Navigation

Page tree
Skip to end of metadata
Go to start of metadata

What are Datastreams?

Datastreams are the files that make up a repository archival information package. All datastreams are generated automatically when an object is added to the repository, Adding, updating or replacing datastreams allows you to manage your objects within CTDA.

In This Guide


Accessing and Downloading Datastreams

  • Navigate to your object and click the download button 

  • The icon will expand to display the datastreams associated with your object. Click on any of the links in the menu to download that datastream.

  • See the sections below if you need to manage, add or, replace a datastream.

Managing Datastreams

  • To view the datastreams associated with an object in the repository, navigate to that object's page. Click the "Manage" tab. 


  • Click "Datastreams".


  • After clicking "Datastreams" you will see this screen. 

  1. Datastream ID (DSID) - an internal unique identifier within the scope of the digital object, for example, RELS_EXT, MODS, or TN.
  2. Datastream Label - the human-readable title given to a data stream. It does not need to be unique but it should be informative enough for a user to understand the purpose of the data stream.
  3. Versions - used to view past versions of the datastreams and revert to an early version of a particular DS to correct a mistake.
  4. Operations - the options in this column allow you to manage each datastream.


Add or Replace a Datastream

All Fedora Digital Objects contain datastreams which can be added, replaced or sometimes deleted. Remember that some datastreams are required and cannot be deleted.

Replace a Datastream

  • Go to the object for which you want to replace a datastream. Click “Manage”.


  • Click “Datastreams”.


  • If the datastream already exists, you can select “replace”.


  • If you want to replace the TN, or thumbnail, datastream click on “replace” on the line for that datastream ID “TN”.


  • In the new window, click "Choose File" and find your new thumbnail. Click "Upload", then "Add Contents".


  • This is the same process to replace any datastream you have permissions to.

Add a Datastream

  • Not all content models come with thumbnail datastream. Sometimes, ingests don’t go as planned and you need to add a datastream rather than replace it.

  • Go to the digital asset where you want to add a datastream.

  • Click “Manage”.


  • Click “Datastreams”.


  • Click on “Add a datastream”.


  • In the new window, fill out:
    Datastream ID – the ID is an autocomplete field and datastream IDs are a controlled vocabulary. MODS is always MODS, TN is always TN, OBJ is always OBJ. If you are adding a datastream that is for preservation purposes, you can create a naming convention for that purpose, i.e. AM for Archival Master.
    Datastream Label – The label can be the filename or just left with the default such as MODS Record.


  • Upload your file and click “Add Datastream”


Deleting Datastreams

You have the ability to delete datastreams from your objects in CTDA, but we recommend that you do not.


Datastream List

IDLabelContent Models
RELS-EXTFedora Object to Object Relationship MetadataAll
MODSMODS metadata record, created at time of ingestAll
DCDublin Core metadata record, created at time of ingestAll
OBJThe original file ingested into the repository (File format for this DS will depend on the content model used)All
TNThumbnail image, created at time of ingestAll
TECHMDTechnical MetadataAll
JP2JPEG 2000 derivative, created at the time of ingestLarge Image
JPGMedium-sized JPEG, created at the time of ingest and used in the standard image viewerLarge Image
MEDIUM_SIZEA compressed version of the image, created at the time of ingest and used on the object's View pageBasic Image
PROXY_MP3MP3 derivative, created at the time of ingestAudio
PDFPDF derivative, created at the time of ingestBook, PDF, Newspaper
RELS-INTInternal Fedora relationship metadata defining the dimensions of the JP2 datastreamBook Page, Newspaper Page
OCROptical Character Recognition fileBook Page, Newspaper Page
HOCRA converted version of the OCR datastream, intended to be more human-readableBook Page, Newspaper Page
COLLECTION_POLICYSpecifies which content models are associated with a particular collectionCollection
FULL_TEXTText of a PDFDocument, PDF
PREVIEWPreview image, created at time of ingestDocument, PDF
MP4MPEG-4 video derivative, created at time of ingestVideo
MKVMatroska video derivative, created at time of ingestVideo
  • No labels