Digital BW, The Print

But, but, but B  i  l  l  l  l.......<whine mode off>

> Hi Mike,
>
> Two points-
> First, on the issue of "microstepping" to achieve any desired resolution
> below 4000 DPI:  If that is true, then why bother "locking" into "optical
> resolution" at all (below 4000DPI)?

Don't know.

>Can you really get, 305 DPI, or 3765

I guess, hmm, let's see, my D4000 has the following optical resolutions;
32, 36, 40, 46, 52, 60, 69, 80, 94, 109, 125, 143, 160, 182, 211, 236, 267,
286, 308, 334, 364, 400, 445, 500, 572, 667, 800, 1000, 1334, 2,000,
4000.....Nope, no 3765dpi.

> DPI, or 3888.8 DPI?  What is really happening (as you alluded to, and as I
> suggested earlier) is that the scanner is scanning at the next higher
> resolution and then interpolating down- but interpolating down, while not
as
> "fuzzy" as up-sampling, still leads to softening of the image.

Not according to Howtek. That only happens if you select a non optical
resolution. But knowing how manufacturer claims are, it would not surprise
me.

> 2nd, and more importantly, the issue of the DPI of the image.
>
> "I set the Q-Factor to 1.5 with a line screen of 205...  I set the desired
> Output size to 20"x25"...Silverfast ... sets the dpi at 308."
>
> Yes, 205 x 1.5= ~308 DPI-  But Mike, that's the DPI of your print @ 20x25,
> not your scan!
>
> In PS, open a new file, 20x25, 308 DPI- at 8 bit it's 45.3 MB; at 16, as
you
> say, it's 90.3 MB.  But more importantly, it's 6160 x 7700 **pixels**.  To
> get a file of 6160 x 7700 pixels from a 4x5, you have to scan at 1540 DPI-
> and that in fact is what you are scanning at, only you are really scanning
> at 2000 DPI and Silverfast is interpolating down, either by averaging
pixels
> or throwing away pixels.  In either case, why not just scan at 2000 DPI,
no
> interpolation, and then print at the resultant PPI, in this case at 400
PPI?

Ok, then lets call the scanners sampling "SPI" or samples per inch.
Obviously the scanner does not scan dpi. It is sampling optically with an
aperture setting equivalent to 308spi. The x axis movements and speed of the
drum determine the spatial resolution. (Did I say that right?).


> Lastly (did I say there were 2 points :^)  on Silverfast, Q-factors, and
> "optimal Scan resolution".  As I understand it, all of this comes from
> pre-press offset scanning; that is the derivation of the "1.5 times
> linescreen" resolution.  I don't believe it has any relevance to inkjet
> printing with Epson's.  Some people maintain that divisibles (?) of
Epson's
> 1440 DPI give the best results- in which case you would try for 180, 240,
> 360, 540, or 720 DPI.

Has this ever really been proven?
I have tried printing at 360dpi file sizes and I have sent the printer
1500dpi files. I see no difference. Recently Vincent Versace taught a
workshop for us, he stated that a print resolution of 360dpi produces the
best print quality. I challenged him and when I showed him a 20"x24" copy
"Magical Light" his jaw dropped.

>Others point out that if you achieve that by
> interpolating up or down, you are adding softness.  Paul has shown that
> above (I think it was) 540 DPI, there is no increase in measurable
> sharpness, and that 360 DPI is for almost all purposes equally as good.
In
> my scanning for print, I try to achieve something close to 360 DPI at my
> largest anticipated print size; beyond that, I let the Epson driver do the
> math.

That is a good posture.
>
> Your prints are still better than mine, dammit, but that's another issue
> altogether!

No they are not. I like your prints. You have been shooting for a long time
as many of us "quasi" old timers.
Bill, the bottom line is that I have no way of testing nor confirming either
what Howtek says or what we have discussed. The key is if it works for you,
do it. There are many ways to skin a cat, and all may be equally viable,
except maybe the religion of Pyro. ;-)

Mike
>
> Bill

Mike- 
Sorry, I wasn't trying to be a pain, just thought I'd give it one more go.
I think, unfortunately, we're talking across each other, here.

Keep up the good work, Mike; this is a small point.

Bill

on 4/6/02 9:36 PM, Michael Kravit wrote:

But, but, but B  i  l  l  l  l.......<whine mode off>




[Non-text portions of this message have been removed]