Collaborative Computing Lab (CCL):

Overview of Lab Setup

Purpose: The purpose of the document is to give a general overview of how things are set up. This will help you to visualize how things are wired up...which will help if you need to figure out a problem on the fly.

Overview: What is the setup?

The overall setup of our usability lab is based on a simple 2-room model: End-user testers are seated at the machine in Rm104a; this is what we'll call the Testing station. The idea is that test users are tranquilly along in Rm104a with your software running on the big screen of the station, on a laptop you've got set up with it, or on a mobile device your software is running on. Just outside in Rm104 there is a machine that has some fancy A/V capture hardware and software on it, which we will call the monitoring station. This is where the audio/video of the users using your software is monitored and recorded, and is where the testing team (that’s you) is seated while the user test is in progress. Audio and video are piped through the wall to Monitoring Station.

Video:

In rm104a, there is a camera on a high tripod designed to shoot over the shoulders between two end-users seated at the testing station. Although this is a digital camcorder, it does not record anything locally on the camera; it merely feeds video down the wire and to the monitoring station. Moreover, it does not record audio of any sort; such audio would be miserable, with the mics being behind the users on the camera. Thus, again, all this camera does is send video down the wire.

Audio:

High-quality stereo audio is provided via two wireless lapel mics (one mic/channel for each test user). The wireless mics send to a receiver which is outside in Rm104, on top of the Monitoring Station. These signals are then passed through a mixing board for modifcation and amplification to "line level", and then into the A/V capture hardware to be dispolayed on the Monitoring Station screen (and potentially recorded). The lapel mics have battery-powered transmitters attached, and the mixing board is powered via an AC adapter.

Recording:

Back in the old days, all video recording relied on VCRs of some sort. This generally worked well (very stable, reliable), but recording quality was limited to the standed NTSC resolution of standard television signals. There were several problems with this. First, the resolution of NTSC video is very limited: just 525 "scan lines". This meant that small details on the recorded image, e.g., where the mouse pointer is, what the value inside of text and checkboxes is, etc., was impossible to discern. This was not fatal, however, and we found plenty of ways to work around it (zooming in more, etc.). Another problem was the medium, namely videotapes. These media were clunky and a bit inconvenient, since you actually needed a VCR to review the recordings. And of course, later, after all testing was completed, the tapes had to be erased or destroyed for privacy reasons.

 

In our modern usability lab, all recordings are made as high quality video clips on the computer. The digital recordings are made by special hardware/software called “Eye-TV”. Once launched, the controls are obvious…pretty much like a VCR. The resulting video file is initially stored and visible in the “Library” window for Eye-TV, named with the time and date of the recording. You should select it and EXPORT it under and appropriate name for your team/session, storing it onto a USB stic or portable HD. Please delete the files from the monitoring computer once you have them stored on your own mass storage medium --- for both data privacy, and for space reasons. These are 1080p files, i.e., they get big fast.

How to retrieve your recorded session:

Bring a USB stick or other storage medium and copy the end-user data file to it. They can easily run into the multiple GB, so make sure you have a large enough medium --- I'm guessing an 8GB stick would be good for the average session. Alternatively, you can SFTP the file somewhere, or mount your shared drive. Email won’t work; the files are too big!.