Flare on Minolta 5400

[Jason]

Hi,

I've got a Minolta DSE 5400 that's a few months old - I've not been using it
much because of issues with Vuescan and a general lack of time, but I've
been putting it through its paces with the latest Vuescan. This seems to
have solved the calibration and bad banding problems I had with it. Hooray!

However....

I seem to be getting very bad flare on my scans (mostly Fuji Sensia) - where
highlights have a sort of halo bleeding into the darker areas surrounding
them. This is really noticeable with some macro pictures of flowers I have,
and general scenes shot in sunlight.

This flare/bloom is _not_ on the original slides. Looking closely at most of
my scans with this, it affects all my pictures - muddying the contrast. I
want nice clean contrasty scans that look like the original slides! It's
also very noticable on the edges of the frame - where colours bleed onto
the slide-mount.

I've done some googling, and the closest I have come to the problem is with
dust getting onto Nikon scanner mirrors - they say Minolta scanners don't
have this problem. For example this site:
http://www.vad1.com/photo/dirty-scanner/

The flare on my scanner isn't as severe as his example, but it's a similaaar
effect. One that looks like the type of scans produces by my 5400 is the
picture halfway down this post on photo.net:
http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=004EWS

Do others using the DSE 5400 have the same problem? Or is mine a lemon?
Anybody have experience of Minolta support in the UK? Because it looks like
the optics in my scanner need cleaning. Is it possible to do this at home?

I want decent high-res scans for posting to a photo library.

Cheers,

J.

 

[Hecate]

Hi,

I've got a Minolta DSE 5400 that's a few months old - I've not been using it
much because of issues with Vuescan and a general lack of time, but I've
been putting it through its paces with the latest Vuescan. This seems to
have solved the calibration and bad banding problems I had with it. Hooray!

However....

I seem to be getting very bad flare on my scans (mostly Fuji Sensia) - where
highlights have a sort of halo bleeding into the darker areas surrounding
them. This is really noticeable with some macro pictures of flowers I have,
and general scenes shot in sunlight.

Do you get the same results if you scan with the Minolta software?

 

[Matthew Dranchak]

Have you tried the Minolta Dimage software? I had very poor results with
Vuescan software and my Minolta DSE 5400 so I have resorted to the Minolta
software that came with the scanner. I am just now realizing the depth of
the Minolta software and really reading the instruction manual that I
downloaded from the Konica-Minolta internet site. Bought my DSE 5400 in July
2004.

 

[Jason]

Hecate wrote:

Do you get the same results if you scan with the Minolta software?

Well, I'm running a completely Linux based system ... but I'll see about
installing Windows on a spare hard disk later today, and putting it in my
laptop...

J.

 

[Jason]

Do you get the same results if you scan with the Minolta software?

Tested with the latest Dimage Scan utility downloaded today, with a fresh
install of WinXP, on my old but serviceable Dell C600. Only USB1.1, so it's
slow, but it works.

Short answer is yes. Same flare - light areas of the scan bleeding into the
dark. The default colour clipping of the Minolta software masks the effect
somewhat, but once the levels are adjusted. Also tried with Vuescan under
windows.

I have a slide of a dandelion on grass taken with my macro lens. Scans like
it's has a luminous yellow glow around it - not on the original slide.

I'll see what Minolta support has to say on Monday - the scanner is still
under warrantee. :-(

J.

 

[Eastside]

Other than the Scitex EverSmart I've never seen a CCD scanner that's good at
controlling flare that results from scanning high contrast film. Try this:
http://www.marginalsoftware.com/LS2000Notes/casestudyIII/FilmMasking.htm

or use a drum scanner.

Dane
....

 

[Hecate]

Tested with the latest Dimage Scan utility downloaded today, with a fresh
install of WinXP, on my old but serviceable Dell C600. Only USB1.1, so it's
slow, but it works.

Short answer is yes. Same flare - light areas of the scan bleeding into the
dark. The default colour clipping of the Minolta software masks the effect
somewhat, but once the levels are adjusted. Also tried with Vuescan under
windows.

I have a slide of a dandelion on grass taken with my macro lens. Scans like
it's has a luminous yellow glow around it - not on the original slide.

I'll see what Minolta support has to say on Monday - the scanner is still
under warrantee. :-(

What film are you using?

 

[Jason]

What film are you using?

Fuji Sensia II, ISO 100 - which I've used for a number of years.

I previously had a Canon FS4000 which was unsatisfactory in a number of
ways, mainly with the IR cleaning in Vuescan, and it was very slow, even
using SCSI. But I got some good scans and didn't noticed any significant
flaring.

Anyway, I've put an image up at:
http://www.cerevisia.demon.co.uk/dandelion.jpg

This is resampled to 1024 pixels jpeg from a full-frame 5400dpi 16bit scan
and matches the original slide density quite well - the greens are a bit
greener on the original, but it shows the flaring.

I take it that this is not normal?

J.

 

[Jason]

Other than the Scitex EverSmart I've never seen a CCD scanner that's good
at
controlling flare that results from scanning high contrast film. Try
this:
http://www.marginalsoftware.com/LS2000Notes/casestudyIII/FilmMasking.htm

Interesting site. But the examples on this site given for flare in the Nikon
8000/9000 suggests that you need to 'enhance' dramatically the image before
it becomes objectionable - even in the case of a hole cut in the film.

Not sure that I want to create a contrast mask for every slide I want to
scan, though. And a drum scanner is currently out of my budget.
Might try a Nikon LS5000 if I justify the expense.

J.

 

[Eastside]

Interesting site. But the examples on this site given for flare in the
Nikon
8000/9000 suggests that you need to 'enhance' dramatically the image
before
it becomes objectionable - even in the case of a hole cut in the film.

Believe me, once you start noticing flare in your scans even without
enhancement the slightest hint of it will stand out like a black eye. In
addition, adjacent areas of lesser contrast are also affected, but the
effect is much less noticeable.

Not sure that I want to create a contrast mask for every slide I want to
scan, though. And a drum scanner is currently out of my budget.
Might try a Nikon LS5000 if I justify the expense.

J.

Strictly speaking all we need is an opaque mask to completely block out the
offending area. A true contrast reduction mask is much harder to make, but
I agree with you there is considerable extra effort involved in using the
method described. I'm not certain that the LS-5000 would show much
improvement, but I'd really be interested. The LS-9000's ability to minimize
flare was an improvement over the '8000, but I can only say that for my
experience with them.

Good luck,
Dane

 

[Hecate]

Fuji Sensia II, ISO 100 - which I've used for a number of years.

OK. First thing is (and as Fuji film user who's going to probably
move on to Kodak) Fuji slide film seems to be notoriously harder to
correctly scan than Kodak (Ektachrome e.g.) I don't know if it's true
of something like Astia, because I've never used, but Velvia,
Provia/Sensia seem to be a bit of a problem.

I previously had a Canon FS4000 which was unsatisfactory in a number of
ways, mainly with the IR cleaning in Vuescan, and it was very slow, even
using SCSI. But I got some good scans and didn't noticed any significant
flaring.

Anyway, I've put an image up at:
http://www.cerevisia.demon.co.uk/dandelion.jpg

This is resampled to 1024 pixels jpeg from a full-frame 5400dpi 16bit scan
and matches the original slide density quite well - the greens are a bit
greener on the original, but it shows the flaring.

I take it that this is not normal?

You mean the halo/ Actually there's a number of reasons that could
happen. One is the film contrast. Another, and I *have* seen this
before, is a lens which has some mould in one end of the optics,
central to the lens axis. (How long have you had the lens and when
was it last serviced?). That's a couple of things. So, no, it probably
isn't waht you'd expect, but no, it doesn't necessarily mean it's the
scanner.

 

[Kennedy McEwen]

Fuji Sensia II, ISO 100 - which I've used for a number of years.

I previously had a Canon FS4000 which was unsatisfactory in a number of
ways, mainly with the IR cleaning in Vuescan, and it was very slow, even
using SCSI. But I got some good scans and didn't noticed any significant
flaring.

Anyway, I've put an image up at:
http://www.cerevisia.demon.co.uk/dandelion.jpg

This is resampled to 1024 pixels jpeg from a full-frame 5400dpi 16bit scan
and matches the original slide density quite well - the greens are a bit
greener on the original, but it shows the flaring.

I take it that this is not normal?

It isn't normal, but it isn't unusual either - its dirt, probably just
dust but possibly more, on the scanner optics.

Scanners must be kept scrupulously clean - especially so for high
performance scanners like the Minolta 5400. Always use a dust cover to
protect them when not in use and never allow anyone who smokes to even
enter the room where the scanner is based - the muck they produce
doesn't just line their lungs. Even with all of these precautions,
regular cleaning every couple of years is still necessary for a well
used scanner.

 

[beenthere]

My first encounter with this kind of problem was on a Polaroid SS4000.
Some suggested that this was due to dirt collected on the scanner
mirror. Following their instructions, I took off the cover but could not
reach the mirror to clean it. I did remove a piece of lint on the mirror
and the flare appeared to be reduced somewhat (placebo?).

Solving this kind of problem in PS is nearly impossible in many
situations. Selections of either the highlighted subject or the shadow
background would end up bleeding into each other. As a result, I had to
avoid scanning highly contrasty slides. I swore that this would be *the*
top priority in choosing my next scanner.

Using an offending image (Velvia 50), I tested a few brand new Nikon
Coolscans in stores and in users' homes. The flare showed up in all of
them, though to a less degree than the Polaroid, it is not what I can
accept. This kind of implies that there is more to it than just dirt on
the mirror. Here's a link on how to clean a Nikon 5000, for those who
are not faint of heart:
http://www.pearsonimaging.com/ls5000cleaning.html

I then tested the same image on two Minolta 5400s, one brand new and one
with quite a bit of dust collected inside. On both, the flare is
*significantly* less than found on the Nikons, and is what I can live
with. I bought a Minolta 5400, and have not been bothered by this
problem since.

All my tests are scanned the same way - raw and without any scanner sw
getting in the way.