|
Hello,
With the absolutely abysmal
weather we are having thanks in no small part to Mr. El Nino, I spent a
good six hours gleaning the internet reading articles with some very good
illustrations about the new CMOS 35mm sized sensors. What has this to do with
astrophotography? Not terribly much, but there is a crossover between film
imaging and CCD imaging of sorts...we are all photographers so maybe this would
be of interest... if not, press the delete key.
The conclusions that I
could make came at a price... I noticed that the reviewers were always leaning
in one direction or another...so one had to weed through the bias.
The CMOS chips are very sharp... but after some
tests, a reviewer finally begrudgingly admitted that a 645 100 ASA chrome film
had better resolution in the same cropping... ie if one took a shot with the
CMOS and framed the same shot with the 645, the 645 showed more real detail.
The CMOS was very good
with color renditions... better than film, and the camera could be adjusted
to portray colors the way you wanted them, either vivid or subdued.
BUT... there are some issues with the cameras that
one could call "too much color" in the form of color fringing in high
contrast boundary lines, and moire patterns when the CMOS chip was
presented linear patterns that caused interference patterns to appear.
The CMOS chip portrays
light in a linear fashion and at a bit depth film can only dream about... BUT...
only in light levels that are friendly to the chip. Go beyond this boundary and
even with a dark subtract there is a lot of noise that appears in the shadow
areas. Film has grain, and that's it... go long or short, shadow noise
stays basically the same. It can reappear in a bad film scan, but a scanner
like the SS120 does a nice job in the shadows... much better than the CMOS
samples that I saw taken under dim light.
The CMOS has less noise
than film grain... a LOT less, but then again, only when the light is strong
enough to make it work. For those of us who
like to experiment with the "fringe elements" of exposures, or shoot the Milky
Way, the CMOS is not going to work ... you still need film.
My choice right now would
be the Pentax 645 N II system... speed and ease of a 35mm camera, image size big
enough to do something with, and economical! Given that the Canon CMOS body
costs around $6000 (I am guessing... it was supposed to cost $9000 but the Kodak
CMOS will go for around $4500 so they can't charge that much)... you can buy an
entire Pentax system for the price of that body. And my guess is that some
day Pentax will bring out a CMOS insert for the 645... if you wanted to
convert.
For now, it seems to
me that there are too many limitations with the CMOS cameras even if they do
perform amazingly well within their boundaries... I guess if I buy a camera
system, I don't want any boundaries... especially if I have to pay twice as much
for it! I don't want color fringing in my high contrast boundaries, I don't want
to worry about moire problems, I don't want to worry about light levels, memory
issues, battery issues... film may be a bit more clunky and it has grain, but it
works... I have noticed within myself an urge to get "the latest" thinking that
it must be "the best"... but after reading these reviews carefully, my
conclusion is that film is still a very viable recording medium and until
digital recording has addressed successfully these mentioned problems, it is my
overall choice for making images the way I like to make them.
Finally, there are some stand
alone digital backs that might work better than the CMOS cameras... but they are
outside my feeble budget and I'll bet they still have the inability to deal with
low light levels... really an important point in the type of imaging that we
do.
Have a Great New Year (we
need one...),
Tony
|