’ of the same ‘product’: each product instance is included since inscriptions, inventory numbers and packaging condition may differ per product instance.

The three levels of the Gevaert data model are based on the naming conventions that Gevaert applied to their products. Most of the photographic papers were named following the same logic: a brand, or ‘product line’, such as Gevarto was registered as a trademark and then further designated according to codes Gevaert established to categorize papers. For every possible combination of surface texture, contrast, bottom layer color, and carrier, Gevaert had a different code.