Built-in audio & mic checks | Voters

Built-in audio & mic checks

complete

Monica Wagner

Like on Skype or Zoom! Easy tests that all devices are functioning properly that give participants infinite opportunities to fix any issues. For the recording check, it would be nice if it automatically detected sound instead of the participants having to check themselves.

August 14, 2020

Jade Pickering (Head of Product)

marked this post as

complete

We've had the 'subsystem' checks in Task Builder 2 for a while now, which automatically checks when a feature uses browser hardware (audio, microphone, camera, full screen, and screen sharing). You can see more here: https://support.gorilla.sc/support/tools/task-builder-2/how-to#subsystems

I'm going to mark this as Complete!

Magdalena Kachlicka (BBK)

Jade Pickering (Head of Product) 👍👍👍

Will Webster

marked this post as

under review

Will Webster

Hey all! For the interim, rather than make changes to the existing zone or create a new zone within Gorilla, we've decided to implement some testing functionality in a Code Editor task. This will make it easier to update/rework the content, as expect it could take a few iterations before we get it right and catch as many of the issues that can occur as possible. My hope is to have this Code Editor task available early next week!

Will Webster

marked this post as

planned

This is now on the internal development map for this month! As has already been pointed out, there are limitations on what can be reliably done within the browser. However, we'll be improving the error detection systems the existing zone has and adding capabilities for you to branch participants out who fail certain checks!

Will Webster

marked this post as

under review

Ferenc Igali EDU

How would you automatically detect sound? If I set a threshold, that threshold could be reached by noise on the mic line (or a blip on gain, someone coughing, etc). There's a reason why Skype, Teams, Zoom and all others make you listen back to yourself - it's to make sure that the recording is clear. The worst thing would be is if thinks it has detected sound and then records patchy and useless audio (and therefore uses up a token). I like the idea and am voting on it (but coming from a programmer's perspective, I have no idea how we would 'automatically' detect sound without a lot of potentially false results).