Re: qcam-0.91

> I've been playing with the -E flag on the qcam, but I'm not sure what
> format and what values it will take. can anyone out there give me/us a
> brief desc of the -E flag and acceptable values?

From the man page:

          Automatically adjust exposure.  -E takes  from  one  to
          five parameters.  They are, in order:

          mode  target-lum  tolerance-lum   std-target-lum   std-
          mode is an integer.  Modes 0 and 1 examine the  average
          brightness  of  the  picture, and adjust the brightness
          until the average pixel luminance is in range.  Mode  0
          looks  at  the  whole picture; mode 1 looks at only the
          center ninth of the picture.

          Mode 2 adjusts contrast to get the  standard  deviation
          of the pixel luminance into range.

          Ranges are specified in the next  four  parameters;  if
          they are not supplied, reasonable defaults are used.

          From getopt's point of view, -E  takes  one  parameter.
          So  if  you're  specifying  more  than  one  parameter,
          enclose the parameters in quotes.

A similar description is in the CHANGES file, under the 0.8 delta
comments.  I would love to make the docs clearer, if you can send
me a suggestion on what to put in the man page instead, I'd be

You're right that this is not very clear on valid ranges.  Target-lum
has a useful range of 0-15 or 0-63, if you're using 4bpp or 6bpp
respectively.  Tolerance should be a non-negative integer; small
numbers give more consistent pictures, but slow things down.  The
Cam keeps exposing until the average brightness is
(target-lum)+/-(tolerance).  Standard deviation should be a small
non-negative integer; the defaults are 10 (for 6bpp) or 2 (for
4bpp).  Low standard deviation == low contrast.  Tolerance for
standard deviation works the same as exposure tolerance.

Default values for 6bpp are 32, 0, 10, 0.  Default values for 4bpp
are 8, 0, 2, 0.