VueScan and Manual Focus

[Louis de Stoutz]

Can someone show me an efficient way to manually focus my Minolta SE
5400 with VueScan? It's so straightforward with the Minolta software,
but in VueScan I can't find any "realtime" indicator of the focus
setting. Don't tell me I have to scan at different settings until I find
the best one ?!?

Thanks,
Louis

 

[Mendel Leisk]

You can't do a realtime focus with Vuescan, assuming you mean the black
and white bars feedback. I would like something like that too.

On the other hand, you can shift your focus point around, and it will
stay where you put it. And, you can take readings at various points, to
get a feel for the curvature of your film.

My workflow is to just ouput a 16 bit linear thru Minolta software,
manuall focussing, with the knob, and then use it as a raw file in
Vuescan. I'm pretty well forced to do this, due to Vuescan's crappy
cleaning

The only feedback as to at what distance you are focussing at, with
minolta sw, is the indent on the knob.

 

[Louis de Stoutz]

Thanks, Mendel, for your answer.

To address the problem of curvature (which in my opinion is unacceptable
using a scanner of that potential) I currently experiment with a
makeshift glass-holder. It seems that the Minolta software can't handle
glass in the light-path, whereas VueScan works ok. I suspect it has to
do with the way the software calibrates the scanner. VueScan asks for
removal of the film before calibrating. It may be something else though.
I'll try to find out as soon as I get to do more tests. But I may have
to stick to VueScan for that reason. (There are worse things...

Louis

 

[Bart van der Wolf]

You can't do a realtime focus with Vuescan, assuming you mean the
black
and white bars feedback. I would like something like that too.

It's one of the feature enhancements I suggested when Ed Hamrick still
participated in this group. Instead of the bars Minolta uses (and I
often get 2 "best" positions), I suggested a graph which would also
give a clue about acceptable range. Another suggested feature would
allow to autofocus on several points in the crop area, and pick the
optimal DOF position for warped film.

The good think though, is that VueScan seems to focus much more
reliably than the Minolta software, especially on dense film areas. If
you want some feedback about how reliable the focus position is, try
switching to manual focus and invoke a refocus several times. You'll
see that, unless the film has not acclimatized yet, there is usually
very little variation which indicates that focus is optimal.

Bart

 

[Louis de Stoutz]

...
The good think though, is that VueScan seems to focus much more reliably
than the Minolta software, especially on dense film areas.
...

I've made the same observation. Where Minolta's SW seemed to be
constantly out of focus, making manual adjustments a necessity (why? the
two bars after all seemed to be a good indicator of sharpness, why
doesn't the SW pick the same information?), it looks like with VueScan
one might be able to rely on the autofocus alone.

Louis

 

[Don]

It's one of the feature enhancements I suggested when Ed Hamrick still
participated in this group.

What does his participation in this group have to do with any of that?

He does have email, you know.

Instead of pining over Ed you can always write to him directly at:
(e-mail address removed)

Of course, your Ed still reads this group religiously so he can send
threatening emails to frustrated Vuescan users who complain about bugs
by telling them he has "blacklisted" them - whatever that means...

Don.

 

[Bruce Gaylinn]

What does his participation in this group have to do with any of that?

He does have email, you know.

Instead of pining over Ed you can always write to him directly at:
(e-mail address removed)

Of course, your Ed still reads this group religiously so he can send
threatening emails to frustrated Vuescan users who complain about bugs
by telling them he has "blacklisted" them - whatever that means...

Don.

Careful Don, by replying to that jab you are admitting that even you
recognize that your abusive posts and dogged attacks have driven people
out of here.

 

[Don]

Careful Don, by replying to that jab

I'm afraid you're confusing a jab with mere plaintive whining and
pining after his Ed. Do check the archives to get the full context.

you are admitting that even you
recognize that your abusive posts and dogged attacks have driven people
out of here.

More confusion, I'm afraid. That doesn't follow on (m)any counts.
Here's two:

1. By his own admission the Vuescan author ran away from hordes of
angry Vuescan victims. From the horse's mouth:

2. Don has never posted an "abusive" post or "dogged attacks". Care to
give an example *with context*? Don only posts facts.

You're probably thinking of rabid Vuescan "fans" struggling with the
program's many bugs and notorious unreliability who frequently vent
their frustration by lashing out with abuse. Often followed by an
exasperated public outcry they themselves are fed up with Vuescan.

Don.

 

[HvdV]

It's one of the feature enhancements I suggested when Ed Hamrick still
participated in this group. Instead of the bars Minolta uses (and I
often get 2 "best" positions), I suggested a graph which would also give
a clue about acceptable range. Another suggested feature would allow to
autofocus on several points in the crop area, and pick the optimal DOF
position for warped film.

It shouldn't be too hard to do a small through-focus series on a tiny but
interesting region so that afterward you can pick out the best position.
Could also be done on multiple locations provided repositioning doesn't take
forever.
The Silverfast software approaches this with a before/after window on a small
ROI. Too bad it hangs all the time on this function, at least on my OSX 10.4 mac.

The good think though, is that VueScan seems to focus much more reliably
than the Minolta software, especially on dense film areas. If you want
some feedback about how reliable the focus position is, try switching to
manual focus and invoke a refocus several times. You'll see that, unless
the film has not acclimatized yet, there is usually very little
variation which indicates that focus is optimal.

The reliability also depends on the width and texture of the top of the
focus function (whatever it is). The Minolta software at least shows the
focus function and the selection in makes so that in the case of a flat
top or a doubtful selection you can try for a better focus point.

-- Hans

 

[Mendel Leisk]

Bart, I also get 2 "best" positions. Starting from the left extreme
(which put the holder furthest to the left, as well), and then moving
to the right, I find scans at the second peak to be sharper.

 

[Bart van der Wolf]

Bart, I also get 2 "best" positions. Starting from the left
extreme (which put the holder furthest to the left, as well),
and then moving to the right, I find scans at the second
peak to be sharper.

Same here with the Minolta software, but there might be differences
between units so two units is too small a sample size to draw
conclusions. Vuescan gives a single, usually (very close to), optimum.

Bart

 

[Mendel Leisk]

But it's a trend This first came to my attention when someone posted
a question re this, on photo.net.

I wonder just what is happening with the double peak. My (incorrect)
hunch would have been the inverse: that the first peak would yield
sharper result: assuming the focussing mechanism had "found" the
emulsion. And the second peak would be due to the mechanism then
finding the backing material.

I'd speculate both peaks are due to the emulsion itself, and there's
just something in the nature of the focussing mechanism that produces
the double peak.

Incidentally, I'm waiting on delivery of some:

Gepe Pro Scan Mount Art. No. 7012

and:

Wess AHX500K

Any opinion/experience with either of these? Or?

I'm hoping one or both of these will improve my corner to corner focus.
I've tried some Gepe Anti-Newton Glass Mounts. I was intitially very
happy with the results. But then, began noticing artifacts in highlight
areas. Reversing the glass in a mount where I drilled out the plastic
rivets holding the thing together, got rid of the artifacts but
introduced some spectacular Newton rings!

At least the glass mounts gave me an idea of the potential there is,
with flat film in the scanner.

 

[Mendel Leisk]

But it's a trend This first came to my attention when someone posted
a question re this, on photo.net.

I wonder just what is happening with the double peak. My (incorrect)
hunch would have been the inverse: that the first peak would yield
sharper result: assuming the focussing mechanism had "found" the
emulsion. And the second peak would be due to the mechanism then
finding the backing material.

Ok, fired up the rest of my brain cells, and now, I think the focus
behaviour does make sense. The scanner's focus is FIXED, at a certain
distance in front of the lens. With the holder at the left extreme,
with the emulsion facing the lens, that focus point is beyond the film,
to the right of it. As you move the knob/slider to the right, the
holder is moving to the right, and that focus point will first
encounter the right, backing side of the film (the first peak), then
travel thru the film, and then the left (emulsion) side of the film.