Don wrote
(in article said:
That's a good plan, too. Once you've narrowed it down, you can have
the winners scanner professionally, if it comes to that.
You have to understand, I was handed boxes and boxes of slides,
that had no sorting or filtering applied to them since they came
back from Kodak 40 years ago. Just stacks of slides, and most
of them shot hand-held, on cameras nowhere near the quality of
modern film cameras. Notable lens flare in some of the shots,
in lighting conditions that wouldn't be considered particularly
'difficult' today. Also, some of the earliest shots in here
(1950s) seem to have been shot with a camera that suffered from
light leaks along the bottom of the film door. It might just be
some common long-term storage failure, but a lot of the slides
have color spikes of red/orange along the bottom edge only, as
if light was let in by the camera. Also, they usually only show
up for frames during or adjacent to brightly lit outdoor shots.
IOW, this isn't exactly a photographer's gold mine I'm working
with.
I just decided to do them all figuring that what's important to me is
not necessarily what's going to be important to people looking at it
in the future. So I just do everything and be done with it. Saves me
agonizing over what to keep and what to toss.
I'm taking them all, but this scanner is literally too slow to
do that for every frame. What is the actual time per slide for
one of the dedicated Nikon slide scanners with the auto-feed
trays, when done at max resolution and color depth? I'm rapidly
coming to the conclusion it may be worth buying yet another
scanner.
I also figure, storage is cheap and these 1000s of today's "huge"
files will probably all fit on a single "future-DVD" with room to
spare.
Storage isn't the issue. I have plenty of that, and it's cheap
to get more. The problem is time. It would take me months,
working 12 hours a day to make it through all of these on this
scanner at max quality.
The thing is manufacturer ratings are notorious for fibbing. There are
ways (the razor edge test) to find out the real resolution which is
often times half that for flatbeds.
Yes, I need to do that test myself today. Good idea.
And believe it or not but 3000x2000 is not that far away from where we
are now. I mean, just compare today's resolutions to VGA of a few
years back. But all that's academic. The bottom line is your choice.
If I could batch scan properly, but I'd be on board with your
idea completely. The problem is, this software is crap, and I
can't do more than three frames at a time in a sequence without
it crashing or giving me runtime errors about running out of
memory. Nevermind that I've watched it carefully during
execution, and it never goes above 1.5GB of RAM in process size,
and I have 4GB installed. Also, I have to manually name each
slide, and fight it over the 'fallback to jpg' problem I
described earlier. Also, about 10% of the time it will write a
zero byte file by mistake and I have to constantly check the
files after saving them to make sure they actually showed up,
and in the proper tiff format. I could go with jpg at 100% to
save some time, but I suspect something is badly wrong there as
well, as it takes a file that is 20MB in tiff and makes a 24kb
jpg out of it. Puhleeze.
My other project for today is to try using VueScan instead and
compare the results.
Exactly! It's the same principle although it's even more important for
digicams because of the so-called "Bayer pattern". The raw digicam
images have 2 green pixels and 1 of each red and blue. The kicker is
they don't overlap so some interpolation is required and there are
many ways to do that. The method used in cameras is usually more
concerned with speed, rather than quality.
I have hope that down the road RAW processing will continue to
improve, and we'll be able to go back to old RAW files and
extract much better data from the same original.
There are two things there. One is selecting which ones are worth
keeping and that's your call. The other is scanning and then it's
definitely worth using maximum resolution. Especially if the originals
are bad! That's because you'll be able to pull out and "rescue" much
more if you use maximum resolution and bit depth. This gives you more
elbow room to do some drastic editing and pull every bit out of what's
left.
Of course, there's no point going beyond what's on the film but to
reach that stage you have to go to about 10,000 dpi (if we go by grain
size alone). However, that's over the top and for practical purposes
the meaningful data levels off at around 5000 dpi for film - assuming
top notch film and steady (tripod) shots, etc.
Another sign of my lack of scanner knowledge, but the marketing
claims on this scanner are "4800x9600dpi". ??? The highest
setting available I can find is 4800. And it crashes trying to
do more than 3 of those at a time. Still, I have no idea where
the 9600 is supposed to come in, or is that some marketing BS
for interpolation of the data?
Another thing is those really mundane, everyday things we forget. You
know, like... I don't know... some food packaging you see in a corner
of a picture or some other such accidental item which makes you go:
"Oh yeah! I used to love that! Whatever happened to it?"
Oh yeah, that keeps happening. The really interesting bit is
that I keep remembering things I didn't realize I knew about
long-dead relatives from these images. Almost creepy how your
mind dredges up things and events long vanished from your
'normal' memory.
And as you say, the "progress" as the skyline keeps getting more
crowded.
True, but I was mostly referring to how blue the sky was back
then compared to today. Even Los Angeles, back in 1962, had a
blue, not gray smog sky. More generally, every outdoor location
seemed to have a MUCH better looking atmosphere than what we
'enjoy' today.
