[argyllcms] Re: Producing target to compare quality of cinematic film scanners
- From: Ben Goren <ben@xxxxxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Fri, 17 Aug 2018 09:11:06 -0700
On Aug 17, 2018, at 2:14 AM, Richard Kirk (Redacted sender "richard" for DMARC)
<dmarc-noreply@xxxxxxxxxxxxx> wrote:
You could use a synchrotron.
Really good discussion there from Richard, especially the overall arc of the
approach.
I just wanted to chime in that graphic arts measurements are the least
demanding form of spectroscopy there is. It’s very, very, very easy to measure
imperceptible differences — or to measure perceptible differences that are
beyond the practical limitations of the output devices (printer, display, etc.).
So...I’d start with something trivially easy. If you’re in the developed world,
mains voltage is more than stable enough for this. If the illuminant looks
evenly diffused, it’s good enough even if your light meter can measure an hot
spot. Your household lightbulb might make a lot of short wavelengths, but it
probably makes more than enough for graphic arts use...and, if it’s too dim and
the results are noisy, just take half a dozen measurements and calculate the
median value — that’s guaranteed to be overkill.
Many of these projects are single-time-use applications; you’ll do it once and
never again. And if you ever do do it again, it’ll be because of some
limitation that you didn’t anticipate with your initial design, so you’ll re-do
it to focus on that particular shortcoming.
And, above all...your first goal should be to get from beginning to end with
“close enough,” to just see if there’re any showstoppers later in your workflow
that you hadn’t even thought of. Then, when you’ve made it to the end, figure
out where your biggest _actual_ problem is and focus on that — and, often
enough, you don’t actually have the problems you thought you would and
otherwise would have wasted all your time fixing.
Cheers,
b&
Other related posts:
