Tag Archives: High Fidelity

Lag contributes to a bad first impression

We all know this situation, don’t we? I’m chatting with someone online — maybe in text, maybe via avatar, maybe with video — and the pauses get awkward. We start talking over each other, then the silence grows while we wait for openings. In video, I’m not always sure if my friend has simply stopped moving, if there’s packet loss, or if the connection has dropped.  In text or avatar chat, maybe it’s lag, maybe the other person has crashed but hasn’t been signed off yet, maybe there is something happening offline that is taking attention, or maybe the conversation is simply going poorly.  Even when I know the other person well, this can lead to misunderstandings and confusion.

Research at the Technical University of Berlin that is published in the May issue of the International Journal of Human-Computer Studies used artificial transmission delays to examine how strangers reacted to delays in video chat. From the summary of the study on Slate’s Future Tense blog today:

People feel they are being interrupted more often even if the person they are talking to didn’t intentionally speak out of turn. Sometimes both people just stop talking, and no one knows who should continue. The conversation gets confused, and people need to explicitly state whose turn it is to talk. It’s all very far from the way conversation flows in real life.

The awkwardness is heightened if people are unfamiliar to each other. They have no prior knowledge about the other person’s personality or how they normally speak. …

For transmission delays of 1.2 seconds, the interlocutor was rated as less attentive, friendly, and self-disciplined than if there was no delay. Our research has also shown that the initial interaction speed of a conversation is essential for how much people realize that there is a technical problem.

It’s not surprising to me that there would be negative consequences, but those are some nasty traits to be rated poorly if one is using video chat in a business context. It emphasizes the importance of controlling latency, as Philip Rosedale covered at the VWBPE conference, for VR to move to a mainstream form of interaction.

Leave a comment

Posted by on May 12, 2014 in In the News, Research


Tags: , ,

High Fidelity’s proposed system architecture


Philip Rosedale posted this on his High Fidelity blog last week and the comments have been thoughtful and constructive so far; be sure to read them if you click over to see Rosedale’s summary. This diagram is a bit out of my depth.  My husband has been building a tool for me in SL and has voiced surprise about how the current architecture there contributes to lag problems; I’ll hope that critically-thinking network minds are examining this closely.

I find myself most interested about topics that came up in the comment stream, like those about cryptocurrency and the idea of setting an upper byte limit for avatars per domain, based on mesh plus textures plus LOD (level of detail). I was mortified recently to discover that my render weight for the low-script outfit I wore to a conference was shockingly high.  I’ve become a bit obsessed with looking at the render weight of avatars wherever I go and keeping mine in the green or yellow range when I leave home. Some sims in SL will ban entry based on script count, but I haven’t come across any that evaluate based on render weight. Penny Patton wrote a thoughtful post last year about ways that creators and builders could improve the experience for all; much of it coming down to the pixel size of textures and physical scale. Individual residents can also use those tips if they have some tech-savvy, but if HiFi plans to allow an upper byte limit, they’ll need to make sure that default behavior by average users does not exceed whatever emerges as the community standard. Starter avatars wearing freebies are lag monsters in SL now. Rumor is that SL plans to offer mesh avatars alongside the standard starter avatars soon (with Ebbe Linden’s avatar being one of the new models coming soon), but while those might look nicer for less render weight, they add to the already complex problem of finding/making clothes and accessories that fit.


My avatar today (un-Photoshopped) with a render weight of 19494.

Leave a comment

Posted by on April 29, 2014 in Side Topics


Tags: , , ,

Philip Rosedale at Virtual Worlds Best Practices in Education 2014

When I can this week, I’m attending the VWBPE conference in Second Life and on the OSGrid. The keynote talk today was given by Philip Rosedale, CEO of High Fidelity and founder/previous CEO of Linden Labs.


Philip Rosedale speaking at the conference

He began by talking about the book Ready Player One, which I haven’t read yet. Guess I should, eh?

Much of his talk was about the restrictions of our current ways to move and communicate in virtual worlds and how we are so constrained by the mouse and keyboard. With today’s limitations, it is challenging and there is a steep learning curve to have rich communication.

Here are a few key points that he explored that resonated with me: Read the rest of this entry »


Posted by on April 9, 2014 in Embodied Experience, In the News


Tags: , , , , , ,

%d bloggers like this: