We had the pleasure this week setting up an Augmented Reality (AR) test for children. While two of our researchers, Michael Brown and Heather Nam have been solving how to achieve our client’s objectives, I have been working on setting up the lab. We often get asked by independent researchers and our clients how we prepare technically for all of the various types of studies we encounter. I thought I would take a break from my typical day to day, and write up how we tackled our latest study from a purely technical perspective.
Our Challenge was that this study involved testing hand held video games, mobile, desktop and web products all in one sitting. Our facility is often and easily configured for mobile testing, but this went beyond our typical mobile setup. We had over a dozen people observing the study and we needed to record the session so that Michael and Heather could go back and cut clips. We needed six different views. Where we typically could use one computer and its attached USB camera, we needed two computers for this study since one AR application required XP, and its own proprietary USB camera. We wanted a “play” view of the room, so viewers could see an angled shot of the kids playing with the camera (something our observers could also see if they choose to watch behind the glass). We also needed an over the shoulder view of the device being used, something we already do with mobile studies. We typically will not attach cameras to the devices, since we feel that it places too much restriction on the user, from either holding it naturally or just looking at it. The disadvantage? The children participating could move and squirm out of camera angle, not a big deal for the researchers, since they are watching live, but a real problem for the observers. We overcome this by either adjusting the camera, or asking the kids to play with the device on the table.
Morae couldn’t be our only solution, we needed a redundant setup. We used Morae to record the sessions, and used HD video cameras as the feed for the observation room. Why didn’t we use Morae for both? Well, we do on some studies. Our issues with Morae to broadcast the session are two-fold. One, it is very tricky to have multiple cameras with Morae. We upgraded the lab machine to an i7 processor and it was still taxed. We needed 3 USB cameras recording in addition to capturing the screen for the desktop. We had to switch study configurations halfway through the study, which wasn’t ideal, but would have been worse if our clients had to wait for it to end, write the file, and restart. Secondly, Morae at its best video quality (which isn’t great on a dual camera capture) buffers anywhere from 15-30 seconds. This means you have to wait that long after a study has concluded before ending the recording, or you will lose the end of the study. We didn’t have that problem with another video setup running parallel.
For the live feed we mirrored the overhead and play USB cameras that we have setup for Morae. We are using two HD cameras, and one 40-inch LCD panel and an overhead projector. The projector was great since it had the picture controls to turn the image upside-down which was required for our over the shoulder shots. Our other 40-inch LCD had a mirror of one of the computers. Live feeds are great, and have been used by the old market research firms for being rock solid dependable. We love our Morae picture-in-picture digital recordings that can be viewed anywhere in the office, but we aren’t Morae fanboys. We felt in this case it had reached its capacity. We will go back to Morae for 90% of our other studies, but between wonky network issues and video quality, we like the HDMI feeds.
What will we do next time? We did run out of wall space. I could have projected another image, but we really didn’t have the space for it. My next purchase is a PIP digital matrix to help optimize our wall real estate by running 2 live images in one display.
If you are interested in general tips for testing or designing for children, see what I wrote up about testing them, and see Heather’s article that was in UXMatters about designing for children.