Component: "speaker" (and may be overall "participant") toolbox

[yarikoptic]

Idea came up while discussing on OHBM 2021 TF slack in connection to #10 . It would be nice to compile materials and references which would assist speakers and participants to prepare and provide "high quality" content (e.g. pre-recorded talks). so here are some ideas on "sub-components"

QA toolkit/container

• automated CC system (or integration with some online) which would would assist testing recording for ability to provide CC for the talk given by the speak on own hardware
  • Sample text (with field specific text) to narrate, do CC, provide feedback on how well it could recognize the voice
  • Ability to record the voice and provide CC for fixup to a. assess quality of CC on a particular text; b. submit CC along with the talk
• could include installation of OBS if in the container... For Linux guest systems singularity container would provide turnkey deployment... not sure if would work/worthwhile for OSX and Windows hosts

Documentation

• Tools/Tips/Tricks for video recording. See e.g. debconf video instructions
  • OBS an awesome cross-platform video recording/casting tool. Sample talks: me in PiP, recorded "live", so no post-editing to get PiP, Adina for OHBM 2020 traintrack with multiple windows, Adina as a "human cursor" - I believe post-edited a bit but may be just live
    • I personally use it all the time also for the "live" virtual talks along with v4l patch/plugin to provide "virtual camera" for zoom/jitsi to use. There is a PR to have that feature included in OBS (for Linux only ATM) itself.
• Recommend some commodity hardware (headphones, camera, video) which is known to work well

Aux ideas

• by @raamana -- suggest institutions to setup a good video recording room (greenscreen, good mic, video) etc to facilitate HQ recordings

[raamana]

Thanks Yarik for diligently maintaining these ideas and available software, esp. the open ones.