Introducing Mirador 3: The next generation image comparison viewer

We are proud to announce the completion of the first phase of development of Mirador 3.  For fourteen weeks between January and April, a team consisting of contributors from four institutions across the US and Europe rebuilt Mirador anew.  Following a comprehensive year-long design process led by Jennifer Vine and Gary Geisler, a dedicated team of engineers from Stanford University, Universität Leipzig, Princeton University and Harvard University followed an agile software development process and produced a feature-rich alpha version that is ready for testing and ongoing development.  

Mirador ( is a web-based digital image viewer that supports deep zoom, analysis, comparison and annotation of high quality cultural heritage images.  It is compatible with the International Image Interoperability Framework (IIIF), and thus has the capability to load and display hundreds of millions of images published on the web by participating IIIF institutions.  Originally built by the team at Stanford, Mirador has been developed and adopted by dozens of libraries, museums, archives and other cultural institutions world-wide. It is being used as the primary image viewer in scores of library and museum catalogs, and is also being used creatively by students and researchers as a sophisticated instructional and research tool. At Stanford, Mirador is being used by a range of library staff, scholars and instructors, and has been implemented as a viewer option within the Spotlight exhibits platform.  The cultural, educational and research potential of Mirador is vast, but the codebase and user interface of Mirador 2 had reached the point where it is prohibitively difficult to extend in a sustainable and maintainable way.  The current Mirador 3 effort is intended to rebuild and re-architect Mirador to fully realize its potential for Stanford and beyond.

Mirador 3 retains the essence and many of the features that users have come to rely on and expect: 

  • Deep zoom and pan of high resolution images
  • Viewing and navigation of multi-image objects (like books and manuscripts), using thumbnails and tables of contents
  • Display of metadata and rights information for the object
  • Interactive display of annotations of regions of images
  • Comparison of multiple images in the same workspace, even from different objects and from different institutions

Mirador 3 has some key improvements and new features that users and adopting institutions told us they need to support Mirador’s use in discovery environments, instructional and research contexts:

  • Multi-lingual support in the user interface elements and metadata (if it exists)
  • Vertical or horizontal display of thumbnail strips, metadata, and other information
  • Choice between a light and dark theme
  • Multiple workspace modes for comparing images: one called mosaic, which is similar to Mirador 2 and force-fits image windows into adjacent spaces; the other called elastic, which enables the user to move windows anywhere on an expandable workspace, even overlapping
  • Support for authenticated access to images that are protected by institutions using user accounts and other methods (must be compliant with the IIIF Authentication Specification)

A major focus of Mirador 3 was to improve the experience of software developers who want to contribute to and extend Mirador, and of institutions who want to implement Mirador in variety of ways. To that end, Mirador 3:

  • Includes extensive configuration support to turn different features on or off, or change defaults.
  • Has broad theming capabilities to allow implementers to customize the look and feel through themes.
  • Includes a sophisticated plug-in system allowing developers and institutions to add unique capabilities without having to fork the code. 

The final demo of the outcomes of Phase 1, can be seen here :



Please note that this is only the first phase of development and we have labelled the release as an alpha, which signals that Mirador 3 is not fully tested and ready to be implemented in a production system.  Our intention with Phase 1 was to build out the essential and core functionality, as well as the plugin system, to enable a broader range of developers to continue development in a distributed way, and to stimulate the build-out of a plug-in ecosystem.  Mirador 3 still requires testing and feedback from its established community of developers, adopters and implementers.  We welcome and encourage you to test it yourself at the demo site: Implementer-facing documentation can be found on the GitHub wiki.  And please go to to learn how to participate and offer your feedback. In particular, we would be happy to hear from and work with institutions who want to do usability and accessibility testing in different environments.

This first phase of work is only the beginning.  You will notice certain features are not yet implemented, including:

  • The ability to create annotations on images.
  • The ability to view objects that have been digitized with multiple layers.
  • The ability to search full-text within an item, which may have text linked to the image via OCR or transcription. 
  • Navigation and related features supporting right-to-left, and top-to-bottom reading documents and languages.  

We hope and expect that these important features and others will be built by the project team or through contributions to the open source project, leveraging its new plugin architecture.  

Many thanks to the team for a productive and gratifying collaboration. Contributors to Phase 1 included:

From Stanford Libraries:

  • Chris Beer
  • Gary Geisler
  • Jessie Keck
  • Jack Reed
  • Stu Snydman
  • Camille Villa
  • Jennifer Vine
  • Drew Winget

From Universität Leipzig:

  • Leander Seige
  • Mathias Maaß
  • Glenn Fischer
  • Christopher Johnson
  • Richard Großer
  • Annika Schröer

From Princeton University:

  • Shaun Ellis

From HarvardX

  • Luis Duarte

Special thanks should also be given to those members of the IIIF and Mirador Communities who contributed their feedback, ideas and testing throughout the design and development process.