Metadata for Images: Remote Team

Capturing Metadata Against Images

If you are part of the Remote Team you need to be aware that you are following behind those who have been capturing the information from the physical images. In other words, if you are remote, the information from the physical image would have been made available to you, if there was any, in the Headline field in the Origin panel. Early on in this process, we had some issues with the quality of the metadata capture by the Local Team. That, however, has been rectified and the Local Team operates within a clear protocol. They now capture the information under the following headings in this template:






It is worth being aware of what each of these headings mean so you know how to interpret the information you are receiving. You can view that under section 5.0 on the Metadata for Images: Local Team page.

If you find that you have been assigned images that do not have this template with associated information in the caption field, then please do not continue capturing information against them. Please notify Karabelo Lenong immediately requesting her to assign such material to the Local Team and assign new material to you that has already gone past the Local Team.

Assuming you have reached the point of being ready to capture metadata in the MEMAT Metadata App, let’s discuss what you will actually capture:

Step 1: Capturing Metadata

As the Remote Team, you will have received an image with information from the physical archive in the Headline field in the Origin panel. This becomes the basis for you creating a metadata record.

As an experienced researcher or archivist, you should be using your wealth of experience to evaluate that information and not simply taking it at face value as the source information is often poor. We explain more about that below.

To create the metadata record you are going to be using the information provided to you in the Caption field captured by the Local Team to create a more comprehensive metadata record filling in a limited number of fields in the General panel and one of the IPTC Extension panels.

When we say, “more comprehensive” we do not mean “comprehensive”. Time and budget constraints do not allow a comprehensive capture of metadata against each file. We are going for accuracy and sufficiency, not comprehensiveness. This means we are limiting you to only certain metadata fields in two panels – the General Panel and the Image Description subpanel of the IPTC Extension schema. We will also use the Dublin panel for a final check of the metadata you have captured.

Research Time and Bulk Metadata Capture

The challenge for metadata capture against images such as these is how much time should you spend on research. For metadata capture against images in this project, we have set a limit of 7 minutes per image. Part of those 7 minutes needs to be spent actually filling in the various metadata fields. Let’s say that takes you 2.5 minutes. That will mean you have to limit your research to 4.5 minutes. So if you come across a single image unrelated to any other, that is going to be your limit.

If, however, you have a whole bunch of images all from the same event, then you have the potential to do a whole lot more research. Let’s say you have five images all from the same event. Then you can write the same caption for all five images as long as you are mentioning the main features and personalities in a general caption that applies to the event even if not everything you mention is represented fully by the contents of all five images. Perhaps filling in the fields now takes 1 minute per image (since you are copying and pasting the same information into all 5 or you are using the functionality to capture metadata into all five images at once), that means you have 30 minutes to do research on this event.

Any research should start with being aware of the whole context of the event in the archive. You can learn how to do that here: Research for Metadata Capture for the ANC Archive

1.1 General panel

1.1.1 Title

Fill in the media type in the title field. You will notice at the top of the Metadata Mode page that there is a folder path that will look something like this indicating the storage location of the image:

The folder-path as above will guide you. “Prints” and, “Photographs Loose, ” says that this is a loose photographic print, so enter “Loose Photographic Print”. Use a capital for the first letter of every word unless it is a joining word such as “with” or “and”. Also, record whether the image is Black and White or Colour or Sepia Toned or whether it is a Newspaper Clipping or Magazine or Book Print. Other possibilities are Poster and Art Print. E.g. Loose Colour Photographic Print.

1.1.2 Caption

As a member of the Remote Team, you will be writing your caption based on the information made available to you in the Headline field of the Origin panel. That field should not be empty but have a template and associated information indicating that it has been worked on by the Local Team. If you find an image or a group of images where this is not the case, please alert Karabelo Lenong.

There are five headings in the metadata template used by the Local Metadata Team namely:






The first four headings and their related information you will leave exactly as you found them, except that you are going to move them to another field once you have completed writing your description/caption. That is the information that you will place greater weight on in terms of writing your caption since it is fairly reliable. You should be aware that this is a liberation archive that was gathered over time by many people and the provenance was not always captured with each item in the physical archive. There are (fairly rare) occasions where the information even in the physical archive may not be correct.

The last heading, INFO NOT FROM PHYSICAL ARCHIVE: is information added by a previous metadata capturer, often without any access to the information in the physical archive. Information not from the physical archive has been found to be very unreliable and you should treat it as such. This information is not only found under this heading in the Caption field but also in other fields that you may find have been filled in such as Location and Headline etc. Treat all with suspicion! In writing your caption you are going to be replacing INFO NOT FROM PHYSICAL ARCHIVE: and its associated information.

Once you are done writing your caption please leave the following headings and their related information in the Headline field in the Origin panel:





While you are leaving the above fields and their related information as is you do need to make sure you delete the following heading and its related information:


The caption that you wrote should be left in the Caption field and the information from the physical archive should be in the Headline field. The Headline field is a temporary place for storing that information until we get a more correct permanent home for it. You won’t need to fix that later on, we will do that in the background once we have identified the permanent home for it.

Now let’s turn our attention to how you capture information under that Description heading which is really your caption.

The caption/description should answer the questions: Who? What? When? Where? How? and Why? if you have the information. The words in the caption will be used in database searches, thus keep the description to the point and use words and word combinations that might be used in a search. Generally, captions should be brief (aim for under three sentences), to the point and specific and correct in terms of grammar and spelling.

Remember, if you have no information about the image, do not make up the information. No information is far better than wrong information. You may make an informed guess but say so. E.g “A man wearing a mayoral chain, possibly the mayor of Worcester…” The caption’s main use is as a descriptive index, not to provide in-depth information which researchers can find from other sources. Adding irrelevant extra information will make the search process harder.

On the whole, Africa Media Online likes to follow the Associated Press Style Guide (this Quick Reference Guide is courtesy of the Colorado Department of Public Health and Environment). One exception being that AMO uses acronyms that are likely to be used as search terms. Use proper sentence case in writing captions.  Here is a useful guide to the rules of capitalisation. And the Faculty of Arts at the University of Bristol provides a useful guide on using numbers. Here is also a useful link to the Guardian and Observer style. This latter document is not a style guide to captioning per se, but rather to how you write certain names etc.

If you have the information available, The caption should speak to the following: Who?

  • Record who is in the picture and where they are in the image. Do not assume that the researcher recognises any of the people. For example, “Alfred Nzo (right) walks with what may be Jimmy Carter (left) and an unidentified man.” Or list people from left to right, or front row, back row, e.g. “At Worcester Railway Station from left to right, Jeff Radebe, President Nelson Mandela, Riyadh Williams, Mayor of Worcester and an unidentified woman.”
  • If there is a child in the picture and you know their age include it in parentheses e.g. “Caitlin Hendry (3)”
  • If it is a portrait mention it. This is quite an important search term as researchers often look for an image of someone in portrait format.
  • Be very careful about the correct spelling of names as they will not be found if misspelt. Keep in mind what name people will search for. A caption listing President Mandela may not be found by someone looking for Nelson Mandela. As per AP style guide, you can refer to someone by their surname after mentioned fully once. Check name spellings (using Google for example), then add them to your spell check dictionary, which will alert you to future misspellings.
  • Use titles if you wish, but they should be the ones current at the time the image was taken. E.g. Archbishop Desmond Tutu from September 1986 to June 1996, but Archbishop Emeritus Desmond Tutu after that.
  • Generally do not use Mr, Esq, Mrs, Miss or Ms. Only use them if you know the surname of the person but do not know the first name. What?

Record what is in the picture in the most factual way possible, in the present tense. If you recognise anybody in the picture, name them and place them, e.g. Alfred Nzo, seated left, speaking to Joe Slovo, right. If not, describe them, e.g. A group of young men listen to a speech, or Two women/Two protesters hold up a placard with slogan: ‘End Apartheid Now!’. Try and pack as much of what is in the photograph into the caption. If the event is known, add it to the caption. “Alfred Nzo gives his acceptance speech after receiving his Honorary Doctorate from the University of Amsterdam.” If nothing is known about the event, just describe what is in the picture. A man gives a speech at a university lectern.

Transcribe words on posters, banners and T-shirts putting them in inverted commas to show the words are as seen in the image, as in the example above. When?

  • Give the day, month and year in words if you know it, in the Caption field. Please use this standard format [DD] [Month Name] [YYYY] e.g. 23 March 1997.
  • If you just have the month and year, then it should be [Month Name] [YYYY] e.g. March 1997.
  • If you only know the year, then it should be [YYYY] e.g. 1997.
  • If you know the date within 2 years either side you can put c. 1997. (Circa abbreviated to c. means about, so c 1997 would be between say 1995 and 1999).
  • If you only know the decade then it should be for example 1990s. Please note, there is no apostrophe between the end of 1990 and the “s”.
  • You can include the date in your description, or you can start the field with it, put a full stop and space, then start the caption.
  • This date information will be repeated in the Date Created field, but this time in numbers. The challenge with the Date Created field is that the order of the day and month will change according to the settings of the computer you are using or the browser on the computer you are using. If your computer is set to the American system, the month will appear before the day mm/dd/yyyy. If you are on the British setting then it will appear with the day first dd/mm/yyyy. When I set Chrome to the South African setting the date in this field appears in this order: yyyy/mm/dd. You should work out what yours is set to. Ideally, set it to the South African or British setting, it is more logical. If you only know the month and year, then use 01/03/1997 to reflect March 1997; or if you only know the year use  01/01/1997. Check that the date captured in the Caption corresponds to the date in the Date Field. Where?

  • Record where the picture was taken, particularly the city, province/state and country. Put a comma between each of these.
  • If you know the actual location record that, e.g. Luthuli House, Johannesburg, Gauteng, South Africa.
  • Repeat this information in the location fields and repeat the most relevant and likely to be used in a search in the Keywords. For example, Mandela’s visit to Zaire should include Zaire in the keywords.
  • Names of locations should be the name that was used at the time of the image. For instance, an image that was pre-1994 taken in Johannesburg would be Johannesburg, Transvaal (now Gauteng) and not Johannesburg, Gauteng. If you are unsure, Google it.
  • If there is a new name for the area, put it in brackets after the old name in the Caption field. For example: “Johannesburg, Transvaal (now Gauteng)”, or “Pietersburg (now Polokwane)”.
  • Be very careful about correct spelling of names as they will not be if misspelt. Why?

Tell us why this picture is significant. This may be background information and context to the photo if it is available. E.g “Nelson Mandela embraces Bibi Mukadam (Bibi Dawood) who was tried for treason with him and others in 1956.” Don’t add information that is not relevant to this situation you are describing. How?

Sometimes the how is an important part of an image and should be mentioned. E.g. “Protestors desperately scramble over each other to escape the police teargas”; “Peter Magubane with camera in hand bravely and defiantly stares down the police controlling crowds at the start of the Treason Trial, Johannesburg, Transvaal (now Gauteng), South Africa, December 1956.” When keywording this aspect one usually uses nouns, however, e.g. anger. The camera angle of the image should be included if it is unusual e.g. Aerial or fish-eye

1.1.3 Free Text Keywords

Our recommendation is to capture between five and 10 keywords according to the principles outlined in Keywording Historical Images.

1.1.4 Location

This is the specific location. For instance, if the place that Alfred Nzo is at in my example is the United Nations building in New York City, New York State, United States of America, then the Location would be “United Nations.”

1.1.5 City

This is the name of the city at the time that the photograph was taken. In our example, it would be “New York”.

1.1.6 Province/State

This is the name of the province or state at the time that the photograph was taken. In our example, it would be “New York” (New York is the name of the state in the US that New York City is in).

1.1.7 Country

This is the name of the province or state at the time that the photograph was taken. In our example, it would be “United States of America.” Use the full official name of the country and not an abbreviation e.g. United Kingdom, NOT England or UK.

1.1.8 Author

  • This is the name of the photographer usually. It may be the name of the artist who made a poster, or artwork or an author.
  • Leave this blank if you do not know who the photographer/author is.
  • If it is already filled in with “ANC Archives” delete that.

1.1.9 Copyright Holder

Quite often the copyright holder is the photographer unless the photographer was working for or commissioned by an organisation. You may come across images that were supplied to the African National Congress by a press agency such as the Associated Press or Reuters or by a newspaper such as The Star. These agencies are likely to be the copyright holders. If you do not know, leave it blank. If an agency is mentioned, then fill it in here.

1.1.10 Copyright Classification

You need to know the date of the photograph in order to select the correct option.


If you do not know the date of the photograph, then leave it blank by selecting this option:


1.1.11 Copyright Status

The default setting on this drop-down is Unknown. Only if you know the file is copyrighted, change it to Copyrighted. This will be true of images that have been supplied from picture agencies or from publications such as a newspaper like The Star. Otherwise, leave this as the default Unknown.

Ignore the fact that all the images are watermarked with “Copyright”.

Files that are older than 50 years (1967 and older) should be marked as Public Domain.

1.1.12 Date Created

This would be the date that the photograph was taken. This should be captured in the yyyy/mm/dd format, e.g. Christmas Day in 1970 would be written 1970/12/25. The system requires all the figures to be filled in, so if you only know the month and year then write it yyyy/mm/01. e.g. December 1970 would be 1970/12/01. If you only know the year write it yyyy/01/01, e.g.1970 would be 1970/01/01. If you do not know the date, do not fill this field in, just leave it blank.

1.1.13 Publication Date

This is only applicable if you know the date that the image was published in a newspaper or magazine etc. Otherwise, leave this field blank.

1.1.14 Author Deceased Date

You need to know who the photographer is in order to fill in this field. Google them to see if they are still alive and if not, when they died. Don’t spend long on the research. This can be filled in later when the image is licensed.

1.1.15 Save the General Panel

Once you have done the above, click Save to save your work.

Apart from the General panel, there are just two other panels that may be applicable.

1.2.0 IPTC Extension Panel

This panel has a number of sub-panels. There are only a few fields here that may be applicable:

1.2.1 Image Description subpanel Person Shown

Fill in the name of each person using the correct case and separating persons by a comma. e.g. Alfred Nzo, Jimmy Carter Featured Organisation Name

Only fill this in if the organisation features prominently in the image. For example, if the image shows the annual congress of COSATU then fill it in as Congress of South African Trade Unions. If the ANC is campaigning for votes and ANC regalia is obvious then add “African National Congress”. This term can be copied from the keywords, as it should feature there if the organisation is prominent in the image.

1.2.2 Admin subpanel Image Supplier Name

If the image comes from an agency, such as the Associated Press (AP) or Reuters or from a newspaper such as The Star, then fill the name of the organisation in here. Supplier’s Image ID

If there is a specific number associated with the image, such as an accession number or some number supplied by a press organisation such as Reuters, AFP or AP, then fill in that number here. Don’t fill in the number of the digital file here. This is only for numbers that come to the archive from elsewhere.

1.3.0 Dublin Panel

Before you move on to commit the image, click on the Dublin Panel, that provides a good summary of what you have captured and check your work carefully. Grammarly also works in this panel, which it may not do in the other panels. Don’t fill in any more fields here. Simply check what you have already filled in elsewhere and is displayed nicely here.

Step 2: Save and Commit

Save each panel as you move from one to the next. Once you have completed the metadata capture, take some time to review everything you have captured making sure you have no spelling errors and no grammatical or factual errors. Once you are content you have captured the material correctly, then click on “Commit” at the bottom in the left-hand sidebar.

Step 3: Select the Next Picture

Back in the Home screen click on the next picture and follow the process again.