I remember the thread, and I went and looked at the software mentioned...
IIRC, the software required multiple images, but could additively combine
the 'good' parts from several into one image that had more detail. It was
not making 'something from nothing' but rather taking the best parts of
several 'somethings' and making one 'super something' out of it.
This is similar to what you can do with Vuescan and multiple passes, to get
more dynamic range out of an image using multiple scans. In the case of the
star trails, I think there were multiple 30-second exposures combined to
produce the equivalent of one multiple-minute exposure without blowing out
the highlights. I would be interested in this software, but I think it only
works with Windoze, and I have a Mac.
--
Jim Brokaw
OM-'s of all sorts, and no OM-oney...
on 6/25/03 6:43 AM, Thomas Bryhn at thomas.bryhn@xxxxxxxxxx wrote:
> That would be interesting. Not only for photography, it could also change
> all sciences dealing with noise and statistics. Imagine to be able to take
> one measurement, run it through software and have an output with more
> significant digits than the input. You could for instance make an election
> poll by asking one person, feed his answer to a computer, and then get a
> good estimate of the election result!
> Sorry to make fun, but one measurement is still only one measurement, you
> can't magically subtract noise from it and be left with the signal. That
> kind of image enhancement is only possible in movies, the same ones where
> they "zoom in" on a digital picture and bring out more and more details.
> BTW, star trails have to my knowledge never suffered from lack of light,
> one usually stops down to about f/8 and then exposes for several hours. The
> exposure is only made this long to make the stars move across the
> film/sensor. Digital cameras will suffer badly from noise with exposure
> times longer than a few seconds, so someone has probably made a program to
> stack several shorter exposures and subtract a dark frame. I would like to
> have a look at that program, though.
>
> Thomnas Bryhn
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|