The Roslin Institute, University of Edinburgh 13/11/15

On friday the 13-11-15 I visited colleagues at the Roslin Institute to demonstrate the latest version of Clipper to colleagues and get feedback. This followed on from a larger earlier meeting in August. What was striking about this meeting was the overalap with the needs of the Royal Conservatoire researchers that we visited the previous week, this was very encouraging and indicates we are on the right path. Thanks are due to Colin and Peter for being so generous with their time.

Colin Simpson RDM Manager

Peter Hohenstein PI / Group leader – my lab generates a lot of time lapse data – very big data dumps.

  • How can the annotations be displayed? Can they be on the video as well as separately? Can we have graphical annotations as well
    • Yes we have had a lot of interest and requests for that
  • I think this has really nice possibilities for the sort of thing we are doing – we were talking about these basically big dumps of data that are generating. We have a serious problem with our time-lapse data – that keep having to duplicate it in order to access it and work with it – so you have many copies and then errors can creep in. If we could have our data in one central place, even just for internal use, then using something like Clipper it would be extremely interesting.
  • I know you said you were not creating an online video editing tool. But really important for here is first of all we are combining different source files into one view (to compare etc.). If we could have this with these sort annotations on the clip then I would be using nothing else for the rest of my life!!
  • There is a second thing. As a scientist there is only one thing that counts and that is [journal] publishers because we need to get our papers and our data out. I would simply say why not include the links to your [Clipper] system in the paper? That would be fine but the publishers might find that difficult. It would be extremely useful for scientists to be able to export their data in the form of Clipper Clips with the video in the form of mp4 or whatever format along with their papers and send that to the publishers. Not being able to do this kind of thing is what is making our life difficult.
  • What about this as a scenario – we create something along the lines that you say as an online service in the cloud with the video reference from there via Clipper? Yes I see what you mean but the publishers really want the data as well. So being able to include the source video data in an export would be still needed.
  • However the publishers are also going to have a problem with the big data that is coming their way – so they may be amenable to a cloud reference solution as you describe.
  • In the local lab environment it would be useful but if we get it out in our papers that would be ideal
  • Some publishers claim the data, but others just want to reference and for this a DOI is used to reference the location of the data (in the institution or elsewhere)
  • Actually if you could turn these Clips etc. into DOIs and if publishers would accept that then that would be a way into a solution
    • That definitely needs to go on the requirements list!
    • Work with publishers on this
  • We would need the ability to transcode any videos as well for import / export to make sure reviewers could se the video
  • We would also need static high resolution images to go into our papers
    • This is a good example of the need for customisation for the toolkit to different user preference
  • A cloud service that could do all the above would be a real winner, including DOI and guaranteeing persistence
  • The problem with the cloud is local policy issues – if Clipper could be part of one stop solution for researchers in institutions that would be ideal
  • With the huge amounts of data that are being generated there is rapidly going to come a turning point where publishers say ‘no I am not going to take your data’.
  • This could be great internal system to reduce data duplication and external in the external world with journals and DOIs
  • Having some video editing capabilities would be a big plus for us
  • Having a range of options for export in the system would be ideal including – reference web ready video and export the whole video file with the paper – export physical hi-def video clips – create and export hi-res stills
  • If we can export the Clipper documents in web format to the publishers that would save a lot of hassle
  • Annotations need to be searchable
  • File Tagging?
    • Yes we can tag different things from whole video files to Clips, annotations and so on
  • Yes the collaboration at a distance would work – for instance one of my collaborators is now in California and we have something stored here locally but we still need to collaborate. We have been sending files over by hard disk. But something like this would work!
  • Something like could work on other projects that use still images. You should also be talking to the IGMM (Institute for Genetics and Molecular Medicine) they are designing a lot of specialist software tools to support their work and they might be interested in Clipper.
    • There are strong links here to the work that Digirati are doing with the IIIF standard for image annotation
  • If we could have a cloud based system where all the stuff is in one place that would be a massive advantage and if the user could choose the resolution / definition of the content they work with that would be ideal to cope with different locations and connections. That would enable all kinds of flexibility for work and collaboration.
  • Being able to export the whole caboodle in different arrangements and formats is crucial for the publishers and therefor the researchers
  • Being able to support the whole research data lifecycle from the labs to the publishers is crucial
  • Being part of a cloud trial with Clipper would be great