Vuescan multi exposure feature?

[Roger S.]

In Vuescan 8.4 release notes:

Significantly improved multi exposure scans
Renamed "Input|Long exposure pass" to "Input|Multi exposure"

Seems interesting but I have no intention of trying it unless other
FS4000US users report it works well. I found 8.3.75 to work correctly
and IR cleaning works and don't need any new aggravation.

 

[Guest]

Seems interesting but I have no intention of trying it unless other
FS4000US users report it works well.

And we care that you're that lazy because...?

 

[Roger S.]

Because it will make you want to rush out and test it for the greater
good of us all.

I've donated countless hours helping Ed bugfix this scanner but have
since retired from the position of unpaid Vuescan FS4000US bug tester ;
)
I suggest just scanning twice and blending scans in Photomatix because
at least I know it works properly.

 

[DenverDad]

Roger,

As a fellow retired unpaid Vuescan FS4000 bug tester, I decided to help
you out and give it a try! After downloading version 8.4.02 I
performed full resolution scans of two different slides - one Provia
100F and the other an older Kodachrome - both with and without the
multiexposure option checked. I also performed long exposure scans
manually (setting the exposure to 11, I think), so I could combine two
scans in Photomatix and compare the resluts.

In summary, what I find is that I do NOT see any of the strange
artifacts with multiexposure that I saw in the past. It seems to work
as it is supposed to - improving the shadow noise in the combined scan,
without any noticeable artifacts and seemingly without any loss of
detail. I was at first somewhat underwhelmed with the AMOUNT of
improvement in the shadows, but when I compared it with the Photomatix
result I found that it was actually about the same in this respect.
When I overlay and toggle the two final images (multiexposure vs.
Photomatix), I find that they are really more similar than different.
Curiously, the main difference I could see was that the Photomatix scan
seemed almost as if it had been sharpened slightly - in the blue sky
area of one slide, for example, the grain was noticeably more visible
than in either of the two scans which were used to create it. Frankly
this was a fairly sublte effect. It is possible I just never noticed
it before.

In any event, from this limited test it would appear that the
multi-exposure seems to be working effectively again on the FS4000 with
this Vuescan version. You may find that you are actually better off
using this feature now (when needed) instead of Photomatix. So give it
a try! I'll be curious to see what you find.

As for me, I may not be able to use this version afterall.
Unfortunately on my unit I STILL see the same old dust removal
artifacts which I have been struggling with since forever. So I may
just stay with 8.3.03. Again, no one but me seems to have this
particular flavor of artifacting, so I suspect the dust removal will
continue to work well for you and others with the latest versions.

Good luck.

Jeff

 

[Roger S.]

Thanks for the report, Jeff. I'm glad your scanner's still working.
If you have any crops you could post that would be great. I might test
it out and add it to my comparison.
I think exp "11" (really, maximum is 6) is probably a bit too much-
I've found exposures of 2 and 4 or 3 and 5 are better for combining as
you avoid some of the CCD blooming.
here:http://jingai.com/vuescan2/long exposure comparison.html

I didn't see any significant IR artifacts with Vuescan 8.3.75 (yes, if
you get up to 400% or so there are little dots as you have shown, but
this is invisible on the 8x12 prints that I do). I can send this VS
version to you if you want. IR exposure for negs has to be bumped up
to 3 I believe (let me double-check- either 2 or 3) and set to heavy to
have a real cleaning effect.

Here's a sample of cleaning:
http://jingai.com/vuescan2/reala-8.3.81-none.jpg
http://jingai.com/vuescan2/reala-8.3.81-Heavy.jpg

8.3.81 had other bugs which made be go back to .75 but the above
samples look about the same.

Roger

 

[DenverDad]

Sorry I haven't been able to get back to this sooner, but the inlaws
have been camping out with us for a while, and... well, you know how
that goes!

So have you had a chance to try out the newer versions? If not I will
try to get some crops posted somewhere. Actually, I am in the middle
of trying to choose a new online photo sharing site (got disenchanted
with my previous one and closed it out), but I'm sure I can find
something.

And yes, you're right, its just those pesky "little dots" that I
continue to see! Thanks for the reminder about the exposure levels.
For some reason I thought the maximum was 8 or 11.

Jeff

Thanks for the report, Jeff. I'm glad your scanner's still working.
If you have any crops you could post that would be great. I might test
it out and add it to my comparison.
I think exp "11" (really, maximum is 6) is probably a bit too much-
I've found exposures of 2 and 4 or 3 and 5 are better for combining as
you avoid some of the CCD blooming.
here:http://jingai.com/vuescan2/long exposure comparison.html

I didn't see any significant IR artifacts with Vuescan 8.3.75 (yes, if
you get up to 400% or so there are little dots as you have shown, but
this is invisible on the 8x12 prints that I do). I can send this VS
version to you if you want. IR exposure for negs has to be bumped up
to 3 I believe (let me double-check- either 2 or 3) and set to heavy to
have a real cleaning effect.

Here's a sample of cleaning:http://jingai.com/vuescan2/reala-8.3.81-none.jpghttp://jingai.com/vuescan2/reala-8.3.81-Heavy.jpg

8.3.81 had other bugs which made be go back to .75 but the above
samples look about the same.

Roger

Roger,

As a fellow retired unpaid Vuescan FS4000 bug tester, I decided to help
you out and give it a try! After downloading version 8.4.02 I
performed full resolution scans of two different slides - one Provia
100F and the other an older Kodachrome - both with and without the
multiexposure option checked. I also performed long exposure scans
manually (setting the exposure to 11, I think), so I could combine two
scans in Photomatix and compare the resluts.

In summary, what I find is that I do NOT see any of the strange
artifacts with multiexposure that I saw in the past. It seems to work
as it is supposed to - improving the shadow noise in the combined scan,
without any noticeable artifacts and seemingly without any loss of
detail. I was at first somewhat underwhelmed with the AMOUNT of
improvement in the shadows, but when I compared it with the Photomatix
result I found that it was actually about the same in this respect.
When I overlay and toggle the two final images (multiexposure vs.
Photomatix), I find that they are really more similar than different.
Curiously, the main difference I could see was that the Photomatix scan
seemed almost as if it had been sharpened slightly - in the blue sky
area of one slide, for example, the grain was noticeably more visible
than in either of the two scans which were used to create it. Frankly
this was a fairly sublte effect. It is possible I just never noticed
it before.

In any event, from this limited test it would appear that the
multi-exposure seems to be working effectively again on the FS4000 with
this Vuescan version. You may find that you are actually better off
using this feature now (when needed) instead of Photomatix. So give it
a try! I'll be curious to see what you find.

As for me, I may not be able to use this version afterall.
Unfortunately on my unit I STILL see the same old dust removal
artifacts which I have been struggling with since forever. So I may
just stay with 8.3.03. Again, no one but me seems to have this
particular flavor of artifacting, so I suspect the dust removal will
continue to work well for you and others with the latest versions.

Good luck.

 

[Roger S.]

So have you had a chance to try out the newer versions? If not I will
try to get some crops posted somewhere. Actually, I am in the middle
of trying to choose a new online photo sharing site (got disenchanted
with my previous one and closed it out), but I'm sure I can find
something.

I haven't tried it yet- I've been processing a lot of raw scans
sitting around on my hard drive.
I've also been comparing scanner profile software. I still think
highly of Scarse but found that LCMS profiles if set to accurate
perceptual are quite good and come in a bit brighter when applied to
slides than Scarse (Scarse is more conservative and tries to keep
highlights from clipping). Both do a much better job than the built
in Vuescan profile engine. I've created profiles with and without
scanhancer. Nominal slide exposure is 2 (1 is too dark unless there
are very bright highlights) and 6 with the scanhancer mounted.

And yes, you're right, its just those pesky "little dots" that I continue to see! Thanks for the reminder about the exposure levels. For some reason I thought the maximum was 8 or 11.

Vuescan used to set 11 as the maximum, but it's really integers from
1-6.