Manual Focusing - VueScan/NikonScan
Tom Harrison Jr.
news:aj0k8b$17gec0$1...@ID-79189.news.dfncis.de...
>
> "Bart van der Wolf" <bvd...@nospam.nl> wrote in message
> news:aj0j6l$80l$2...@news1.xs4all.nl...
> >
> > It is only a different scaling, so setting a value to 0.122 or 0.123
will
> > achieve the same effect. You will see that sometimes the number jumps to
> > another value if the hardware cannot accomodate such fine increments.
>
> Hi Bart... yes, i realise it's a only a difference in scaling... it's this
> difference that i'm trying to arrive at.... meaning, in VS, 0.008 to 0.032
> is a difference of .024 how does this compare to, for instance,
NikonScan's
> difference of 220 to 228 = 8?
>
> I'm trying to resolve the areas that will remain in critical focus using
the
> LS4000 and VueScan....
>
[snip]
Hi --
I think this just some linear algebra? So if -1 to 1 in VS maps to 0 to 320
in Nikon LS-40, then
N = 160( 1 + V )
or
V = (N/160) - 1
where N = Nikon setting and V = VueScan setting.
Or perhaps I am missing the question...
Tom Harrison
Jay Jhabrix
Thanks for your reply....
I think you're right... it's probably just a question of relationships....
:)
However, my maths is rather bad and i wouldn't have a clue as to how to go
about it.... keeping in mind also the fact that ViewScan in manual focus
mode goes to 3 decimal places... (or is that not important?)
Think you could work it out for me? In effect, what i'm looking for is the
relationship... meaning if on the NS scale, the difference of 220 and 221
="1", then on the VueScan scale what does that "1" equal. Is it 0.003 or
0.033 or 0.???
The reason for this question is as follows... the LS 4000ED has a
smaller/shallower depth of focus than most other scanners... so film
flatness is critical. Film is generally not flat. Now there was this article
that said that for critical focus a range of 12 points was a must... meaning
if your focus readings showed 220 at the lowest end and 232 at the highest
then setting your focus to 226 was result in everything between 220 and 232
being in sharp focus. Or, if the readings were 220 to 240 then you'd have to
decide on the critical areas and set your manual focus so that 6 points on
either side of that setting would be sharp and the rest a bit softer. Have i
been able to explain my problem?
I suppose, put another way it's just this... using VS i need to know what is
the range that will make sure that everything is in sharp focus on either
side of the mid-point....
Ever so grateful for any help... virtual beer? :)
Thanks...
JJ
"Tom Harrison Jr." <tom.ha...@nospam.attbi.com> wrote in message
news:Zl069.76720$UU1.14909@sccrnsc03...
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.381 / Virus Database: 214 - Release Date: 02/08/2002
Tom Harrison Jr.
> Tom...
>
> Thanks for your reply....
>
> I think you're right... it's probably just a question of relationships....
> :)
>
> However, my maths is rather bad and i wouldn't have a clue as to how to go
> about it.... keeping in mind also the fact that ViewScan in manual focus
> mode goes to 3 decimal places... (or is that not important?)
>
> Think you could work it out for me? In effect, what i'm looking for is the
> relationship... meaning if on the NS scale, the difference of 220 and 221
> ="1", then on the VueScan scale what does that "1" equal. Is it 0.003 or
> 0.033 or 0.???
Well, ok, so according to Ed's numbers, the LS4000 focus settings of 0 to
450 map to VueScan's -1 to 1 range. This makes the formula:
VueScan Setting = (Nikon setting/(450/2)) - 1
So if your Nikon setting is 220, the corresponding VueScan setting
is -.0222;
( 220/(450/2)) - 1 = (220/225) - 1 = .9777 - 1 = -.0222
If your Nikon focus setting is 221, the corresponding VueScan setting
is -.0177
Further, this means that a single focus point on the Nikon is represented by
a difference of .0045 in VueScan.
I am not sure how precisely any scanner focuses -- it is probably closely
enough, but the fact that there is a scale from 0 to 450 does not
necessarily mean there are 450 discrete focus points, or even if there are,
that the distinction between any two matters. VueScan is controlling the
hardware and chances are that it is rounding the value entered such that it
can set each integral point that the Nikon can set. Put in other words, I
think it's very probable that VueScan can set the exact same settings as the
Nikon.
But I think we may be focusing on the wrong thing (ok, the pun was
intentional :-). As you point out, the curve of the film is important, and
this will vary for each frame of each roll -- indeed, it may even vary
between the time you preview and final scan due to the heat of the scan.
In the case where the film is curved there will be no one correct focus
point for all parts of the film. According to an expert who posted on
this newsgroup a while back, the optimal place to focus is something like
60% of the way from the center (presumed to be the furthest away from the
film) to the edge (presumed to be the closest to the film). This probably
assumes that your film is in a holder and held down pretty well on four
edges by the holder. This accounts for the fact that depth of field is
shorter the closer the lens is to the subject.
But again, I think this may be an exercise in theory. Can you actually
observe a difference? If so, forget the theory and go directly to practice
and set the focus that looks best.
And if your film is that badly curled, you may want to consider a different
photfinisher -- mine comes out nearly flat, and if I treat it well, it stays
that way (I am not sure if this is because I am using a professional to
develop or because I am using pro film and they might treat it better :-)
Tom
Jay Jhabrix
few questions tho' if you could help out....
> Well, ok, so according to Ed's numbers, the LS4000 focus settings of 0 to
> 450 map to VueScan's -1 to 1 range. This makes the formula:
> VueScan Setting = (Nikon setting/(450/2)) - 1
That's a bit that's got me confused... What does Ed mean by mapping to 450?
The focus points in NS 3.1.2 on the LS 4000 only go up to 323 (0-323).
> Further, this means that a single focus point on the Nikon is represented
by
> a difference of .0045 in VueScan.
Thanks for the math :) Just one point... my point of 323 vs 450 above....
> I am not sure how precisely any scanner focuses -- it is probably closely
> enough, but the fact that there is a scale from 0 to 450 does not
> necessarily mean there are 450 discrete focus points, or even if there
are,
> that the distinction between any two matters.
mmmmmm... makes sense....
> But I think we may be focusing on the wrong thing (ok, the pun was
> intentional :-).
LOL!
> In the case where the film is curved there will be no one correct focus
> point for all parts of the film. According to an expert who posted on
> this newsgroup a while back, the optimal place to focus is something like
> 60% of the way from the center (presumed to be the furthest away from the
> film) to the edge (presumed to be the closest to the film). This probably
> assumes that your film is in a holder and held down pretty well on four
> edges by the holder. This accounts for the fact that depth of field is
> shorter the closer the lens is to the subject.
Right.... that's why i'm trying to work out the depth of focus/field using
VS...
> and set the focus that looks best.
Sometimes difficult to judge on screen.... :(
> And if your film is that badly curled, you may want to consider a
different
> photfinisher -- mine comes out nearly flat, and if I treat it well, it
stays
> that way (I am not sure if this is because I am using a professional to
> develop or because I am using pro film and they might treat it better :-)
35mm film always has that inherent curl... but that per se is not such a
problem Tom... i do use a Kodak pro lab for all my processing... and
generally - hell, 99% of the time - i use the FH3 holder.... i find that the
focus points usually fall comfortably within the range of '12' using the
NikonScan. It's just that i want to start using VS more and am trying to map
that '12' to the VS numbers....
Again, thanks a lot for the clarifications and the maths... :) really am
grateful... could never have worked it out myself... (history honours
student!) Just, if you could resolve that 450 vs 323 business for me... i
think i'll be on my way!!!!
Cheers and all the best...
JJ
Jeff Randall
Your first post in this thread indicated that you have experimented
with NS and found that a "spread" of 12 NS units keeps everything in
focus for you. Why don't you do the same experiment with VueScan to
determine its "focus range" rather than fussing with algebra and
theory? This would also directly eliminate the 0-450 or 0-323
problem.
Jeff Randall
"Jay Jhabrix" <jha...@yahoo.com> wrote in message news:<ajffp9$1b7rdv$1...@ID-79189.news.dfncis.de>...
Jay Jhabrix
news:fdd0dd01.0208...@posting.google.com...
> Jay:
>
> Your first post in this thread indicated that you have experimented
> with NS and found that a "spread" of 12 NS units keeps everything in
> focus for you. Why don't you do the same experiment with VueScan to
> determine its "focus range" rather than fussing with algebra and
> theory? This would also directly eliminate the 0-450 or 0-323
> problem.
Jeff...
Yes and no.... i came across an article and then tried out the
recommendations and found them to be pretty much correct... but it was an
extremely painful exercise... :( the thought of going thru it all over again
is well.... not too happy a scene... Hence, was looking to see if there was
some way to relate the VS scale to the NS scale. Tom is definitely helping
out here.... the only question at this stage seems to be whether it's 450 or
323. If that's resolved, then i can use Tom's calculations and home in on
the correct settings....
As i've hinted at, my problem is that numbers/figures are, by and large,
double dutch to me... i'm the kind of person who really needs a calculator
to add up three numbers.... :(
Regards... and thanks for taking the time to write...
Tom Harrison Jr.
question. Ed is usually right :-)
But the Nikon software may not even represent the same number of focus
points as the actual scanner can control (maybe they use the same software
as for the other scanners that have a different range). So when you say
there is a 12 point range of tolerable focus, those 12 points may even be on
a different scale than supported by the underlying hardware!
I have to agree with Jeff's post, however, if this is something you cannot
see, then this is a pointless exercise in math.
There are several points you should consider:
* every frame will have a different "correct" focus
* the depth of field of your scanner is independent of the software that
controls it
* VueScan will set focus exactly the same as the Nikon software
* The range of focus points is pretty much arbitrary. It could be from 1 to
200, 17 to 512, -1 to 1, 0 - 450, whatever. Like Celsius versus Fahrenheit,
there are simply two different ways of expressing how cold or hot it is.
* And all bets are off if the focus range is not linear, since that 12 point
range you mention may only apply at a given starting point. For example, if
you start at 200 you can go plus or minus 6 and still have acceptable focus.
But what about if you started at 15? Is the range still from 9 to 21 (plus
or minus 6)? I don't know, but very little in optics is linear, so I doubt
it. Which is only just to say that it is questionable that the first
(painful) exercise you went through may have lead you to a false conclusion.
* If you cannot see whether focus is acceptable on your monitor, all you
have to do is zoom in on the image. At some level of zoom, the monitor's
resolution is plenty to see the actual pixels.
Tom
"Jay Jhabrix" <jha...@yahoo.com> wrote in message
news:ajgk6e$1bik1k$1...@ID-79189.news.dfncis.de...
Jeff Randall
> "Jeff Randall" <jran...@ch2m.com> wrote in message
> news:fdd0dd01.0208...@posting.google.com...
> > Jay:
> >
> > Your first post in this thread indicated that you have experimented
> > with NS and found that a "spread" of 12 NS units keeps everything in
> > focus for you. Why don't you do the same experiment with VueScan to
> > determine its "focus range" rather than fussing with algebra and
> > theory? This would also directly eliminate the 0-450 or 0-323
> > problem.
>
> Jeff...
>
> Yes and no.... i came across an article and then tried out the
> recommendations and found them to be pretty much correct... but it was an
> extremely painful exercise... :( the thought of going thru it all over again
> is well.... not too happy a scene... Hence, was looking to see if there was
> some way to relate the VS scale to the NS scale. Tom is definitely helping
> out here.... the only question at this stage seems to be whether it's 450 or
> 323. If that's resolved, then i can use Tom's calculations and home in on
> the correct settings....
>
> As i've hinted at, my problem is that numbers/figures are, by and large,
> double dutch to me... i'm the kind of person who really needs a calculator
> to add up three numbers.... :(
>
> Regards... and thanks for taking the time to write...
>
> JJ
Jay:
It's really not rocket science. All you need to do is to drag the
VueScan focus point to your choice of focus area on your test negative
or slide and manually focus with VS. Then vary the VS focal point in
increments of say 0.005 (or if you want quicker results maybe 0.01) on
either side of the VS focus value and check the clarity at 1:1
magnification in your photoeditor. Note: Don't use the spin buttons
or the slider to adjust the focus point -- manually enter the values.
For example, if the VS focus value is 0.004, then you should increase
the focus to 0.009, 0.014, ... etc and then -0.001, -0.006 ... etc.
In about 10 to 12 scans you should arrive at your focus range.
Jeff Randall
Jay Jhabrix
you're absolutely right... i think i'm tying myself up in knots, getting
myself into a situation where i can't see the wood for the trees....
Thanks Tom for the calculations.... they will help as a starting point....
Thanks both of you for putting things in perspective :)
Cheers... and have a great weekend!
JJ
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.381 / Virus Database: 214 - Release Date: 03/08/2002
Pasi Savolainen
> Tom... Jeff....many thanks for your comments, feedback and suggestions....
> you're absolutely right... i think i'm tying myself up in knots, getting
> myself into a situation where i can't see the wood for the trees....
I don't remember your sacnner, but why don't you use VS:s autofocus for a
slide?
I mean, look at slide, notice where it's curve is in '50%' range of whole
range of curvature, then set VS:s autofocus point (cross thing) at that
point (there needs to be contrasty area there, just like with older camera
AF), and scan. After that turn manual focus on, and there'll be the focus
point VS got (remember to check scan for sharpness). Do the same for 0%
point and 100% point (probably in dead-center and very near frame border),
and you'll get your scanner's AF-variation and deviation caused by
manufacturing imperfections.
--
Psi -- I need no force <http://www.iki.fi/pasi.savolainen>
Vivake -- Virtuaalinen valokuvauskerho <http://members.lycos.co.uk/vivake/>
Jay Jhabrix
"Pasi Savolainen" <pvsa...@cc.hut.fi> wrote in message
news:ajigad$dmbf$2...@midnight.cs.hut.fi...
Pasi...
Thanks for the suggestion.... will try it out....
Cheers...
JJ
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.381 / Virus Database: 214 - Release Date: 02/08/2002
PW
with my Coolscan IV.
Peter
Jay Jhabrix
"PW" <pw...@powerup.com.au> wrote in message
news:3D5CF0B7...@powerup.com.au...
> Glass mount the slide and save yourself a lot of time. I gave up on manual
focus
> with my Coolscan IV.
I do that sometimes... but prefer not to... reasons as follows:
1/ More handling of the slide
2/ Storage is fussy ( i prefer to keep them in strips of 6 in the wallet)...
i never have my slides mounted
JJ
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.381 / Virus Database: 214 - Release Date: 02/08/2002
Jay Jhabrix
a bit of seriious scanning... :) and my thanks to all of you are long
overdue.... so:
Thanks to all of you who replied and attempted to help me. Special thanks to
Tom Harrison Jr.
Tom, those formulae work perfectly... thanks again... and yes the difference
is excatly 0.0045! I did another test... fortunately, the film holder stays
static in the LS4000... so it worked.... made a careful note of two points
(well, several actually) in Vuescan and NS and used the formula to work it
out/crosscheck... the difference was bang on.... Now have it all set up in
Excel so just enter the figures and know if i'm within the 'range'. Your
point about the range not being linear is kept in mind but i find that by
and large this gives me a quick method to check if i'm in the ballpark. One
advantage i have is that i'm usually scanning just a handful of slides at a
time.... so it seems to be working for me.
Thanks again Tom.... Thanks again guys....
Cheers....
JJ
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.385 / Virus Database: 217 - Release Date: 04/09/2002