[Phys-L] physics of mirrors and non-mirrors

[John Denker <jsd@av8n.com>]

Since we are now talking about mirror physics,
it is worth keeping in mind the age-old puzzle:
What does a mirror really do?  Some say it 
reverses left and right ... but if you are
clever you realize that there are problems with
that.
 -- What exactly is wrong with the R/L notion?
 -- Can you give a concise and correct explanation
  of what a mirror really does?

Answer: https://www.youtube.com/watch?v=msN87y-iEx0

On 05/06/2015 07:51 AM, Bob Sciamanda wrote:
> Try shaving (or hair grooming) while looking at your video image on
> your computer (or camera) screen - this is not a mirror image, this
> is how your barber sees you.

That is likely to work for the VoIP (video over IP) 
application on your computer ... but it does not work
for the typical camera app on a smartphone.

Fact:  I have at least six different camera apps on 
various phones.  Precisely none of them give a non-
inverted "barber" image in the real-time viewer;  
they all give the mirror image.  This behavior is
built into the hardware at a rather low level.
   In contrast, however, if you click the shutter, 
   the /saved/ image file is non-mirrored, but that's
   because the bits get swapped by the software.

It turns out there is a good reason for this.  A lot
of non-experts say "people expect it to act like a
mirror" but this is *not* a good reason, as evidenced
by our observations of the VoIP applications.  A
non-mirrored image draws no complaints, and actually
has some advantages.  So then, what is the real reason,
and why does this reason apply more to camera phones 
than to desktop or laptop computers?

If you shop around long enough you can find specialized
apps for your smartphone that exhibit more-or-less the 
same behavior as the desktop VoIP applications.  I hate
to mention it, but here is an example:
  https://play.google.com/store/apps/details?id=com.wckd_dev.mirror
I do *not* recommend it in the usual sense, because I
find it to be phenomenally buggy.  The best I can say 
is that you might kinda maybe sorta get it to do part
of what you want (for a while, until it crashes).  If
anybody knows of a better app with non-mirror mode, 
please let me know.

Leaving aside the problems with that implementation,
there are fundamental physics issues to be resolved.
What _should_ such an app do, to produce a non-inverted
"barber" image?

I claim that fundamentally when you are looking at
the barber, he is looking at you from a different
angle, 180 degrees opposite.  It's a rotation.

It is well known that a mirror produces a parity
inversion plus a rotation.  A proper rotation can
always be produced by /two/ reflections, using
two mirrors with artfully-chosen orientations.

So, in terms of mirrors or otherwise, what is the
non-inverting imaging app /supposed/ to do, and
why is this tricky?  I claim that part of the
problem has a nice solution and part of it doesn't.

Usual jsd puzzle rules apply:  Everything I've
said is (to the best of my knowledge) true and
helpful.  It's not a word game.  The answer does
not come from nitpicking the question;  it comes
from understanding the fundamental physics.

Hint:
Pbafvqre gur pnfr jurer gur cubar vgfrys vf ebgngrq
sbegl-svir qrterrf, unysjnl orgjrra cbegenvg naq 
ynaqfpncr bevragngvba.

Meta-hint:
http://www.rot13.com/