Comments please - off the list!
I am posting this protocol to the list in general so that people more
knowledgeable than I can comment. Please direct comments to
me *directly*. Please do NOT make this into a list thread. I would
much rather iron out the bugs off the list than to take up bandwidth
of people who are not interested. This is also in view of current
bandwidth issues of the list.
If you would like to take part, and have a Nikon scanner, please
email me for the target files. Please *don't* take any photos until
people have had the chance for feedback in case the protocol
needs modification. I don't think any changes to the actual photo-
graphic process will be needed, possibly only the scanning.
Thanks in advance,
Rob
=====================================================
Introduction: Why test films?
Since I bought my Nikon LS30, I've been suprised by the intensity
of "grain" in scans - or inkjet prints made from those scans - compared to
photographic enlargements of the same size using the same source
film. It turns out that this "grain" is an artifact of the scanning
process,
caused by the interference between the spacing and size of the CCD
elements and the random film dye clouds. The effect is enhanced by the
collimated light source used by the Nikon scanner and the sharpness
of its focus. Other factors may also be involved.
While scanning various films, it has become obvious that there are
huge differences in the "grain" that appears in the resulting image.
Consequently, it makes sense to use the films that demonstrate
this effect the least!
What is indended:
o To give some visual, qualitative guide to which films scan best
on Nikon scanners (and probably others by inference)
What is NOT intended:
o It is not a buyer's guide to film scanners, hence only Nikons
are used in the testing process
o It is not a means to quantitatively measure the performance
of Nikon scanners. More rigorous testing would be required.
o It is in NO way a comparison of the sharpness of various
cameras, lenses or even film. The appearance of "grain" is
the issue here, not the sharpness of the target image. It is
entirely possible that a film may seem grainy when scanned,
but not when printed photographically, and the same goes for
sharpness.
Caveat: Your mileage may vary, but I already know that the
differences in "grain" can be huge (eg. between Sensia II 100
and Provia 100F slide films).
What targets were chosen and why?
1) Bob Murphy suggested using the cross target. Photographed
as specified above, it gives an idea of the interference between
the "grain" and straight lines, as well as the intensity of the
"grain" in areas of a single colour.
2) Photographing a grey card was suggested since it is a standard
target in film resolution tests and is calibrated in colour.
3) I suggested the use of the USAF resolution target since it gives
some idea of the separation of parallel lines in scans, and
hence resolution. Anyone who knows how to interpret this
target should be aware that the line pair resolution from this
test should *not* be regarded as an accurate measurement
of the film resolution, scanner resolution, or combination. It
should *only* be used as a guide to compare different films.
The Protocol for target scans
======================
Part 1: Preparing the targets
I printed the USAF and Cross targets on my Epson Stylus 700
using Epson Photo paper on A4 paper at 1440dpi. If you don't
have a colour printer of comparable (or better) resolution, do
not print the cross target, but print the USAF target on A4 paper
using a laser printer of at least 600dpi. The grey card should
be a standard Kodak 10x8 photographic grey card (other
brands could be used assuming they are proper calibrated
cards). A4 and US letter are sufficiently similar in size!
Part 2: Taking the photos
1. You will need an SLR camera with spot metering and a 28mm
lens or 28mm-"n"mm zoom set to 28mm, preferably with aperture
priority.
2. Use a tripod to ensure no movement
3. Tape the target to a flat, vertical surface outside in open but
solid shade (not dappled light from a tree or in full sun!)
4. Position the camera 2 metres from the target. The camera's focus
ring should be sufficient as a guide, but preferably check with a
measuring tape to the film plane.
5. Carefully focus the camera on the target and center it in the
viewfinder. Photograph in landscape orientation NOT portrait!
6. Set the exposure for at least f8. The camera should be allowed to
determine correct shutter speed using spot metering. Do not use
exposure compensation or flash.
7. Photograph the target.
8. Preferably photograph the other targets overlaid on the first so the
camera is never moved and the conditions are as similar as possible.
You may need someone to hold the grey card in place - they're too
expensive to tape!!
9. Process the film
Part 3: Scanning the images
Please make notes of the equipment, setup and exposure
for each frame that you used so that I can include it with the actual scans.
1. Use Vuescan 6.1and a Nikon LS30 or LS2000
2. Delete vuescan.ini. Start Vuescan and set "Bits per pixel" to 48RGB.
3. Set "media type" to "Colour Negative", "Slide Film" or "B/W Negative" as
appropriate
4. Set the frame number appropriately
5. Tick the "Output Raw File" check box, clear "Output TIFF File" checkbox.
6. Use the Preview/Device menu entry to preview the frame.
7. Set the crop size *inside* the boundaries of the target using the
preview.
This should be an area of about 2.3x2.9mm according to the "Crop" tab.
8. Name the raw TIFF file as follows: "<film>_cross.tif", "<film>_grey.tif",
"<film>_usaf.tif" as appropriate - cross for the cross target, grey for
the grey card, usaf for the usaf resolution target. Insert the name of
the film you are scanning where "<film>" appears eg.
"fujivelvia_cross.tif".
9. Email the file(s) to [EMAIL PROTECTED] The TIFFs should be around
500KB each. Please send them one at a time and do NOT send entire
frame TIFFs. Alternatively if you can put sized raw files on an FTP
site, please do so and send me the URL. In any case please ensure they
are cut to the size specified above. TIFFs should be compressed by
Vuescan by default.
10. I will process the raw files using vuescan and post the results on my
web site http://wordweb.com
11. It should be possible to cut final images down to PNG format files
<100K each. Jpegs will be used on the pages as a pointer to the
PNG files. The PNG files should be used for actual comparisons.
====================
Rob
PS Looking at the USAF target test shots I took with Konica 100VX, there
is a distinct improvement in separating lines in one direction to the other.
The scanner seems to get better separation *across* the film compared to
*along* the film.
====================================================================
The filmscanners mailing list is hosted by http://www.halftone.co.uk
To resign, <mailto:[EMAIL PROTECTED]> with UNSUBSCRIBE FILMSCANNERS in the
title, or UNSUBSCRIBE FILMSCANNERS_DIGEST if you are reading the Digest.
