?
-
Many people are wondering if the 4870 is worth the upgrade from a 3200, 2450
or 3170. One of the first questions people seem to ask is, "Has the scanner
really improved, especially when compared to the 3200?" My tests only
provide insight in regard to actual resolution. Remember, resolution is only
one factor in a purchase decision... but a big one
.
I completed a number of test scans but have only posted the ones that seemed
most relevant. You can find them available in the files section of the Epson
2450/3200 Users Group at
http://groups.yahoo.com/group/epson3200/files/Epson 3200 vs. Epson 4870/ .
If that link doesn't work, just go to the homepage at:
http://groups.yahoo.com/group/epson3200/ . (I chose to post in 3200 group
because it already has the most number of members.)
All files were saved as "maximum" quality jpegs (#11 of 12). The files were
all scanned on the same computer using the same version of VueScan. Exposure
was set to 16-bit black and white, "manual" color settings with the default
white and black points used.
No adjustments have been made to the files after scanning. The best way, in
my opinion, to evaluate these files is to open a few of them up in Photoshop
and then go View > Actual Pixels which should set the magnification at 100%.
This should eliminate any video interpretation issues. I think the best way
to compare the 3200 ppi files against a 4800 file is to view the 3200 ppi
file at 100%, open the 4800 ppi file so it slightly overlaps the 3200 file
and then reduce the 4800 ppi image until the top and bottom lines of the box
around the test chart match up as the same equal distance as the 3200 ppi
file. Maybe others will have more appropriate methods.
Remember, you can eliminate all of the palettes, etc. on the Photoshop
desktop by cycling through with the tab key. This will give you more room to
view multiple files.
Unsharp masking is just a fact of life with flatbed scanners. I would
encourage you to test unsharp masking techniques on both the 3200 and 4800 p
pi files to see how much each file type can take before getting unsightly.
Due to the well-known variability from scanner unit to scanner unit within
this line of Epson scanners, your scanner may or may not have produced the
exact same results as mine.
To give you an idea of how large each of these test images is, the black
border around the test image is 2.5 mm square. In other words, one would
need to place 22.8 of these side by side to cover the 57 mm side dimension
of most "6x6" images. Around 520 (!) of these images would fit in the
standard 57 mm x 57 mm area of a "6x6" image.
If a file name indicates that the scan was masked, that means I used matte
black finished poster board to cover unoccupied/unused openings in the film
holder as well as areas of the scanner's glass bed that were not covered by
the holder itself. Due to the design of the actual target itself, there
clear border gaps right next to the matrix of individual target images.
Things I noted when comparing scans (remember.just one person's
observations):
* As has been recently discussed, masking out extraneous light does seem to
have benefits.
* The 4870's scan at 3200 ppi appears to me to be very slightly sharper than
the 3200's scan at 3200 ppi. I honestly didn't think there would be any
noticeable difference.
* I scanned the images in the middle of the scanner bed as well as at the
top of the bed and did not find any noticeable difference change sharpness.
* I did notice that the scanner definitely scans sharper in one direction
than in the other. In short, it appears to me as though these scanners will
scan a line placed vertically on the scanner bed more sharply than the same
type of line that is placed horizontally on the scanner bed. If the test
target was placed on the scanner for scanning so that it was oriented just
as it appears in these scan files, lines that appear as vertical in the
resulting test scan are sharper. If scan were oriented on the scanner bed
for scanning in a lengthwise position and then scanned (thus needing a 90
degree counter clockwise rotation to appear correct), then the horizontal
lines in the scan appear sharper.
* Newton Rings are the less of a problem than the observed decrease in
sharpness when film was placed directly on the glass.
We all know that these scanners are less sharp than dedicated film scanners.
No doubt about that. These scanners do a decent job with medium format and
larger when you consider the price-to-performance factor though. Remember
that you are looking at a test measuring incredibly fine detail. Just make
sure you are up-to-speed with decent unsharp masking skills
Doug
---
Doug's "MF Film Holder" for batch scanning "strips" of 120/220 medium format
film:
http://home.earthlink.net/~dougfisher/holder/mfholderintro.html
x
or 3170. One of the first questions people seem to ask is, "Has the scanner
really improved, especially when compared to the 3200?" My tests only
provide insight in regard to actual resolution. Remember, resolution is only
one factor in a purchase decision... but a big one
.I completed a number of test scans but have only posted the ones that seemed
most relevant. You can find them available in the files section of the Epson
2450/3200 Users Group at
http://groups.yahoo.com/group/epson3200/files/Epson 3200 vs. Epson 4870/ .
If that link doesn't work, just go to the homepage at:
http://groups.yahoo.com/group/epson3200/ . (I chose to post in 3200 group
because it already has the most number of members.)
All files were saved as "maximum" quality jpegs (#11 of 12). The files were
all scanned on the same computer using the same version of VueScan. Exposure
was set to 16-bit black and white, "manual" color settings with the default
white and black points used.
No adjustments have been made to the files after scanning. The best way, in
my opinion, to evaluate these files is to open a few of them up in Photoshop
and then go View > Actual Pixels which should set the magnification at 100%.
This should eliminate any video interpretation issues. I think the best way
to compare the 3200 ppi files against a 4800 file is to view the 3200 ppi
file at 100%, open the 4800 ppi file so it slightly overlaps the 3200 file
and then reduce the 4800 ppi image until the top and bottom lines of the box
around the test chart match up as the same equal distance as the 3200 ppi
file. Maybe others will have more appropriate methods.
Remember, you can eliminate all of the palettes, etc. on the Photoshop
desktop by cycling through with the tab key. This will give you more room to
view multiple files.
Unsharp masking is just a fact of life with flatbed scanners. I would
encourage you to test unsharp masking techniques on both the 3200 and 4800 p
pi files to see how much each file type can take before getting unsightly.
Due to the well-known variability from scanner unit to scanner unit within
this line of Epson scanners, your scanner may or may not have produced the
exact same results as mine.
To give you an idea of how large each of these test images is, the black
border around the test image is 2.5 mm square. In other words, one would
need to place 22.8 of these side by side to cover the 57 mm side dimension
of most "6x6" images. Around 520 (!) of these images would fit in the
standard 57 mm x 57 mm area of a "6x6" image.
If a file name indicates that the scan was masked, that means I used matte
black finished poster board to cover unoccupied/unused openings in the film
holder as well as areas of the scanner's glass bed that were not covered by
the holder itself. Due to the design of the actual target itself, there
clear border gaps right next to the matrix of individual target images.
Things I noted when comparing scans (remember.just one person's
observations):
* As has been recently discussed, masking out extraneous light does seem to
have benefits.
* The 4870's scan at 3200 ppi appears to me to be very slightly sharper than
the 3200's scan at 3200 ppi. I honestly didn't think there would be any
noticeable difference.
* I scanned the images in the middle of the scanner bed as well as at the
top of the bed and did not find any noticeable difference change sharpness.
* I did notice that the scanner definitely scans sharper in one direction
than in the other. In short, it appears to me as though these scanners will
scan a line placed vertically on the scanner bed more sharply than the same
type of line that is placed horizontally on the scanner bed. If the test
target was placed on the scanner for scanning so that it was oriented just
as it appears in these scan files, lines that appear as vertical in the
resulting test scan are sharper. If scan were oriented on the scanner bed
for scanning in a lengthwise position and then scanned (thus needing a 90
degree counter clockwise rotation to appear correct), then the horizontal
lines in the scan appear sharper.
* Newton Rings are the less of a problem than the observed decrease in
sharpness when film was placed directly on the glass.
We all know that these scanners are less sharp than dedicated film scanners.
No doubt about that. These scanners do a decent job with medium format and
larger when you consider the price-to-performance factor though. Remember
that you are looking at a test measuring incredibly fine detail. Just make
sure you are up-to-speed with decent unsharp masking skills
Doug
---
Doug's "MF Film Holder" for batch scanning "strips" of 120/220 medium format
film:
http://home.earthlink.net/~dougfisher/holder/mfholderintro.html
x