Canon FS4000US focus

[Brian Wickham]

I'm not very happy with the scans I get from my FS4000 as they all
seem to be just ever so slightly out of focus. I know many of the
negs and slides I scan are pin sharp but they never seem to come out
that way when I scan them. Is there some secret to the FS4000 that I
don't know about? Is there as way to guarantee that the unit is
focusing properly?

Brian

 

[Brian Wickham]

I'm not very happy with the scans I get from my FS4000 as they all
seem to be just ever so slightly out of focus. I know many of the
negs and slides I scan are pin sharp but they never seem to come out
that way when I scan them. Is there some secret to the FS4000 that I
don't know about? Is there as way to guarantee that the unit is
focusing properly?

I should have mentioned that I have been using Vuescan for the past
few years and am currently using v. 8.3.80

Brian

 

[Steven Saunderson]

I should have mentioned that I have been using Vuescan for the past
few years and am currently using v. 8.3.80

Have you tried the standard Canon software to see if it does any better?

The focus operation is fairly automatic from the PC perspective. The PC
software just issues the command and the scanner does the rest. It is
possible (if I remember correctly) to manually adjust the focus but I
don't know any software that tries to do this.

 

[DenverDad]

Brian,

I can report that my FS4000 has always focused well, as have those of
others I have talked with on here. If anything, the focusing ability
of this scanner is considered one of its strengths since (at least by
some accounts) it has greater depth of focus than some of the other
brands. Hopefully, focus deficit you are seeing is just the result of
software settings, as opposed to a hardware problem - which obviously
would be a more $eriou$ problem.

The first thing I can suggest is to see what VueScan settings you are
using for focus. I don't have the software in front of me at the
moment, but as I recall the settings for focus are all under the Input
tab. Make sure that you are set to autofocus as opposed to manual
focus; then check which particular autofocus setting you are using.
As I recall you can select to focus during the preview only, during
the scan only, or always (other settings?). It is probably best to
select "always" so that there is no doubt that the scanner will try to
find best focus each time. The next thing to check is where in the
frame the scanner is measuring focus. You may be aware of this
already, but there is a small "+" symbol which appears in the preview
scan. This is the location in the frame where the focusing routine
measures to find best focus. Since there is always at least a little
bowing of the film in its holder, it is important to set the mark to
an appropriate location. Make sure that it is NOT placed near the
edge of the frame (or worse yet, out of the frame). Rather than
placing this mark right in the center of the frame, I find that it is
usually best to place it somewhat away from the center - say half way
between the center and an edge. This makes best use of the depth of
field of the scanner optics and usually insures that the entire frame
will be in sharp focus. You should be careful that this mark falls on
an area which has high contrast and is not too dark. Otherwise the
scanner may not have enough to work with in determining the best
focus.

If all these software settings seem right, the next thing you might
want to try is to perform a manual focus test. Without the software
in front of me I don't remember exactly how you go about selecting
manual focus. But once you have it selected, perform a number of
scans varying the focus depth over some range. You should see that
there is an optimum depth at which the scans appear well focused, with
the scans being blurrier on either side of this. If the scanner's
focus mechanism is not working (or if the software is not accessing it
properly) these scans will probably all appear to have the same degree
of focus (i.e., equally blurry). On the other hand, if you can
clearly see that there is a best focus but the best scan is STILL not
acceptably sharp, then it may indicate an optics problem in the
scanner.

Finally, Steven's suggestion to try FilmGet is a good one. You could
also try a different version of VueScan to make sure that this isn't
just a bug with that particular version.

Anyway, that's all I have for you at the moment. Let us know if any
of this helps!

Good luck,

Jeff

 

[Roger S.]

I agree with everything Jeff said. To add to it try to select an area
with some contrast (instead of a uniformly bright or dark area try
picking the border between the two).

 

[Colin_D]

I'm not very happy with the scans I get from my FS4000 as they all
seem to be just ever so slightly out of focus. I know many of the
negs and slides I scan are pin sharp but they never seem to come out
that way when I scan them. Is there some secret to the FS4000 that I
don't know about? Is there as way to guarantee that the unit is
focusing properly?

Brian

Do you have sharpening enabled (if the scanner can do that) or have you
tried sharpening in your image editor?

All digital images, whether from a camera or a scanner require
sharpening to get the best results.

Colin D.

 

[Steven Saunderson]

I'm not very happy with the scans I get from my FS4000 as they all
seem to be just ever so slightly out of focus.

Do you have access to another scanner so you can compare results ?
Perhaps your scanner has a film on the mirror or lens that is affecting
the results.

I believe the auto-focus operation uses only one vertical line in the
scan. The horizontal position (scan line #) is determined by the
carriage position and the vertical top and bottom are specified in the
command parameters.

Auto-focussing might not be that precise an operation but if all your
scans are soft (rather than just some being soft) I would suspect the
fault is somewhere else.

 

[Roger S.]

Auto-focussing might not be that precise an operation but if all your
scans are soft (rather than just some being soft) I would suspect the
fault is somewhere else.

Feel free to post a crop at 100% at 4000dpi and I can let you know how
it compares to my scanner. Slides absolutely need 2 passes (capture
and output) with a program like PhotoKit Sharpener so as not to look
soft. Color negative benefits from it as well to a lesser degree.

You can tell if focus is off by whether the grain gets more or less
distinct across the frame (chances are part is more in focus than
another part). If the grain is uniformly soft, move the autofocus
point and try again. If the grain is sharp but image is soft you need
to sharpen or work on capture technique.

 

[Brian Wickham]

I believe the auto-focus operation uses only one vertical line in the
scan. The horizontal position (scan line #) is determined by the
carriage position and the vertical top and bottom are specified in the
command parameters.

If I read you right an improperly aligned strip in the carriage could
have a deleterious effect on the focusing function. Is that correct?

Brian

 

[Brian Wickham]

Do you have sharpening enabled (if the scanner can do that) or have you
tried sharpening in your image editor?

All digital images, whether from a camera or a scanner require
sharpening to get the best results.

This is a point I was unaware of. I did try sharpening in Vuescan and
found that it increased grain, but nevertheless did increase the
readability of some details.

Brian

 

[Brian Wickham]

Brian,
You may be aware of this
already, but there is a small "+" symbol which appears in the preview
scan. This is the location in the frame where the focusing routine
measures to find best focus. Since there is always at least a little
bowing of the film in its holder, it is important to set the mark to
an appropriate location. Make sure that it is NOT placed near the
edge of the frame (or worse yet, out of the frame). Rather than
placing this mark right in the center of the frame, I find that it is
usually best to place it somewhat away from the center - say half way
between the center and an edge. This makes best use of the depth of
field of the scanner optics and usually insures that the entire frame
will be in sharp focus. You should be careful that this mark falls on
an area which has high contrast and is not too dark.

Jeff & Roger,

Thanks for reminding me of the focus indicator. I had been neglecting
to observe its position on each scan. I tried all of the suggestions
concerning manipulating Vuescan and I found that the best results, for
me, come from making a multi-pass scan.

Since I'm in the habit of scanning complete rolls I don't normally
bother with multi-passes. I will do that in the future on individual
frames that are to be displayed or printed.

Thanks to you two plus Colin D. and Steven Saunderson for the input.
All of the information was helpful in analyzing my work habits and my
equipment.

Brian

 

[Steven Saunderson]

If I read you right an improperly aligned strip in the carriage could
have a deleterious effect on the focusing function. Is that correct?

Yes it could but I haven't expressed myself very clearly here. The
carriage holds the film holder and is positioned by the fine stepper
motor. The carriage contains a coarse stepper motor which moves the
film holder. All this movement is along the X-axis. Focussing is done
by moving the lens along the Z-axis. It would be difficult to misalign
a film strip in the holder but I have had trouble with curled film that
doesn't sit flat in the holder.

The PC software will focus the scanner using two commands. The first
will move the film holder (using the coarse stepper motor) to select the
position on the X-axis. Then the focus command is done and the FS4000US
moves the lens (on the Z-axis) and finds the sharpest scan of the
vertical line. Two parameters in the focus command specify the limits
(on the Y-axis) of the scanned line.

Vuescan allows you to select the focus point but after I viewed a log it
looks like only the scan line number is used. Vuescan calls this the
Y-point because the scan lines in the display are horizontal rather than
vertical as they are in reality. The other setting (X-point) appears to
be ignored and the area starts at pixel 500 and stops at pixel 3500.

I think Roger's offer of commenting on one of your scans is excellent.
If you like I can upload a test scan so you can view it for comparison.
If you need a temporary location for uploading one of your scans please
ask and I will arrange something.