FS4000 + Vuescan Dust Cleaning Misalignment

[Glenn]

Hi All,

While scanning slides, I'm having intermittent problems with the
cleaned pixels being misaligned when I use the FARE IR cleaning on my
FS4000US. Filmget generally performs the cleaning properly (if
over-zealously), but the misalignment is especially evident when using
Vuescan. Sometimes it nails the dust great, other times it totally
misses it. It's all very random. In all other respects Vuescan is
working fine and far exceeds Filmget.

Reading through past posts, I see this problem has been mentioned
before -- however I haven't found a solution. Is this a known bug or a
system problem, and is there a work-around available?


System details...

I have a PC running XP, Vuescan 8.0, scanning Kodak, Fuji Velvia, and
other slide flavors, no special processing other than batch, just
single passes typically at 2000dpi but occasionally at 4000dpi.

Cheers,

Glenn

 

[Kennedy McEwen]

Hi All,

While scanning slides, I'm having intermittent problems with the
cleaned pixels being misaligned when I use the FARE IR cleaning on my
FS4000US. Filmget generally performs the cleaning properly (if
over-zealously), but the misalignment is especially evident when using
Vuescan. Sometimes it nails the dust great, other times it totally
misses it. It's all very random. In all other respects Vuescan is
working fine and far exceeds Filmget.

Reading through past posts, I see this problem has been mentioned
before -- however I haven't found a solution. Is this a known bug or a
system problem, and is there a work-around available?

As you have seen from the other posts, this is a known bug with the
Canon scanner, but perhaps the film cleaning algorithm used in Vuescan
is less tolerant of it than FARE.

The Canon scanner captures the image in two passes - one for the RGB
image and another for the IR image, used to detect the presence of
defects. Two separate passes means there is a possibility that the
image and the defect data will be misaligned - quite a a high
probability with 4000ppi. Most other scanners capture the image and IR
data in a single pass of the scan head, thus avoiding any possible
misalignment of the defect detection from the image itself.

I haven' seen any workarounds for this. :-(

 

[bmoag]

Cleanliness is next to digigodliness in scanning slides.
Multiple passes in Vuescan may decrease the need to use FARE.

 

[Andrey Tarasevich]

...
The Canon scanner captures the image in two passes - one for the RGB
image and another for the IR image, used to detect the presence of
defects. Two separate passes means there is a possibility that the
image and the defect data will be misaligned - quite a a high
probability with 4000ppi. Most other scanners capture the image and IR
data in a single pass of the scan head, thus avoiding any possible
misalignment of the defect detection from the image itself.
...

I get very good alignment with multipass scanning on FS4000 and at the
same time rather large and obvious misalignment of VueScan dust cleaner.
This is definitely not a hardware problem, but a bug in VueScan.

 

[Kennedy McEwen]

I get very good alignment with multipass scanning on FS4000 and at the
same time rather large and obvious misalignment of VueScan dust cleaner.

You may well do, its a question of tolerance build up so some samples
are better than others, but Glenn stated quite clearly, as have others,
that he is getting misalignment with FARE which is more evident in
Vuescan. The issue is fundamentally a hardware problem, however it is
possible that the Vuescan algorithm is more susceptible to it - possibly
because it uses a smaller border around detected defects than the Canon
FARE algorithm.