[AG-TECH] lighting

[Markus Buchhorn]

At 09:08 11/04/01 -0600, Gary Orser wrote:
>On another topic.
>
>It has been suggested that high data rates and low fps from
>a camera are affected by florescent vs incandescent lighting.
>
>How large is the effect, and how much should one invest in
>switching the lighting in agnode space to incandescent?

Hi Gary

When I have measured the effect, it has been as large as a net factor of
approx. four, through a combination of reduced bandwidth (factor of
~1.5-2.5) and increased frame rate (factor of ~1.5-2). It depends a *lot*
on the fluorescents in use - some appear to run at higher frequencies, or
have better coatings that cause less beating with the video capture. It
also appeared to depend somewhat on the capture system, some boxes were
more sensitive than others. In a few cases I saw very little improvement
(but still some).

I've also seen this with ISDN-H.320 based VC systems.

Note that daylight coming from an external window can also cause similar
grief (specular reflections, varying intensity, etc.). The effect isn't
usually as large though.

Rule of thumb when I set up a video capture - take a desktop incandescent
with you and try it out (turn off the fluoros). Almost all rooms I have
tried this in I have seen a healthy improvement. However, most of the rooms
have been on one campus, where lighting fixtures are centrally managed - so
we tend to get the same lights installed.

The only hassle then is that desktop incandescents usually have a horrible
colour distribution/temperature :-) The eye copes, but video cameras
capture it in raw glory.

Cheers,
	Markus

Markus Buchhorn, Faculty of Engineering and IT,          | Ph: +61 2 61258810
email: markus.buchhorn at anu.edu.au, mail: CSIT Bldg #108  |Fax: +61 2 61250010
Australian National University, Canberra 0200, Australia |Mobile: 0417 281429