Delicious LiveJournal Links for 7-8-2008

[andrewducker]

• Why do I look fat in photographs?
  Fascinating article on why the single viewpoint of the camera affects how you look.
  (tags: perception photography vision)
• Love really is blind, or at least blinkered
  Students primed with thoughts of love took significantly less time to identify shapes after viewing an attractive face of the opposite sex, compared with those who had written essays on happiness.
  (tags: love neuroscience psychology science)

Comments

[pizza.maircrosoft.com (from livejournal.com)]

That 'fat' one doesn't explain the common phenomenon where the photographed is horrified at how fat they look, but none of their friends can see it ...

[channelpenguin.livejournal.com]

never mind fat - why do I look more *wrinkly* in photos than in the mirror?? :-)

[call-waiting.livejournal.com]

I'm really not buying the "Why do I look fat in photographs?" argument at all. It has no references, makes some rather large and awkward leaps, takes the dubious step of selectively picking the 'best' bits of background.

A more accurate approach would have been to additively overlay them, which would lead to a narrower central section of background at full insensity fringed by half-intensity sections, which would eliminate the effect of the 'wider' background.



The background in this doesn't seem much wider than either of the individual images to me.

We'd also expect that the effect wouldn't be observable in
stereographic images. I don't have a pair of red/green glasses at the moment, but that doesn't stop me GIMP'ing it:



We'd also expect to observe some corollorary effects; we'd expect that the 'fattening' effect would be more pronounced in photographs, appear (through other cues, eg. focal depth, luminosity) to be far from their backgrounds.

Also, we'd expect that objects which are narrow enough that in real life we'd get a complete picture of the background would exhibit some tendency towards asymptotic behaviour, as the proportion of the amount of background that shouldn't be obscured per unit width of the foreground object increases rapidly at that point.

Moreover, I'd expect that this effect would be small. The claimed effect may be observable for objects the size of a coffee mug, but the effect will be much smaller for objects larger than the 8cm between your eyes. The extra background we'd expect to see always scales to the equivalent of about 4cm at the Z distance of the foreground object, and while this effect may be significant (if it's actually present) for a 10cm mug, it would be less so for a human being say 30cm wide.

(note: I have no idea how wide human beings are, now that I come to think of it...)

I'd expect that this effect would be dwarfed by the combination of very real effects of distortion through perspective (people seldom see themselves in mirrors from the same perspective that cameras see), unfamiliar angles... (for example, I'm a lot balder in photographs than I am in the mirror) all of which sound a lot more plausible as explanations.

There's also the fact, I think, that a lot of people are just too used to seeing pictures of skinny celebrities, and placing themselves in the same analytical framework, they make the comparison.

[ashfae.livejournal.com]

That camera one is really fascinating!