Lighting for videoconferencing
mbietz January 21st, 2010
I spend a lot of time in videoconference meetings. I live in San Diego, but I’m a post-doc in the CSC Lab at the University of Washington, and also have an appointment in the Institute for Software Research at UC Irvine. I attend 3-5 distributed meetings in a typical week. Most of them are videoconferenced, and in most of them, there is a group of people on the other end of the line.
I’ve noticed is that the way I am lit seems to affect how others see me. These meetings are often my only real-time interaction with my collaborators, and I want to be perceived as an important part of the group. Of course, I don’t want to miss what is going on, but I also want my collaborators to know that I am attentive and care about what they have to say. I don’t think good lighting will make it seem like I’m paying attention when I’m not. But I do think that bad lighting can give the wrong impression.
These are 3 pictures of my normal videoconferencing setup taken within moments of each other. In the picture on the left, I have only the regular room lights (an overhead fixture), with the camera set for automatic exposure control. In the middle picture, I’ve turned on the proprietary (and pretty amazing) Logitech RightLight feature, which finds the face and sets the brightness and contrast so that the face will look decently lit. In the picture on the right, I’ve turned on the lights. The colors are brighter, the shadows aren’t as deep, and there’s more contrast with the background. And even though I’m in the same position and have the same expression as the other two photos, I think I look more engaged.
I have a cheap but effective lighting setup. Two 25w Kvart clamp lamps from Ikea ($6.99 each) clamped to chairs positioned at about a 45° angle on the right and left. The fronts of the lamps are covered with tracing paper so the light isn’t quite so harsh. I also have a cheap gooseneck lamp on the floor behind me to provide some backlight. The camera sits on a cookbook stand that’s a little higher than the screen of my laptop (just about at chin level for me). The separate stand also makes it jostle less as I type. When I’m at the computer, there are two windows on my left, and behind me is a translucent curtain in front of a large window, so if it’s sunny outside, I’ll only use one lamp in the 45° position to my right. Even with all 3 lights I can get the whole thing set up in just a couple of minutes.
I use a high-quality webcam, but good lighting helps even low-end cameras.
I haven’t conducted rigorous trials, but I get the sense that when I am well lit, the people at the other end are more likely to include me in the conversation. Good lighting helps me feel less far away.
Facebook has been playing with their friend suggestion algorithms, and people are starting to notice. Last week FB suggested I become a friend with one of the participants in my research. I am pretty liberal with accepting friend requests, but that is a line I try not to cross. But what surprised me was that we have no common friends and have no other common interests or memberships on the site. But in the end my own reaction was mostly one of curiosity – what data sources led them to make this match?
The 
