RE: [APML] More SS4000 scanning questions

["Lane, Jason R" <jason.lane@navy.mil>]

Hi guys,

Sorry I'm probably jumping into this thread a bit late, but thought I might add these comments.  I've never thought about breaking the numbers down before, but seems like it might be helpful.  Nyquist sampling theory states that your sample frequency should be at least twice that of your highest input frequency...for imaging this means your pixels (or grain, or whatever) should be at least twice as small as your smallest feature...the smallest feature being a film grain in our case.  By imaging the film grain correctly, then we know for sure that we've gathered every useful bit of information off the film that we can.  Imaging at a lower resolution will cause aliasing.  So the question is..how does the size of our scanner pixel compare to the size of our film grain?

2700 dpi = 2700 dots per inch, or 106.3 dots (pixel) per mm.  This means that each pixel is 9.4 um dimension or looking another way, corresponds to about 53 cycles per mm (or lines per mm..they are approximately the same thing).  Without looking at the film specs, I think this is pretty comparable to the size of your typical film grain or grain clumping (there are exceptions...Provias, Tech Pan, etc.).  Therefore, you are going to have aliasing problems scanning at 2700 dpi.  I've seen this myself in my scans on HP s20xi, where the scanned images look way more grainy than they should.  Some were worse than others, Konica Centuria Pro seemed especially bad but Supra 400 seemed extra grainy from the scan as well.  This is why I intend to upgrade my scanner.  I've looked at all the options and I've got my sights set on an Epson 4870.  

Hope this helps,
Jason

P.S.  I just did a search on film grain size, and found this page on aliasing and sampling with respect to scanning of films:  http://www.photoscientia.co.uk/Grain.htm   Looks like a pretty good write-up on the subject.

Also, as everyone else is saying, the optical resolution is what's important.  That site that I link to suggests defocusing slightly to reduce aliasing effects if your scanner only scans up to 2700 dpi and the software supports changing the focus.  

-----Original Message-----
From: astro-photo-bounces@seds.org
[mailto:astro-photo-bounces@seds.org]On Behalf Of Chris Cook
Sent: Thursday, November 11, 2004 10:02 AM
To: Discussion of Film Astrophotography
Subject: Re: [APML] More SS4000 scanning questions


> 4. I've asked this before and the answer made sense. For 35mm color
> negative film, is a scan setting higher than 2700 dpi useful? I had
> someone tell me in an off-topic discussion on the Losmandy group that I
> should be scanning at the highest resolution, 4000dpi in this case.

Depends.  My feeling is that the only 35mm films the would benfit for a
greater than 2700dpi scan would be Tech Pan, and extremely fine grained ones
like Provia 100 or Velvia.  For any ISO400 speed neg(Supra etc...), you're
not going to gain anything since the grain/film's resolution will limit you
first.  If you scan a ISO400 neg at 4000dpi, you're just ending up with a
larger file that occupies more of your HD... ;-)

Chris

-----------------------------
Chris Cook Photography
www.cookphoto.com





_______________________________________________
Astro-Photo mailing list
Astro-Photo@seds.org
http://seds.org/mailman/listinfo/astro-photo
_______________________________________________
Astro-Photo mailing list
Astro-Photo@seds.org
http://seds.org/mailman/listinfo/astro-photo