Practical DPI versus grain for consumer 35mm negative film (eg Kodak Gold 200)

R

Rowan Crowe

I have a Canoscan 5000F which is a 2400 DPI flatbed scanner with a
transparency adaptor. It can't hope to compete with Nikon 4000DPI or
Flextight 5760DPI scans of a Velvia positive trans (which I
outsource), but I've noticed that grain is visible on a Kodak Gold 200
neg even with the much "softer scanning" of the 5000F.

I want to do prints around the 8x10" size which is pushing the
resolution limits of this scanner, but I'm not sure whether I'd see
much benefit by sending off the negs to be scanned at a higher res.

In general, what would be considered some practical DPI limits for
various types of popular consumer neg film?
 
K

Kennedy McEwen

Rowan said:
I have a Canoscan 5000F which is a 2400 DPI flatbed scanner with a
transparency adaptor. It can't hope to compete with Nikon 4000DPI or
Flextight 5760DPI scans of a Velvia positive trans (which I
outsource), but I've noticed that grain is visible on a Kodak Gold 200
neg even with the much "softer scanning" of the 5000F.

I want to do prints around the 8x10" size which is pushing the
resolution limits of this scanner, but I'm not sure whether I'd see
much benefit by sending off the negs to be scanned at a higher res.

In general, what would be considered some practical DPI limits for
various types of popular consumer neg film?

Don't confuse the visibility of grain and film resolution, they are
related but one does not necessarily drive the other.

If you look at the Kodak and Fuji data sheets of the equivalent
professional film, they will give you resolution and MTF figures to use.
There is little difference between professional and consumer films other
than the storage conditions that they are kept under before they are
sold to you. Most modern 200 ASA negative films will resolve around
100cy/mm, which requires a scanner of at least 5000ppi to pull
everything off the film that can possible be recorded there. Generally
speaking though, the resolution of popular consumer films don't
determine the resolution of popular consumer images nearly as much as
how steadily those popular consumer cameras are held when taking the
popular consumer photographs. ;-)
 
B

Bart van der Wolf

SNIP
Generally speaking though, the resolution of popular consumer films
don't determine the resolution of popular consumer images nearly as
much as how steadily those popular consumer cameras are held
when taking the popular consumer photographs. ;-)

How very true :-(

I've seen only too many messages from scanner users, claiming there is
no more resolution than their scanner can extract from the film.
However, a view with an optical microscope, or a good scan from a
"sharp" film image (low ISO, good lens, and steady tripod), usually
reveals that more detail is available (despite graininess).

The current sweetspot seems to be around 6000-8000 ppi, but that
requires a dedicated filmscanner that secures good film flatness and
has high contrast optics. However, color accuracy is also a factor in
scan quality, so the quality balance is tipped depending on the final
purpose.

Bart
 
F

false_dmitrii

Kennedy McEwen said:
Don't confuse the visibility of grain and film resolution, they are
related but one does not necessarily drive the other.

In my limited forays into Neat Image with my noisiest Epson 1200dpi
negative scans, I've had some trouble getting the software to detect
enough sky and shadow noise without turning distant branchy trees into
gray blobs. Seems as though the grain and the perceived detail are
intersecting.

Is it accurate to say that a high-res scan well over the photo's
detail level can still offer improved grain reduction? My
understanding is that at lower resolutions, some of the
sub-pixel-sized grain will end up dominating the output pixel,
essentially enlarging and "hardening" the grain just like sub-pixel
image detail. Thus, a higher-resolution scan will usually smooth out
the edges of the grain and even shrink some of it in the process. I
don't have the technical terms, but is this the right idea?

Generally
speaking though, the resolution of popular consumer films don't
determine the resolution of popular consumer images nearly as much as
how steadily those popular consumer cameras are held when taking the
popular consumer photographs. ;-)

As well as the autofocus, autoexposure, and built-in lens quality, no?
Having no manual photography experience, I can only guess.

Regards,
false_dmitrii
 
K

Kennedy McEwen

In my limited forays into Neat Image with my noisiest Epson 1200dpi
negative scans, I've had some trouble getting the software to detect
enough sky and shadow noise without turning distant branchy trees into
gray blobs. Seems as though the grain and the perceived detail are
intersecting.

Is it accurate to say that a high-res scan well over the photo's
detail level can still offer improved grain reduction? My
understanding is that at lower resolutions, some of the
sub-pixel-sized grain will end up dominating the output pixel,
essentially enlarging and "hardening" the grain just like sub-pixel
image detail. Thus, a higher-resolution scan will usually smooth out
the edges of the grain and even shrink some of it in the process. I
don't have the technical terms, but is this the right idea?
That is a reasonable description of the effect of aliasing. Small
details and grain that should be sub-pixel have sufficient contrast to
dominate the pixel output and are thus enlarged in the output. That is
why software grain reduction functions such as Neat Image and others
can't really remove grain without also having an effect on sharpness -
the grain has been enlarged so that it is indistinguishable from larger
real image details. They usually try to get around this by some
adaptability, reducing the grain reduction when edges in the image are
detected, but that just means more grain is left visible around edges
and the image looks a little unnatural, almost plastic. They can be
acceptable if applied with a very light touch though, but they always
sacrifice image sharpness to some degree.
As well as the autofocus, autoexposure, and built-in lens quality, no?
Having no manual photography experience, I can only guess.
Fair guesses, and these are contributing factors though usually less so
than just simple camera shake.
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top