Recent dispatches from the outside world...
PlayOn: Collaborative building in Second Life
The legendary Palo Alto Research Center have been conducting studies in Second Life, and recently published some of their findings on their blog, PlayOn. To investigate collaborative building, PARC researchers observed teams of volunteers completing specific building tasks.
One of the interesting observations, however, wasn't specifically about building: Most SL users, they discovered, do not maintain the default camera position of looking over their avatar's perspective (or "in-avatar"). Instead, only 43% of their volunteers remained in-avatar, while 57% used their camera controls to move away from their avatar's position (or "in-camera").
What's more, this phenomenon changed drastically in direct relation to experience:
We divided our users into groups based on experience and found that experts spent 90% of their time in-camera, while novices stayed within their avatars most of the time. (Emph. mine.)
The PARC researchers were curious if this caused users to disassociate themselves from their avatars, but that wasn't the case. Instead, when speaking to other Residents, they still referred to objects in-world in relation to their avatars, as opposed to their camera's location. (As in, "The wall behind me, the hill to my left," etc.)
This is striking, and should change our understanding of social interactions in Second Life. From my perspective, their findings (while limited to a small group) resonate with me. When I'm in SL, I constantly move back and forth between in-avatar and in-camera, even when I'm talking with someone standing right in front of me. However-- and this is a crucial point, I think-- I still "feel" that person's proximity to me. In other words, for me the experience of Second Life is simultaneously both subjective to my perspective and objective to my environment. I'm curious to find out if this is a common perception.
As for collaborative building in SL, here's the nugget in PARC's highly sensible, metaverse-centric conclusion:
When people build things together in real life, they frequently have to collaborate synchronously around individual objects: for example one person might hold two pieces of wood in place while another screws them together. In previous research, users were sometimes forced to collaborate in this way, by implementing gravity and friction into the virtual environment. But gravity and friction are usually not necessary in virtual worlds. Experienced Second Life builders make use of the fact that objects will stay hanging in the air, or even pass through other objects. Rather than force real-world modes of collaboration, it might be better to embrace the non-physicality of virtual worlds, allowing people to avoid the kind of synchronous 3d collaboration that the physical world forces upon them.
Image: blogs.parc.com/playon/. Hat tip: Jeanricard Broek.
I don't think it's really any different than what one does when using binoculars or a telescope in RL--it takes something far more immersive to make one shift one's point of view.
Not being given to out-of-body experiences, the default camera position seems as arbitrary to me as any other. Mouselook, which I would think of as really being a "natural" point of view, has limitations that preclude me from using it very much.
Posted by: Melissa Yeuxdoux | Wednesday, October 22, 2008 at 05:48 PM
Embodiment, which SL forces on us by putting us in our avatars, is one of it's greatest strengths. When that is solid one retains situational awareness of the spatial relations around one's avatar and can use the camera, just as one looks around one's own body in real life without losing spatial perspective.
This is one of the problems with the Google Lively way, where the avatar is on the other end of your camera view. You are never entirely aware of yourself spatially. When it looks like i'm talking to someone on my screen it may seem that i'm talking to someone else to others. It keeps avatar interactions at the puppet level rather than true embodiment, and misses the key strength of virtual worlds. We aren't there to look at ourselves, but to look at what is around us. Not being embodied in our avatar as a point of reference severely curtails our angular, depth and general spatial awareness.
As for the Parc study, most of their findings seem valid, but there's a few bugs in there. In SL you _can_ tell where somebody is looking when they are in camera view - they are looking at the point which they used as the pivot point for their camera. This information is lost when folk wear an animation override device, so most pro builders I know do not use these when collaboratively building so that other builders know precisely what they are looking at.
Because the camera can move from that point to look at other things, old builders deliberately focus when drawing attention to a point at the actual thing they wish to look at rather than an arbitrary point in space which may be convenient. When not specifically drawing attention to a point in space builders will mostly be looking in the general direction of what they're working on, and will usually only tightly focus their attention on an exact point should they wish to convey this information to others. As SL is still fairly young this behavior may not have developed in their test group, but it certainly exists and these findings are based on a misunderstanding of the finer points of the SL camera system.
There's a point I'd like to address in their conclusion also. One of the fundamental rules of design is to ensure that elements you put in are unsurprising unless you wish them to draw attention. We're not all Jonny Mnemonic flying around in the pure data feed of cyberspace... we need solid reference points. Lack of reference causes two problems.
Firstly, psychologically speaking, things floating in space are somewhat surprising. We may get used to them, but on an unconscious level there will always be something slightly "wrong" or uncanny about them. Having talked to folk in SL long enough (where floating objects are all over the place) it is obvious to me that folk are aware of this, but it can take extended scrutiny (untrained as we are in floating things) for them to figure out what exactly is out of whack. It may be perhaps barely perceptible, but distracting.
Even if these physics bending glitches are only subtly psychologically distracting it is better to be aware and eradicate the distraction. There are enough distractions already in working in a virtual world, and if all the small ones are ignored they'll just mount up into a poor user experience - on the surface it has all the right ingredients but in use somehow frustrates.
Expert builders of course will be able to work floating in space, but working collaboratively requires strong spatial references; this means the horizon plane, floors, walls etcetera, without which we can not see perspective. Our everyday life experience and our amazing visual systems in the eye and brain prepare us for dealing with normal 3d space. As soon as you take the normalcy away our visual perception is severely compromised.
A simple way to see this is to go to a sandbox and rez a prim on the ground - this is an easy task even for a newcomer... hit the build button and click on the ground. A prim! Walk around a bit and then turn and look at the prim. Because it's on the ground you will have an immediate sense of how far away it is from you. Now if you're still in build mode, grab the little arrow pointing up and drag the prim into the air. If you walk around a bit and look at it again you will find it much harder to figure out how far away it is. You may feel a subtle confusion about it, as even though perspective is still there and the size should tell you about the distance, something critical is missing. It's the ground, pure and simple. (Please now dispose of your prim properly :)
So sure expert builders can build floating in space and so on, but generally they shouldn't. Doing so loses familiar subtle spatial reference points which our many years of life on earth have trained us to understand. Builders collaborating without this spatial reference tend to make more mistakes in terms of scale and proportion. As soon as these builds are put into a more naturalistic context their faults become embarrasingly obvious. (Even if the faults are subtle they at the very least become distracting.)
Just my two cents as usual :)
Posted by: Pavig Lok | Thursday, October 23, 2008 at 01:55 AM
Pavig, the sort of skillset you're speaking off has become more prevalent as more people enter Second Life who are already acclimatised to other more limited forms of working in 3D (e.g. first-person shooters and 3d-based MMOs, as well as 3D generative applications such as Blender et al.)
I would fret a little, but errors such as parallax error that occur within the collaborative realm could be reduced or removed by more regular checking of prim coordinates, the use of the Copy Selected tool that is tucked away in the Edit panel along with all the possible prim choices for rezzing, as well as the use of builder tools such as PrimDocker that are built with such collaboration in mind.
Preplanning a build is equally another way of getting everyone on the same page and reducing errors in building as a result. If you know what the distances are and where everything should be placed roughly, you create a sort of sanity check on prim placement. This even allows a bigger build to be apportioned off into smaller pieces for separate teams.
I'd suggest putting attention into ensuring that builders in SL know how to cope with the SL camera controls, what options they have for ensuring that prims match up properly, and the rest will fall into place.
Posted by: Patchouli Woollahra | Thursday, October 23, 2008 at 09:48 AM
> Not being embodied in our avatar
> as a point of reference
> severely curtails our angular, depth
> and general spatial awareness.
Why would that be? Visual perception of a 3D scene works fine from any ecological camera position, whether it coincides with your avatar or not. Constraining the camera to your avatar's head leads to something like tunnel vision, compared to a more flexible camera, which has pros and cons depending on design goals. The immersion and collaboration advantages are more obvious, if that's what you're after.
Posted by: Mark Young | Thursday, October 23, 2008 at 10:23 AM
If the animations in an AO are uploaded at the correct priorities, an avatar should be able to turn its head to look at things and point its arm at what it's editing. The content creators I know who avoid wearing an AO are usually just more paranoid about lag than most, and they also tend to avoid things like prim hair and shoes.
The greatest challenge involved in collaborative building in SL, at least for my purposes as the owner of a content development company, is not anything to do with avatars or camera issues. It's the permissions system.
Posted by: Kimberly Rufer-Bach | Thursday, October 23, 2008 at 11:48 AM
"When I'm in SL, I constantly move back and forth between in-avatar and in-camera, even when I'm talking with someone standing right in front of me. However-- and this is a crucial point, I think-- I still "feel" that person's proximity to me. In other words, for me the experience of Second Life is simultaneously both subjective to my perspective and objective to my environment. I'm curious to find out if this is a common perception."
Yes, this is exactly the way I use the camera in SL, and yet I still feel the same way about my 'position' in the world as you do. I think it is because it's hard to make do with the limited AV perspective once you get good at camming (usually caused by hours spent shopping). You can be talking to someone and examining their face, their boots, a distant object, etc. but you still feel right 'there'. If I may quote from 'Red Dwarf' (from when an android becomes human):
Lister: Any problems?
Kryten: Well, just one or two. In fact I've compiled a little list if
you'll indulge me. Now then, uh, my optical system doesn't
appear to have a zoom function.
Lister: No, human eyes don't have a zoom.
Kryten: Well then, how do you bring a small object into sharp focus?
Lister: Well, you just move your head close to the object.
Kryten: I see. Move your head . . .
Posted by: Valentina Kendal | Thursday, October 23, 2008 at 12:02 PM