Derek - you have gone to a lot of trouble to compare the focus performance
of the two different holders for your scanner, but it is really hard to
draw conclusions or interpret this.
Can I suggest that to get a more direct comparison, just stick a film in
each holder, and measure the 'auto' focus position at several positions
including at least each corner, at the middle of each side, and at the
centre of the image. This will give you at least 9 focus positions. If
the film was perfectly flat and perfectly aligned to the scan path, all
these numbers would be the same. If the film is curved or not parallel to
the scan path, the numbers will vary. The amount of the variation tells
you the DOF you would need to correctly scan the image so that it was all
in focus.
In the end you get only one number for each holder - the total variation in
focus position (= the range of these numbers)- which is easy to compare
with the number for the other holder.
When I did it, I got a total variation of over 60 nikon focus units for the
auto feeder, and at best around 20 for the manual holder. SInce my DOF is
about 12(critical) to 20 (acceptable) this means that I have a problem with
anything except the manual holder on its best behaviour. I have swapped
scanners since then and the new one is somewhat better.
Julian
At 18:37 06/02/03, you wrote:
>Julian,
>
>I went back and used your methodology described below to determine the
>critical focus range of my LS4000 and my results seem to be generally what
>you thought they might be.
>
>Namely, critical focus is still intact at -5/+5 Nikon Units from focal
>point (0). Somewhere between -5/5 and -10/10 it softens enough to remain
>within acceptable focus, but by -10/10 it seems to be clearly out of
>critical focus. By -15/15 the focal shift is significant. However, one
>caveat: all my above determinations are being made when viewing a very
>small portion of the total 24mm x 36mm image and at a highly magnified view.
>
>Given this, I am still hard pressed to say what would really appear as out
>of critical focus when the total image has been printed, at lets say 8x10,
>11x14, 16x20, and viewed in total. Tough to say. A bit of a slippery
>slope, really.
>
>However, if one is going for the most pristine, most ideal scan and one
>wants to make sure that even a small selection of the total image is
>printed at a decent size, I would imagine one would want film curl not to
>range outside of the -5/5 range if possible and definitely not outside the
>-10/10 range.
>
>To this end, I have been considering using the slide mounts with
>Anti-Newton glass as quite a few posters have mentioned
>previously. However, results appear mixed. Some say it works very nicely
>for them, keeping film flatness within a 0-3 Nikon Unit range w/o any loss
>of image clarity and sharpness. Yet, others seem to indicate this is not
>the case at all, claiming that the scanner even picks up the diffused
>patterns characteristic of the AN glass.
>
>What have your findings been in this area? I would be very interested to
>hear.
>
>Lastly, I would like to post my DOF test results as they relate to the
>FH-3 manual strip holder vs. the SA-21 auto strip feeder. And here, I'm
>afraid my findings still tend to stand out as anomalous.
>
>The reason here being that I pretty clearly got better DOF results (in
>direct enlargement to enlargement comparisons) with the SA-21 Auto Strip
>Feeder than with the FH-3 Manual Strip Holder.
>
>I ran the tests on 2 different images: one was on the inside of a film
>strip and the other was on the end of the strip. Each one was scanned
>with the SA-21 and the FH-3 adapters (in both the LS4000 UPRIGHT and SIDE
>orientations) and then compared at various view sizes and at various DOF
>settings.
>
>Here is the side by side breakdown of the results (granted, this is
>ultimately a subjective determination):
>
>INSIDE IMAGE:
>
>At Focal Point (0): SA21 sharper than FH3
>(Side and Upright about the same sharpness; Side perhaps a little sharper).
>
>At (+5): SA21 sharper than FH3
>(Side clearly sharper than Upright).
>
>At (-5): SA21 sharper than FH3
>(Upright now sharper than Side).
>
>At (+10): SA21 sharper than FH3
>(Side especially sharper; Upright closer to FH3)
>
>At (-10): SA21 sharper than FH3
>(especially Upright; Side better than FH3, but not as sharp as Upright)
>
>Summary:
>
>SA21 sharper than FH3 at all dofs. Side and Upright alternate; Side
>sharper at focal point and points above; Upright sharper at points below
>focal point.
>
>
>
>OUTSIDE IMAGE:
>
>At Focal Pt (0): FH3 a little sharper than SA21 in some places.
>
>At (+5): SA21 becomes a little sharper than FH3 in some places, otherwise
>even.
>
>At (-5): FH3 clearly sharper than SA21
>(except at lower left corner)
>
>At (+10): SA21 clearly sharper than FH3
>(except at lower left corner)
>
>At (-10): FH3 clearly sharper than SA21
>(except at lower left corner)
>
>Summary: SA21 and FH3 alternate at differing dofs; FH3 sharper at focal pt
>and at points above, while SA21 sharper at points below focal pt.
>
>Regards,
>
>Derek
>
>
>----- Original Message -----
>From: "Julian Robinson" <jrobinso@pcug.org.au>
>To: <golder@netzero.com>
>Sent: Saturday, January 25, 2003 1:26 PM
>Subject: [filmscanners] Re: Nikon LS-4000ED Depth of Field Revisited
>
>
>Hi Derek,
>
> >In the web site above, critical focus is maintained with a +/-12 Nikon
> >unit range, and decent focus within a +/-24 range. How this translates to
> >the LS4000, I don't know.
> >
> >I have looked at the scans super magnified and tried to discern what makes
> >for a critical focus range in the LS4000 and what makes for a decently
> >focused range (using NikonScan focus units) by looking at grain structure,
> >but I fear I have not been to successful with this method.
>
>I was very interested to read your comments - it is good to hear that at
>least some examples of the LS4000 seem to work well re focus. As Peter
>said, the extent of the focus problem definitely varies a lot of between
>individual scanners.
>
>But I am bothered that you couldn't determine the critical focus range -
>particularly since I have been waiting for a year or so for someone to do
>this on an LS4000!! (it was my page you were looking at). Can you have
>another go? It would be very instructive for LS4000 owners, of which I am
>not one, and also for me to know if the LS4000 is an improvement over the
>LS2000 in this respect.
>
>It should be easy to do, unless I am missing something about the LS4000
>that interferes with the method. Let me try another description to see if
>it helps. Pls don't be insulted by the level of detail, I am trying to make
>sure we are doing the same thing.
>
>1) Use a negative, neg is better because scans have more apparent grain to
>play with.
>2) Do a preview and crop the image on the preview to be a small area around
>some part of the neg with obvious grain
>3) Do a "manual autofocus" on that point, read the focus number in Nikon
>units - call your reading X. By "manual autofocus" I mean :
>
> - hold down on the control or command key and click on
> the focus button (the one like a checkered flag)
> - now click on your test area on the preview (the cursor
> should have changed to a gunsight)
>
>4) Scan, save the scan and enlarge in PS or whatever. Note that the grain
>is sharp.
>5) Now, manually set the focus point to X + 5 or 10 units. To do this,
>type the required value directly into the "Manual Focus Adjustment" box on
>the Scanner Extras palette. Repeat the scan and check if the grain is
>still sharp.
>6) Repeat 5) as often as necessary increasing the focus point value each
>time, (moving the lens more and more away from the correct focus point)
>until the resulting scan has clearly lost grain sharpness
>7) Repeat 5) and 6) but this time setting the focus point to LESS than the
>auto-derived focus value (i.e. X-5, X-10 etc), until the image again has
>definite soft grain.
>8) You should now have a series of little images with names like "+5",
>"-20". Line them up in Photoshop or whatever, in order, and pick the two
>(a plus value and a minus value) at which the grain first becomes
>definitely soft. The difference between them is the DOF in Nikon units -
>to grain sharpness level. I did the same exercise again, but looking at
>the *image* sharpness disregarding the fact that the grain was obviously
>soft and got another figure - a greater range - over which the image was
>acceptably sharp for my purposes. This gave me a kind of "worst case" -
>the actual range which I had to keep my film within if the image was to be
>usable.
>
>The first time you do step 5, I suggest you choose an outlandish figure
>like X + 50 just to check that the method is working. If the resulting
>test image is not way our of focus then there is a problem with my
>description.
>
>Hope this helps, because it is not much use knowing the curviness of your
>images if you don't know the scanner DOF. I look fwd to your results. If
>any other LS4000 user has done this measurement can you tell us your
>results pls? (Or LS2000, 30, 8000 for that matter).
>
>Julian
>
>
>
>----------------------------------------------------------------------------
>------------
>Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
>filmscanners'
>or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
>or body
>
>
>
>
>----------------------------------------------------------------------------------------
>Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
>filmscanners'
>or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
>or body
Julian
Canberra, Australia
Satellite maps of fire situation Canberra and Snowy Montains
----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or
body
