Digital BW, The Print

I have some 6cm x 6cm b&w negatives exposed back in the early '60s on 
Tri X Pan Professional. They blew up to 20" x 16" quite successfully 
using a DeVere cold-cathode enlarger. 

Recently I had some of these negatives scanned by a bureau through an 
Imacon Flextight machine, but grain has become so pronounced that the 
scans are virtually unusable.

I understand the problem. Light in a scanner is highly collimated - 
the light beams are nearly perfectly parallel. So the grains don't 
just block the light - they scatter it creating greater apparent 
density. The so-called "Callier Effect." And it will have been made 
worse because I neglected to tell the bureau *not* to sharpen the 
image!

However I wondered what 'work-round' others in this group have tried 
to reduce grain on silver negatives. 

I have tried blurring  the image lightly in Photoshop, followed by 
unsharp masking but am not very happy with the results.

I have also read somewhere that some old-timers digitise their images 
via a conventional photographic print and a flat-bed scanner. By 
printing with a diffuse light source and a relatively soft grade of 
paper they suppress grain while capturing a tonal range that can be 
enhanced in Photoshop. Doubtless this works, but it does seem 'the 
long way round' and since it introduces an extra step in the process, 
image quality is bound to suffer.

Of course I'm now running trials with the newer chromogenic films but 
that doesn't solve my problems with the archival images I still have.

Any ideas on this theme would be most welcome!

There is a product I was just told about, but have had little chance to
test--called Neat Image (http://absoft.hotbox.ru): it is designed to reduce
noise in digital camera files, but I intend to try it for this very problem,
it looks promising and its free!

Wendel

> -----Original Message-----
> From: atmcintyre2001 [mailto:amcintyre@...]
> Sent: Saturday, January 26, 2002 7:53 AM
> To: DigitalBlackandWhiteThePrint@yahoogroups.com
> Subject: [Digital BW] Excessive grain in scanned images
>
>
> I have some 6cm x 6cm b&w negatives exposed back in the early '60s on
> Tri X Pan Professional. They blew up to 20" x 16" quite successfully
> using a DeVere cold-cathode enlarger.
>
> Recently I had some of these negatives scanned by a bureau through an
> Imacon Flextight machine, but grain has become so pronounced that the
> scans are virtually unusable.
>
> I understand the problem. Light in a scanner is highly collimated -
> the light beams are nearly perfectly parallel. So the grains don't
> just block the light - they scatter it creating greater apparent
> density. The so-called "Callier Effect." And it will have been made
> worse because I neglected to tell the bureau *not* to sharpen the
> image!
>
> However I wondered what 'work-round' others in this group have tried
> to reduce grain on silver negatives.
>
> I have tried blurring  the image lightly in Photoshop, followed by
> unsharp masking but am not very happy with the results.
>
> I have also read somewhere that some old-timers digitise their images
> via a conventional photographic print and a flat-bed scanner. By
> printing with a diffuse light source and a relatively soft grade of
> paper they suppress grain while capturing a tonal range that can be
> enhanced in Photoshop. Doubtless this works, but it does seem 'the
> long way round' and since it introduces an extra step in the process,
> image quality is bound to suffer.
>
> Of course I'm now running trials with the newer chromogenic films but
> that doesn't solve my problems with the archival images I still have.
>
> Any ideas on this theme would be most welcome!
>
>

> I had some of these negatives scanned by a bureau through an
> Imacon Flextight machine


> I understand the problem. Light in a scanner is highly collimated -
> the light beams are nearly perfectly parallel. So the grains don't
> just block the light - they scatter it creating greater apparent
> density. The so-called "Callier Effect."

Hum...  I believe the light source in the Imacon is a diffuse light source
(OSRAM L 8W/12-950 Lumilux De Luxe Daylight 5400K lamp), not a point light
source?  If so, that is to your advantage!  It wouldn't be "parallel" light
(as you call it), but diffuse...and I believe the effect would be the
opposite you are talking about.

Austin

You would be better off having the negs rescanned from scratch, 
rather than trying to save them at this point. In the FlexColor 
software, a USM setting of -60, that's negative 60, in effect, is 
zero unsharp masking. Why, I don't know, but it is. Imacon 
recommends scanning negs at -60, and transparenicies at 
about 250, due to difference in the grain. I'd ask your scanner 
person to rescan them at -60 and then look at a sample.

Personally, I'd scan negs at 100-150. If the grain is there, I'd 
rather it be sharp than murky. Nothing worse to me than an 
image with the grain not sharp. (Think of darkroom prints with 
the enlarger head out of alignment -- where the grain gets soft 
and mucky on the edges of the frame -- that's what I'm talking 
about).

But it's your call. That's what I can offer.

Mark Tucker

There has been extensive discussion of this problem and the 
phenomenon of "grain aliasing" in general on the Usenet 
newsgroup comp.periphs.scanners, so you might post a 
message there .... It seems that this effect is far more 
pronounced on some scanners than others. Many Nikon 
scanner users say it's not a problem for them. I use a Polaroid 
Sprintscan 4000 and haven't found it to be a problem, but I've 
used a Kodak Rapid Film Scanner in the past, and the grain was 
extremely pronounced.

Peter

--- In DigitalBlackandWhiteThePrint@y..., "atmcintyre2001" 
<amcintyre@b...> wrote:
> I have some 6cm x 6cm b&w negatives exposed back in the 
early '60s on 
> Tri X Pan Professional. They blew up to 20" x 16" quite 
successfully 
> using a DeVere cold-cathode enlarger. 
> 
> Recently I had some of these negatives scanned by a bureau 
through an 
> Imacon Flextight machine, but grain has become so 
pronounced that the 
> scans are virtually unusable.

Is there no Mac version?

thx,
t

"Wendel White"  wrote:
> There is a product I was just told about, but have had little 
chance to
> test--called Neat Image (http://absoft.hotbox.ru): it is designed 
to reduce
> noise in digital camera files.

I've been experimenting with beta versions Neat Image for about 3 
weeks.  They have just released a fully functional Demo version. It 
can be downloaded at:  http://absoft.nm.ru/download.html .

Take their advice and print out the User's Manual and **READ** it. It 
will save you time and frustration. 

It is very interesting software.  It analyzes and adjusts the 
grain/noise at three frequencies for each YCC channel (or RGB channel 
if you choose),and also allows differential sharpening at the three 
frequencies and channel.  I have found it to do a great job on my 
negatives (color and b/w).  [Note:  You currently have to convert 
your b/w to 24-bit RGB to use NI.] 

It is very slow (lots and lots of calculations).  It also tends to 
leave a few strange artifacts that are visable if you print at 1:1.

Jeff Randall

ternahan <ternahan@s...> wrote:
> Is there no Mac version?


No Mac.  Also no PhotoShop plugin.  The plugin is under concideration 
for a future release.  The Neat Image development team is apparently 
focusing on getting the standalone version up and running first.


The current v1.0 Demo only reads 24-bit files.  48, 8, 16 bit 
capability is planned.  It can read BMP, JPG, and TIF files but only 
outputs BMP.  The soon to be released v1.0 Home version (payware) 
will output to all three. 


Jeff Randall

thanks.....rats!

> No Mac.  Also no PhotoShop plugin.  The plugin is under concideration
> for a future release.  The Neat Image development team is apparently
> focusing on getting the standalone version up and running first.
> 
> 
> The current v1.0 Demo only reads 24-bit files.  48, 8, 16 bit
> capability is planned.  It can read BMP, JPG, and TIF files but only
> outputs BMP.  The soon to be released v1.0 Home version (payware)
> will output to all three.
> 
> 
> Jeff Randall