Re: [APML] Film vs Digital - is film really dead?

["Paul Howell" <paul@howell-ltd.com>]

Hi Andrew -


> Professional astronomers embraced the CCD so quickly I think in part
because
> of the quick turnaround.

Yes, but that is the smaller part of it. First and foremost, it's the
big gain in quantum efficiency. Most professional imaging is either
spectroscopic, or is photometric. In either case, it is likely that
the object is rather small. So catching more photons per unit time is
like shooting with a bigger scope for the cost of a CCD detector. The
other prime reason is analytical. Film is horribly non-linear. This is
actually advantageous for aesthetic purposes, but nightmarish for data
reduction. CCD's are linear, and of course the data are read directly
into the computer.

This is what makes the choice of CCD so obvious for professional use,
but paradoxically makes it less clear for amateur use. As you point
out, CCD cannot be justified from a purely economic standpoint
presently for amateur astronomers. Also the affordable CCD size
formats currently available don't come close to competing with medium
format film for coverage (unless you mosaic). This is a hobby, so the
justification for CCD performance is the same as it is for all other
hobbies: "I can afford it so I want it".

> Maybe in 2010 we'll see a flood of products based on the new digital
imaging
> technology that is to replace the CCD, CMOS-based imaging. CMOS is
> ultra-cheap stuff as your computer chips are all CMOS based. Maybe
then the
> economics will change, but no matter how, there will probably always
be use

Maybe, but I am a doubter in this regard. As you point out, the great
majority of all silicon wafer fabrication lines in the world are CMOS.
Hence, the per unit area cost (this is one of the things that drives
chip cost) is lowest for CMOS due to the high absorption of fixed
costs. Specialty processes like CCD have inherently higher cost. Thus
CMOS sensors are indeed cheaper than like CCD sensors.

The problem with CMOS is that it is significantly less sensitive and
noisier than a CCD device. For 8-bit consumer cameras, that is
probably ok, but not for high-end professional photography cameras
that demand 12 or 14 bit data, and certainly not astronomical cameras
that need to digitize to 16 bits. CMOS technology will progress of
course, but so will CCD technology. Whether the quality gap will ever
close enough to make CMOS attractive to astronomical applications,
well we will see...

-P



--  APML Archives at <http://astro.umsystem.edu/apml/>  ---
             Unsubscribe at <majordomo@seds.org>