Based at the University of Bristol, the 'Historical Photographs of China' project began work in 2006. It grew out of a 'Dissemination award' to Robert Bickers, as leader of an Arts & Humanities Research Council funded project on the 'History of the Chinese Maritime Customs Service'. This developed into an initiative that searched for material far beyond descendants of the staff of the Customs Service. Initially, support for publication online was provided by generous collaborators at the Institut d’Asie Orientale, Lyon. But 2017 saw the launch of the predecessor of the current platform, refreshed in 2022, and wholly based at the University of Bristol, and designed and maintained by the University.

The project located, digitised, and published online photographs of China held, largely, in private hands outside the country. The aim was to help make this virtual photographic archive of modern China publicly available, open access, which is done under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International Licence (CC BY_NC_ND 4.0). Download, transmission, or print, for research or educational purposes, is permitted, providing that the provenance is acknowledged, but no commercial reuse is permitted without the permission of the owners of the material.

From 2007 onwards the project has been recognised by the British Academy as an Academy Research Project. To find out more about the project's various activities and history please look at its Visualising China blog. You can also find it on Bluesky, Twitter, Facebook, Instagram and Sina Weibo. You can watch a short film about the project here, and listen to a BBC Radio 4 programme about it here.

In March 2021, after 15 years, the project ceased active work, with the transfer of its digital assets to the University of Bristol Special Collections, which was accompanied by the accession of a large collection of original material. This is discussed in this blog post. Queries about the project including requests for reuse should be directed to Special Collections.

Robert Bickers, Professor of History, Project Director

Technical Notes

(These notes were drafted when the project was active, and are provided for historical and technical interest).

Where do our images come from?

The majority of the collections that are presented here came from, and remain, in private hands. In the main, correspondents, alerted to the project by word of mouth, or who have come across the platform itself, the blog, or other forms of publicity, reached out to the team. Generally, they had a historic family connection to China. Those who originally created the collections came from all walks of life, and lived or worked in, or visited, many different parts of China. Collections came from across the UK, as well as the United States, Canada, Australia, France, India, Hong Kong, and China. Their current locations range from within walking distance of the office, to Sydney. Some corporate and institutionally-held material was also digitized.

The photographs held were taken, collected, bought, or otherwise acquired by those individuals and families within whose collections they now sit. While we have some holdings of professional, or expert amateur photographers, much of what is showcased is vernacular photography: photographs taken by women and men with no particular photographic skill, taking advantage of technical innovations in photography that made it progressively easier and cheaper.

What do we do when we are offered them?

A collection may consist of prints in albums, loose photographs, negatives, 35mm transparencies (slides), magic lantern slides, published books, post cards, memoirs, scrap books and ephemera. We can arrange for a secure courier to collect material, which may have been donated or loaned to the HPC project. When new material arrives in the HPC project office, its extent is recorded, and digitisation progress is logged. If the collection is a donation, it is formally accessioned into the University of Bristol's Special Collections, allocated a DM number and stored in acid-free folders or archival boxes.

If material has been loaned to the project, the owner is invited to complete a licence form, that was developed by the University’s Secretary’s Office, and which is signed on the University’s behalf by them. This protects the rights and property of the owners of the physical material that we receive, and it gives the project permission to copy and store the photographs and to upload low-resolution jpegs to the project web sites. If a publisher, for example, seeks to use an image from the platform, the request is relayed to the owner of the photograph. We provide owners with a set of the digital images when returning their original material. 

Digitisation

Images are digitised with a Canon EOS 5D SLR digital camera, mounted on a Kaiser RS10 copy stand. The lights on each side of the copy stand are set at 45 degrees and metered – the camera can then be set at an exposure of f11 and (about) one fifth of a second, at 100ASA (an underexposure of a third of a stop is advisable). A 100mm macro lens is used to re-photograph originals as small as a postage stamp, while other lenses (50mm and 28mm) cater for larger originals up to around 30 x 40mm. Mirror lock up is enabled. A small spirit level is used to ensure that the camera is level. A remote shutter release and an angle viewfinder are attached to the camera. All photographs are re-photographed in colour (Adobe RGB) to capture the warm tones.

When copying shiny originals (e.g. albumen prints) care is taken to avoid reflections appearing on the original photograph (use of black masking tape, black paper, no overhead lighting, reflections from light-coloured clothing, hands, cables, etc). Similar precautions apply when a sheet of glass is laid on top of an original to make it flat when it is photographed. To avoid unwanted reflections appearing in the digital image, originals are photographed through a large sheet of black card, with a lens-sized hole in the middle, held just below the lens of the camera.

Whole pages of albums are copied as well as the individual prints. Albums are propped up with foam supports to ensure that the page is level. Damage to the album is avoided as the album does not have to be fully opened.

RAW (.CR2) image files are transferred from the camera and opened in Adobe Bridge, where they are rotated where necessary and saved as both dng (digital negatives, i.e. masters) and tiffs at 300ppi (these image files being about 40Mb, after cropping to original format dimensions). In Adobe Photoshop, the tiffs are finely straightened and cropped, as required. Optimisations include adjusting levels, curves, contrast and brightness. Images for exhibition can be spotted and sharpened. Use of a Wacom Graphics pad and pen avoids RPI due to excessive Photoshopping. The dngs and tiffs are labelled in Adobe Bridge and transferred to the University of Bristol’s Preservica Digital Asset Management System (DAMS).

A lightbox placed under the camera on the copy stand is used to re-photograph black and white negatives. Negatives are placed emulsion side down on the lightbox and held flat with metal rulers or a sheet of glass. The camera is set to 100ASA, f11, aperture priority. The shutter speed is set to suit the density of the negative. The viewfinder is filled as much as possible. In Photoshop, an ‘action’ is set up to invert the image (which turns it into a positive) and to de-saturate it by 100% (which removes any residual colour cast/stains in the celluloid or glass of the negative), leaving a black and white image to be balanced, as described above. Monochrome negatives of any size can be digitally copied quickly and inexpensively with this innovative technique.

Preparing images for publication

Information about an album is initially recorded on a spreadsheet. Any captioning in the album or on the photographs themselves is transcribed, although on the whole we receive scant information this way. The names of people depicted, locations, dates, the name of the photographer and so on, are recorded, along with links to sources of corroborating evidence or other helpful information. The Historical Photographs of China (HPC) image reference numbers (unique identifiers) of any related images is noted (‘related’ images may be scattered elsewhere in the same collection or in another HPC collection altogether).

The columns on the spreadsheet relate to the fields on the platform and all metadata revolves around the HPC image reference number. The metadata on the spreadsheet is transferred to the DAMS in bulk, or inputted directly. In future, the DAMS will add this metadata to the pages on the HPC platform, along with low resolution jpegs (96ppi).

In addition to basic information about each photograph, the team add additional metadata to support searching. The experience of processing more than 22,000 images does mean that the team is quite familiar with many of the topographic subjects that recur across collections. On the whole, the project aims to provide sufficient minimal data to support discovery by users through keyword searches, or tags. Further research by users has significantly enriched the data we provide, but we do not undertake advanced research ourselves unless preparing blog posts.

The platform

The HPC platform was developed by the University of Bristol’s Research IT team, using the open-source content management framework Drupal and the MySQL relational database. The platform provides a mechanism to manage the images and add and enrich metadata prior to their publication on the website. The platform has a mirror site in China, hosted at Shanghai Jiaotong University.

The future

Originally conceived as a pilot for a potential University of Bristol developed digital assets management system (DAMS), this plan was superseded by the decision to purchase the Preservica DAMS. In 2020 work began to move HPC assets into the DAMS as part of its implementation. At the same time, Research IT began to rebuild the public platform using the Python/Django framework which will integrate with the DAMS. The new version of HPC will take advantage of containerization by using Docker, and will be hosted on Microsoft’s cloud-based platform, Azure. The new version was launched on 15 December 2022.