Is the oculus rift sexist?

3D IMAXYet another biological difference between men and women in the brain?

danah boyd wrote an interesting piece in Quartz about her observations that men and women prioritise different depth cues. She has personal experience in a CAVE Automatic Virtual Environment that made her puke and she can’t see IMAX movies.

In short: artificial 3D environments, depth cues have to be programmed in. There are a lot of depth cues, and we don’t need all, but motion parallax is pretty easy to render in 3D, so that gets in. Motion parallax, according to boyd, is the one men’s brains pick out as the most important cue, but women prioritise on shape-from-shading, which is a lot more complicated. Therefore, for men in general, 3D environments work well, but for women, the poor rendering of shape-from-shading causes disorientation and nausea. This phenomenon may also be related to why some transsexuals experience strange visual side effects from their treatment.

If this is the case, there is indeed a problem with 3D technology. dana points out that a lot more research is needed.

I have been in the CAVE of the Centrum for Wiskunde & Informatica in Amsterdam which  was totally awesome, and a very rare experience. No nausea experienced. However, IMAX theatres have been around for quite a while and are common. Some people do indeed experience motion sickness with them, but shouldn’t it have come to light by now if this was something particularly affecting women?

HT to Hank Greely for bringing this to my attention.

4 thoughts on “Is the oculus rift sexist?

  1. It seems a number of variables may come in here, all interesting, but not all acknowledged. Hanah Boyd seems to have hormonal correlations in mind, and to think of the processing of visual depth cues as the site of divergence in response to 3D simulation… And she’s surely right that if a media-technology project complacently focuses on how well it works for (a subset of) men, sexism is not far offstage.

    Yet it’s odd that her hypothesis summary at the end was that her brain “simply wasn’t picking up on signals the system was trying to send me;” the rest of the piece suggests to me that nausea may be provoked partly by the *dissonance* between two kinds of cue. If her brain were not picking up on motion cues at all, she surely would not have been suffering from (perceived-)motion sickness… Instead, her brain was getting visual motion cues of one kind (the kind that strikes many men as “sufficient”), while coping with their *absence* within another visual neural network. In general, I suspect that many motion-sickness phenomena involve the incongruity of sensory signals across channels, which is a particular kind of stress on our nervous system… And here, gender-attentive (non-essentialist) research on something like “field independence” and “field (inter)dependence” — if researchers can manage to avoid treating field independence as if it were intrinsically a positive trait! — may also be interesting. Generally, it’s not “motion” that makes anyone sick, but conflicts between visual cues, balance cues, and proprioceptive-kinetic cues (one is “sitting still” yet also moving).

    There may be lots of roughly-gender-correlated influences on vulnerability to motion sickness in general (anecdotally: having been through a gestational hormone phase seems to correlate somewhat with greater vulnerability to dizziness, sea-sickness…). Clearly, the inner ear is also a site of some interesting hormonally-correlated variations. If we can learn to talk about embodied variation without oversimplifying (about what “male” and “female” mean, or about how rigidly “programmed” these differences are), that will be a big step from the current state of these fields…

  2. The title of this post is a bit unfair, given the content. The Rift is specifically designed to work with PC and game consoles on a 1;1:1 user, headset, and device basis, and is primarily intended for games. If any configuration holds promise for prioritizing different cues for different groups or individuals, it’s one like this.

    First, while shading is more technically challenging than (some kinds) of motion parallax, it has been a focus of 3d rendering research from the start. The Phong shading model was developed in a thesis turned in in 1973, and Jim Blinn’s early NASA fly-by videos show increasing sophistication in their shading over time (Here is a very early one (1978) where the shading applies primarily to Jupiter and the moons: http://www.youtube.com/watch?v=o4xIJlEV8Kw and a later one (1980) showing more on the craft itself as it rotates: http://www.youtube.com/watch?v=zd9TOvFelFg (most evident starting around 3:08). It is considered a primary criterion for realism in rendering. Similar depth cues, such as depth from a repeating pattern shown at an angle, are easily achieved by using high-resolution texture maps, which have been standard for many years now.

    And even if it’s true that current systems have not been developed to render a good set of cues for female users (which is very easy to believe, even setting aside the tiny number of personalized 3d systems sold so far), your assertion about the extent to which cues must be “programmed in” doesn’t apply very well to the systems the Rift is likely to be used with. Most gaming platforms specify virtual environments in terms of shape (approximated by fine meshes of polygons) and textures. The rendering engine then does the work of transforming these into (possibly stereo) images, which allows the process to be tuned at run-time in many ways. Many games even allow replacement of textures, shapes, and even animated characters, with users trading alterations in the form of software patches.

    It may be that while members of both sexes suffer from motion sickness using common 3d technologies, women are more susceptible to the problem. But addressing that problem is one of Oculus’s greatest concerns. Most of the technology improvements in its intermediate and much-beloved “Crystal Cove” prototype addressed this problem (See http://www.wired.com/2014/01/oculus-rift/ ). (3D rendering is now sufficiently accurate to deliver any of the pictorial depth cues. The main remaining gap appears to be in discrepancies between head and eye position and the image seen. In films, the camera zooms and pans without a viewer’s head moving, or worse, with the viewer’s head moving out of sync. Many gamers have trouble playing non-stereoscopic 3d games for longer than a few minutes because of the way the view changes on the monitor as they “move around”. The more recent Rift prototypes have two features — head tracking and a “low persistence display” — targeted at this problem.)

    This has probably been more detail than necessary, but I thought you should know that the 3d gaming community is both focusing on this problem, and is probably in the best position to solve it for either sex.

  3. Thank you for your comments, Elise and Skef. I completely agree with you and danah boyd that this is something that should be further investigated.
    Skef, thank you for your elaborate reply, it is very good to know that the gaming community, which does have a few more problems with sexism in other realms, is looking into this. I would argue, though, that rendering in itself is programming, and although my rephrasing of danah’s piece is a bit coarse, whatever is rendered is still something that is programmed in.

  4. That makes sense. The point I was trying to make is that unlike with a film, person A can play a given game with one set of rendering parameters, and person B with another. So if A and B need different things to stand out or even to be omitted, the overall framework can easily handle that.

Comments are closed.