pixels per inch for film?

What is the maximum achievable result for scanning in different film
types in dots per inch? I know scanners can scan to very high
resolution but there must come a time when it is not doing anything
productive. I guess that for the amateur medium format flatbed
scanners with a transparency hood it will be way lower. But by how
much?

On 21 Nov 2004 23:47:15 -0800, (RolandRB)
wrote:

What is the maximum achievable result for scanning in different film
types in dots per inch? I know scanners can scan to very high
resolution but there must come a time when it is not doing anything
productive. I guess that for the amateur medium format flatbed
scanners with a transparency hood it will be way lower. But by how
much?


Don't confuse the scanner's resolution with sharpness.

A 2500 dpi film scanner with good optics is worth
more than a 4800 dpi film scanner with poor or
marginal optics.

There are a lot of cheap flatbed/film scanners now
in that second category.

That said -- with really good optics, the practical
limit seems to be around 4000 to 5000 dpi, though
it really depends on the image (film type, camera,
lens, technique, lighting, etc.)


rafe b.
http://www.terrapinphoto.com

On 21 Nov 2004 23:47:15 -0800, (RolandRB)
wrote:

What is the maximum achievable result for scanning in different film
types in dots per inch? I know scanners can scan to very high
resolution but there must come a time when it is not doing anything
productive. I guess that for the amateur medium format flatbed
scanners with a transparency hood it will be way lower. But by how
much?


Don't confuse the scanner's resolution with sharpness.

A 2500 dpi film scanner with good optics is worth
more than a 4800 dpi film scanner with poor or
marginal optics.

There are a lot of cheap flatbed/film scanners now
in that second category.

That said -- with really good optics, the practical
limit seems to be around 4000 to 5000 dpi, though
it really depends on the image (film type, camera,
lens, technique, lighting, etc.)


rafe b.
http://www.terrapinphoto.com

In article ,
rafe bustin wrote:

Don't confuse the scanner's resolution with sharpness.

A 2500 dpi film scanner with good optics is worth
more than a 4800 dpi film scanner with poor or
marginal optics.

There are a lot of cheap flatbed/film scanners now
in that second category.

That said -- with really good optics, the practical
limit seems to be around 4000 to 5000 dpi, though
it really depends on the image (film type, camera,
lens, technique, lighting, etc.)

It also depends on your basis for comparison. Is the OP looking for distinct
pixels on his computer screen, or for extinction resolution?

If we turn a traditional debate topic on its head and ask what area of film
is equivalent to digital capture, I'm finding that with modest consumer
equipment (an Epson 4870) and old TLRs, albeit ones with good optics, I'm
getting results which seem to compare, pixel for pixel, with decent digital
capture at about 2500-3000 pixels per inch, with slide film. My results also
suggest that the limiting factor here is the scanner, so your figures of
4000-5000 for better scanners seem reasonable.

"RolandRB" wrote in message
om...
What is the maximum achievable result for scanning in different film
types in dots per inch? I know scanners can scan to very high
resolution but there must come a time when it is not doing anything
productive. I guess that for the amateur medium format flatbed
scanners with a transparency hood it will be way lower. But by how
much?

For high-quality scanners, the time when it is not doing anything productive
has yet to be reached. However, if you're making an 8 x 10 enlargement, you
surely don't need 4000 dpi. There are those who claim that you should still
scan at max res and reduce image size in Photoshop to maintain the highest
degree of fidelity to the original capture.

--
Regards,
Matt Clara
www.mattclara.com

"RolandRB" wrote in message
om...
What is the maximum achievable result for scanning in different film
types in dots per inch? I know scanners can scan to very high
resolution but there must come a time when it is not doing anything
productive. I guess that for the amateur medium format flatbed
scanners with a transparency hood it will be way lower. But by how
much?

For high-quality scanners, the time when it is not doing anything productive
has yet to be reached. However, if you're making an 8 x 10 enlargement, you
surely don't need 4000 dpi. There are those who claim that you should still
scan at max res and reduce image size in Photoshop to maintain the highest
degree of fidelity to the original capture.

--
Regards,
Matt Clara
www.mattclara.com

On Mon, 22 Nov 2004 14:30:47 GMT, Chris Brown
wrote:

In article ,
rafe bustin wrote:

Don't confuse the scanner's resolution with sharpness.

A 2500 dpi film scanner with good optics is worth
more than a 4800 dpi film scanner with poor or
marginal optics.

There are a lot of cheap flatbed/film scanners now
in that second category.

That said -- with really good optics, the practical
limit seems to be around 4000 to 5000 dpi, though
it really depends on the image (film type, camera,
lens, technique, lighting, etc.)

It also depends on your basis for comparison. Is the OP looking for distinct
pixels on his computer screen, or for extinction resolution?

If we turn a traditional debate topic on its head and ask what area of film
is equivalent to digital capture, I'm finding that with modest consumer
equipment (an Epson 4870) and old TLRs, albeit ones with good optics, I'm
getting results which seem to compare, pixel for pixel, with decent digital
capture at about 2500-3000 pixels per inch, with slide film. My results also
suggest that the limiting factor here is the scanner, so your figures of
4000-5000 for better scanners seem reasonable.


"What area of film is equivalent to digital capture"
Good question, sort of.

But then the rest of the paragraph leaves me very
confused. What direct digital capture are you
comparing to your 4870-scanned TLR images? And
how do you make such a comparison?

I see two other ways to compare direct-digital-
capture to scanned film:

1. useable image detail per unit sensor area
2. quality of image data per megapixel

In both of these measures, direct-digital-capture
holds up well against film, but film still wins,
because a) there's more of it, and b) you can
scan it at very high resolution.

In #1, a sensor 15mm x 23mm holds up reasonably
well against 24mm x 36mm of film. The ratio of
sensor areas is 2.5 to 1.

In #2, compare 20 Mpixels from a 35 mm film scan
(at 4000 dpi) to 6 Mpixels from a 10D -- again,
the 10D image holds up surprisingly well. The
ratio of pixels is 3.3 to 1.


rafe b
http://www.terrapinphoto.com

On Mon, 22 Nov 2004 14:30:47 GMT, Chris Brown
wrote:

In article ,
rafe bustin wrote:

Don't confuse the scanner's resolution with sharpness.

A 2500 dpi film scanner with good optics is worth
more than a 4800 dpi film scanner with poor or
marginal optics.

There are a lot of cheap flatbed/film scanners now
in that second category.

That said -- with really good optics, the practical
limit seems to be around 4000 to 5000 dpi, though
it really depends on the image (film type, camera,
lens, technique, lighting, etc.)

It also depends on your basis for comparison. Is the OP looking for distinct
pixels on his computer screen, or for extinction resolution?

If we turn a traditional debate topic on its head and ask what area of film
is equivalent to digital capture, I'm finding that with modest consumer
equipment (an Epson 4870) and old TLRs, albeit ones with good optics, I'm
getting results which seem to compare, pixel for pixel, with decent digital
capture at about 2500-3000 pixels per inch, with slide film. My results also
suggest that the limiting factor here is the scanner, so your figures of
4000-5000 for better scanners seem reasonable.


"What area of film is equivalent to digital capture"
Good question, sort of.

But then the rest of the paragraph leaves me very
confused. What direct digital capture are you
comparing to your 4870-scanned TLR images? And
how do you make such a comparison?

I see two other ways to compare direct-digital-
capture to scanned film:

1. useable image detail per unit sensor area
2. quality of image data per megapixel

In both of these measures, direct-digital-capture
holds up well against film, but film still wins,
because a) there's more of it, and b) you can
scan it at very high resolution.

In #1, a sensor 15mm x 23mm holds up reasonably
well against 24mm x 36mm of film. The ratio of
sensor areas is 2.5 to 1.

In #2, compare 20 Mpixels from a 35 mm film scan
(at 4000 dpi) to 6 Mpixels from a 10D -- again,
the 10D image holds up surprisingly well. The
ratio of pixels is 3.3 to 1.


rafe b
http://www.terrapinphoto.com

In article ,
rafe bustin wrote:

But then the rest of the paragraph leaves me very
confused. What direct digital capture are you
comparing to your 4870-scanned TLR images? And
how do you make such a comparison?

I'm comparing with the results I get from my EOS 10D, and evaulate the
images subjectively on the basis of getting similar image quality from
equivalent pixel-count crops of both images. I appreciate that this isn't
especially scientific, but my end point is to have something which
subjectively looks nice as a print, not a hard and fast numerical
comparison.

I see two other ways to compare direct-digital-
capture to scanned film:

1. useable image detail per unit sensor area
2. quality of image data per megapixel

Number 2 - assuming I can get the field of view I want, the sensor area is
of only academic interest.

In both of these measures, direct-digital-capture
holds up well against film, but film still wins,
because a) there's more of it, and b) you can
scan it at very high resolution.

I agree. With the relatively inexpensive setup I have, I'm finding that I
can get 10D-style results at around 36 megapixels (6000*6000) with 6*6 and
moderately decent optics. The largest print I can make myself from these
images is 13*13 inches (Epson 1290), and they do look very good indeed -
much better than the DSLR stuff when put side by side.

In article ,
rafe bustin wrote:

But then the rest of the paragraph leaves me very
confused. What direct digital capture are you
comparing to your 4870-scanned TLR images? And
how do you make such a comparison?

I'm comparing with the results I get from my EOS 10D, and evaulate the
images subjectively on the basis of getting similar image quality from
equivalent pixel-count crops of both images. I appreciate that this isn't
especially scientific, but my end point is to have something which
subjectively looks nice as a print, not a hard and fast numerical
comparison.

I see two other ways to compare direct-digital-
capture to scanned film:

1. useable image detail per unit sensor area
2. quality of image data per megapixel

Number 2 - assuming I can get the field of view I want, the sensor area is
of only academic interest.

In both of these measures, direct-digital-capture
holds up well against film, but film still wins,
because a) there's more of it, and b) you can
scan it at very high resolution.

I agree. With the relatively inexpensive setup I have, I'm finding that I
can get 10D-style results at around 36 megapixels (6000*6000) with 6*6 and
moderately decent optics. The largest print I can make myself from these
images is 13*13 inches (Epson 1290), and they do look very good indeed -
much better than the DSLR stuff when put side by side.