13 More Playtest!

Happy Easters! On Monday and Wednesday, we have conducted two more play-testing in our project room and hall of architecture, from which we received a great number of helpful feed back from Cari Marty, Jibran Khan, Yiheng Zhu, Mike Christel and John Dessler. The help is much appreciated.


Recording video clips during the playtest, we have analyzed the video and gathered testers’ opinions together.


These videos reflect the three procedures of our playtesting, including free navigation, Q & A, and opinions on new UX features.



Abstracted from the videos, the feedback can be summed up as follows:

In order, play-testers were motivated by:
(1) find what objects we have more details for;
(2) see how we overlay virtual with real (the AR)
(3) see how we tie in models and give me the opportunity to “zoom out” of an artifact and see its full context;
(4) see the casts’ details up close by letting me zoom into fine detail
Because of their emphasis on those points, some of them neglected a chance to drop into story and read through what we offer – people would do that if they had more time, and as they suggested, such information is a great hook to get people to visit the website later and learn more about mausoleums, Turkish Muslim architecture, the Knights of Malta, etc.  We offered more than they had a chance to sample without keeping people in the museum for too long (and people did notice a watch check so they wanted to stay on time).
Mike thought we may have hit him with too much all at once (or he missed the intro of the interface element overlays), e.g., with his focus on the casts he really wanted to see and interact with them first.  The blue dots obscured a good portion of the cast and he thinks might have come onto the screen immediately with the artifact.  If the cast came on first, undecorated, and then people zoom in to look at its detail, no blue dots or instruction text obscuring, great.  If people do nothing for a few seconds, then introduce a screen toggle (with a clear toggle button, not a slider looking UI icon) to turn on/off hotspots so that when on as they would be when first introduced to the screen, blue dots, instruction text appear and he can click to get details on the hotspot, when turned off he gets to inspect the artifact clearly with no obscuring.  That might have helped, for two reasons:
(a) we gently increase the complexity of the interface: the target cast first by itself, and then some information overlays like hotspots;
(b) we let them selectively turn off layers (e.g., hotspots) if I want them out of the way to visually inspect the artifact via rotation, scaling to get fine detail, etc.
Moreover, some of our testers are not sure why it took them a bit of time to appreciate that they had to put the model on the floor to grow it.  They did see the red-while-on-wall green-on-floor signalling but didn’t notice any helper text if we introduced such on the screen at lower left.  We’ll need to make the entrance of such tips text more dramatic visually if we want to take testers’ eyes away from the real world and from the cast in the screen display or its model wireframe.
Biggest missing piece for testers: understanding how the cast fits into the larger whole, e.g., the first cast flat piece might be an exterior decoration above an entrance, a floor panel, a wall panel.
Biggest missed opportunity for testers: taking a picture with the model and the room (probably easier if they had their family with them rather than trying to attempt a selfie of such – they could take them posed with the model) – they are glad we have photo taking options and need to check it out later.
Most enjoyable moment: Many, including having the Tower of the Winds grow when placed on the floor, and then walking around and into it.
As a result, we are cleaning up UI including shining hotspots, highlight casts in buildings, hand gesture/other UX tips, re-designing of markers/navigators, font size, style of toggle and camera range.
​Finally, there will be a document to let clients know how to operate it.​