AV Artifact Atlas: By the People, For the People

In this interview, FADGI talks with Hannah Frost, Digital Library Services Manager at Stanford Libraries and Manager, Stanford Media Preservation Lab and Jenny Brice, Preservation Coordinator at Bay Area Video Coalition about the AV Artifact Atlas.

One of my favorite aspects of the Federal Agencies Digitization Guidelines Initiative is its community-based ethos. We work collaboratively across federal agencies on shared problems and strive to share our results so that everyone can benefit. We’ve had a number of strong successes including the BWF MetaEdit tool, which has been downloaded from SourceForge over 10,000 times. In FADGI, we’re committed to making our products and processes as open as possible and we’re always pleased to talk with other like-minded folks such as Hannah Frost and Jenny Brice from the AV Artifact Atlas project.


That’s not just a pretty design in the AVAA logo. That’s a vectorscope displaying SMPTE color bars. Photo courtesy of AVAA.

The AV Artifact Atlas is another community-based project that grew out of a shared desire to identify and document the technical issues and anomalies that can afflict audio and video signals. What started out as a casual discussion about quality control over vegetarian po’boy sandwiches at the 2010 Association of Moving Image Archivists annual meeting, the AV Artifact Atlas has evolved into an online knowledge repository of audiovisual artifacts for in-house digitization labs and commercial vendors. It’s helping to define a shared vocabulary and will have a significant impact on codifying quality control efforts.

For an overview of AVAA, check out The AV Artifact Atlas: Two Years In on the Media Preservation blog from the Media Preservation Initiative at Indiana University Bloomington.

Kate:  Tell me how the AV Artifact Atlas came about.

Hannah: When we get together, media preservation folks talk about the challenges we face in our work. One of the topics that seems to come up over and over again is quality and the need for better tools and more information to support our efforts to capture and maintain high quality copies of original content as it is migrated forward into new formats.

When creating, copying, or playing back a recording, there are so many chances for error, for things to go sideways, lowering the quality or introducing some imperfection to the signal. These imperfections leave behind audible or visible artifacts (though some are more perceptible than others). If we inspect and pay close attention, it is possible discover the artifacts and consider what action, if anything, can be taken to prevent or correct them.

The problem is most archivists, curators and conservators involved in media reformatting are ill-equipped to detect artifacts, or further still to understand their cause and ensure a high quality job. They typically don’t have deep training or practical experience working with legacy media. After all, why should we? This knowledge is by and large the expertise of video and audio engineers and is increasingly rare as the analog generation ages, retires and passes on. Over the years, engineers sometimes have used different words or imprecise language to describe the same thing, making the technical terminology even more intimidating or inaccessible to the uninitiated. We need a way capture and codify this information into something broadly useful. Preserving archival audiovisual media is a major challenge facing libraries, archives and museums today and it will challenge us for some time. We need all the legs up we can get.

AV Artifact Atlas is a leg up. We realized that we would benefit from a common place for accumulating and sharing our knowledge and questions about the kinds of issues revealed or introduced in media digitization, technical issues that invariably relate to the quality of the file produced in the workflow. A wiki seemed like a natural fit given the community orientation of the project. I got the term “artifact atlas” imaging guru Don Williams, an expert adviser for the FADGI Still Image Working Group.


This DV Head Clog artifact may be the result of a clogged record head when taping over a recycled piece of tape. Photo courtesy of AVAA.

Initially we saw the AV Artifact Atlas as a resource to augment quality control processes and as a way to structure a common vocabulary for technical terms in order to help archivists, vendors and content users to communicate, to discuss, to demystify and to disambiguate. And people are using it this way: I’ve seen it on listservs.

But we have also observed that the Atlas is a useful resource for on-the-job training and archival and conservation education. It’s extremely popular with people new to the field who want to learn more and strengthen their technical knowledge.

Kate: How is the AVAA governed? What’s Stanford Media Preservation Lab’s role and what’s Bay Area Video Coalition’s role?

Hannah: The Stanford Media Preservation Lab team led the initial development of the site, which started in 2012 and we’ve been steadily adding content ever since. We approached BAVC as an able partner because BAVC demonstrates an ongoing commitment to the media community and a genuine interest in furthering progress in the media archiving field.

Jenny: Up until this past year, BAVC’s role has primarily been to host the AVAA. We’ve always wanted to get more involved in adding content, but haven’t had the resources. When we started planning for the QC Tools project, we saw the AVAA as a great platform and dissemination point for the software we were developing. Through funding from the National Endowment for the Humanities, we now have the opportunity to focus on actively developing the analog video content in the AVAA. The team at SMPL have been a huge part of the planning process for this stage of the project, offering invaluable advice, ideas and feedback.

Over the next year, BAVC will be leading a project to solicit knowledge, expertise and examples of artifacts found in digitized analog video from the wider AV preservation community to incorporate into the AVAA. Although BAVC is leading this leg of the project, SMPL will be involved every step of the way.

Kate: You mentioned the Quality Control Tools for Video Preservation or QC Tools project. How does the AVAA fit into that?


This tracking error is caused by the inability of video heads to follow correctly the video tracks recorded on a tape. Photo courtesy of AVAA.

Jenny: In 2013, BAVC received funding from the NEH to develop a software tool that analyzes video files to identify and graph errors and artifacts. You can drop a digital video file into the software program and it will produce a set of graphs from which various errors and artifacts can be pinpointed. QC Tools will show where a headclog happens and then connect the user to the AVAA to understand what a headclog is and if it can be fixed. QC Tools will make it easier for technicians digitizing analog video to do quality control of their work. It will also make it easier for archivists and other people responsible for analog video collections to quality check video files they receive from vendors, as well as accurately document video files for preservation. The AVAA, by providing a common language for artifacts as well as detailed descriptions of their origin and resolution (if any), helps serve these same purposes.

Kate: My favorite AVAA entry is probably the one for Interstitial Errors because it’s an issue that FADGI is actively working on. (In fact, when I mentioned this project in a previous blog post, you’ll notice a link to the AVCC in the Interstitial Error caption!) What topics stand out for you and why?

Jenny: When I first started interning at BAVC, I was totally new to video digitization. I relied heavily on the AVAA to help me understand what I was seeing on screen, why it was happening and what (if anything) could be done. The entries for Video Head Clog, Tracking Error and Tape Crease hold a special place in my heart because I saw them often when digitizing, and it took many, many repeat views of the examples in the AVAA before I could reliably tell them apart.


Debris on video heads can prevent direct contact with the videotape and result in an obscured image or a complete loss of image. Photo courtesy of AVAA.

Hannah: There are so many to choose from! One highlight is SDI Spike, because it is a great example of a digitization error – and pretty egregious one at that – and thus demonstrates exactly why careful quality control is critical in preservation workflows. The DV Head Clog entry is noteworthy, as the clip shows how dramatic digital media errors can be, especially when compared to analog ones. Other favorite entries include those that give the reader lots of helpful, practical information about resolving the problem, as seen in Crushed Setup and Head Switching Noise.

Kate: Where do you get your visual examples and data for the Atlas? Are there gaps you’re looking to fill?

Hannah: Many of the entries were created by SMPL staff, drawing on research we’ve done and our on-the-job experience, and most of the media clips and still images derive from issues we encountered in our reformatting projects. A few other generous folks have contributed samples and content, too. We are currently in the process of incorporating content from the Compendium of Image Errors in Analogue Video, a superb book published in 2012 that was motivated by the same need for information to support media art conservation. We are deeply grateful to authors Joanna Phillips and Agathe Jarczyk for working with us on that.

Our biggest content gaps are in the area of audio: we are very eager for more archivists, conservators, engineers and vendors to contribute entries with examples! Also the digital video area needs more fleshing out. The analog video section is pretty well developed at this point, but we still need frames or clips demonstrating errors like Loss of Color Lock and Low RF. We keep a running list of existing entries that are lacking real-life examples on the Contributor’s Guide page.

Kate: I love the recently added audio examples to augment the visual examples. It’s great to not only see the error but also to hear it. How did this come about and what other improvements/next steps are in the works?

Hannah: Emily Perkins, a student of the University of Texas School of Information, approached us about adding the Sound Gallery as part of her final capstone project. Student involvement in the Atlas development is clearly a win-win situation, so we encourage more of that! We are also currently planning to implement a new way to navigate the content in terms of error origin. The new categories – operator error, device error, carrier error, production error – will help those Atlas users who want to better understand the nature of these errors and how they come about.


The new Sound Gallery feature allows users to hear examples of common errors. Photo courtesy of AVAA

Jenny: As part of the NEH project, we want to look closely at the terms and definitions and correlate them with other resources, such as the Compendium of Image Errors in Analogue Video that Hannah mentioned. We also want to include more examples – both still images and video clips – to help illustrate artifacts. As QC Tools becomes more developed, we want to include some of the graphs of common artifacts produced by the software. The hope is that users of the AVAA or of QC Tools will have more than one way to identify the artifacts they encounter.

Kate: It can be challenging to keep the content and enthusiasm going for community-based efforts. What have you learned since the project launched and how has it influenced your current approach?

Hannah: So true: keeping the momentum going is a real challenge. Most of the contributions made to date have been entirely voluntary, and while the NEH funding is a welcome and wonderful development – not to mention a vote of confidence that the Atlas is a valuable resource – we understand fully well that generous donations of time and knowledge on the part of novice and expert practitioners will always be fundamental to the continued growth and success of the Atlas.

It definitely takes a core group of committed people to keep the momentum going and you always need to beat the bush for contributions. In our day-to-day work at SMPL, it has come to the point where I routinely ask myself about a problem we encounter: “is this something we can add to the Atlas? Have we just learned something that we can share with others?” If more practitioners adopted this frame of mind, the wiki would certainly develop more rapidly! I also try to remind folks that you don’t have to be an expert engineer to contribute. Practical information from and for all levels of expertise is our primary goal.

Kate: Is there anything you’d else like to mention about AVAA?

Jenny: We’re hiring! Thanks to funding from the NEH, we are able to hire someone part-time to work exclusively on building out content and community for the AV Artifact Atlas. If you are passionate and knowledgeable about video preservation, consider applying. We’re really excited to hire a dedicated AVAA Coordinator and to see how this position will help the Atlas grow!