[filmscanners] Re: film and scanning vs digital photography

["" <lotusm50@sprynet.com>]

R. Jackson wrote:
> It depends, really. Like, I was scanning some old Ektachrome 400
> today. The images were coming out at at 4374 x 6400 pixels. That's
> about 28 megapixels and the scanner still wasn't clearly capturing
> the grain structure. Looking at it closely you can see what looks
> like noise, but is actually imperfectly resolved grain.

The thing about properly resolving the grain, is that the grain creates
the image.  It is responsible for acuity, contrast, resolution and
detail.  Without properly capturing it, in a scan, you are not getting
all the sharpness and detail that is there.   And I think this goes to
my comment about "proper scanning".   Further, current, modern slide
films are much better than the old 400, even noticeably better than they
were 5 years ago. They are sharper, finer grained and higher resolving
than before.


> Now,
> Ektachrome wasn't the finest-grained kid on the block, but the grain
> is fine enough that 28 megapixels isn't getting there.

That is another downside for film.  Once properly scanned, it takes a
lot more storage space and computer prower to handle.  That said,
handling the 220 MB files from a 5400 dpi scan of a 35mm slide hasn't
presented much of a problem for me.

> Of course,
> some of it is undoubtedly me hitting the optical limits of my
> Microtek scanner. That said, I've taken 5 1/2 megapixel images with
> my old Olympus E-1 that give some of my Ektachrome slides a run for
> their money when it comes to resolving detail.

Maybe an old Ektachrome 400 slide,  but not a current slide film.  And I
think it also depends on the image subject to some extent (and perhaps
what you are looking for in an image).   Some things just don't have
alot of detail to show so you are not losing anything by not having it,
and in some cases you don't need or care about the detail.  And if you
are not printing large, say if 8x10's are all you do, you might not
notice a difference.  But take the film to it's limits and you will
definitely notice the additional detail present in the film image (and
how interpolation of the digital file to produce a large print, creates
fake detail to fill in the extra pixels).

> It's just that the
> actual image on a slide doesn't begin to cover the amount of
> information contained in the slide and if you want it all you have to
> scan huge to get it. I just shot a couple of rolls of Efke 25 in my
> Mamiya 7. Those 6x7 negatives contain WAY more than four times the
> amount of information on those 35mm slides. I wouldn't be surprised
> if it took 175 megapixels to properly resolve the grain structure.
> And that's the real problem with comparing film and digital. 10-12
> megapixels will certainly give you images every bit as detailed as
> you're used to getting from film. Yet to capture the beauty of film's
> grain you have to scan at a level of detail that's really kind of
> impractical.
>

Yes.  I have to say that full 4000 dpi scans of 6x7 slides is scanning
huge  It creates a 500+ MB file.  This is not the easiest thing to work
with even with a good computer, and the scanning takes time.  But I have
to say that the results can be stunning -- with Astia 100F as well as
Efke 25 -- and virtually grainless.  And the only thing that can match
it is a $30,000 39 mp digital back.  I've scanned 4x5 film a few times,
but if I am going to do that seriously as I am planning, I'm going to
need a bigger computer.

> I had a test arrangement with a camera manufacturer last year to do a
> telecine of some old 8mm film to HDV. They wanted to know how it
> performed. They may have been thinking of looking into an HDV
> telecine product, I don't know. Anyway, the results were mixed. The
> 720x1280 images from the camera captured all the detail that the lens
> on the 8mm camera original delivered to the film. I'm fairly
> confident of that. But the camera didn't even begin to resolve the
> grain structure. In fact, after talking to their engineers I found
> out that the mpeg encoder saw the grain as high frequency noise and
> tried to suppress it. So I was seeing a kind of cross-hatch pattern
> on individual frames that had replaced the grain structure. Now, when
> the image was in motion you couldn't tell you weren't just looking at
> grain, but pausing on a frame left an impression of some kind of jpeg
> compression gone wrong or something. Obviously this wouldn't be the
> case with uncompressed recording, but then the file sizes would be
> immense and I'm pretty sure 720p doesn't even approach the level of
> detail needed to resolve the 8mm grain structure.
>

Yes, in some ways these things aren't set up for film.  It can be hard
for something set up for digital to adequately deal with analog grain.
but there are product designed to deal with grain very effectively.  You
can indeed reduce grain like you would reduce noise.  I've applied Neat
Image to scanned film images to reduce grain (when I want to take
something big and I don't want grain to intrude) and the results can be
amazing.  It is very effective

> So you've got kind of a mixed bag, IMO. You can replace film with
> digital at a fairly low resolution, IMO. 6-10 megapixels will usually
> yield comparable imagery, IMO. And yet to fully resolve the grain
> structure of film takes WAY more resolution than you need to replace
> it as a capture medium.

That 6-10 range depends on the type of film used and is pretty
consistent with what I have suggested.  It takes 10mp to produce
comparable image quality current 100 ISO color slide films from Fuji and
Kodak, probably 8 mp to be comparable to 400 ISO film, and 6 mp (if you
have one of these early DSLR's that is low noise) to produce what
800-1600 ISO film can (although it just won't have the charm and feel of
some of these films).

----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 
body