I took a NBS lens resolution test chart and scanned it in at 1000dpi
(native) on my flatbed. Then I shrunk it down by a factor of 6.25 so
the resolutions in lpmm on the chart would be reduced by a factor of 4,
giving a range of 3 lpmm to 20 lpmm (wishfull thinking).
There are six targets on the chart, I set one pair to 100%, the next to
67%, and the third to 33%, photoshop grayscale. Then I applied a USm to
the righthand set of three.
Printed it out on the Epson Backlight film I've been working on for
enlarged negatives.
Results:
100% 6 lpmm, 8 lpmm horiz only
67% 7 lpmm, 8 lpmm horiz only
33% 8 lpmm, 10 lpmm horiz only
This is better than I expected, and encouraging for 11 x 14 inch prints.
Saw very little difference for the USM side, if any.
The biggest caveat is that even at the lower lpmm bars, the dots along
the axis of the bars were visible, not smooth and continuous as in film
tests I have done.
So that answers one 'cold hard data point', still says very little about
how these negs would compare with an equivalent analog print made by
someone who knew how to get the most from either process (they surely
wouldn't match, but how close could the Epson digital neg get?).
I'd really be curious about how how close this is to David's method in
in this respect (cause I'm too cheap and lazy to deal with shops with
imagesetters). If anyone who is currently producing neg's that way would
like to tag this test chart on one of thier negs I'll carry the ball and
send a zip, floppy, file via the net, one of my print samples, or
whatever. Please let me know. The file right now is a ridiculous 6.25K
dpi at 1.05 x 1.52 inches = ~60 Mbytes, so you'd need to specify how I'd
downsample it to 450 or 600 dpi to get it to you, whatever your
preference.
Also in case anyone is interested, I'll post this on the web for
downloading but it'll be a day or two.
FYI, I have it from an informed source that an 8x10 print will be
percieved as sharp at about 5 lpmm, when held at a normal viewing
distance, an 11x14 would be viewed from a greater distance and so be
percieved as sharper. So this is over the hump. My impression (meaning I
cant and dont care to back it up) has been that large format work on can
reach as high as 20 lpmm on an 8x10 print and viewers looking at this
print side by side with perhaps a 10 lpmm print can percieve a
difference.
All this says nothing about tonality however.
Larry