Silverfast Studio AI 6.5.0r4 Multi-Exposure-Useless

[viermalfuenfinches]

hi, i was very happy to see the preview-results on the homepage and decided
to buy the upgrade.
i thought with ME its possible to get most from the slides.

i wait 35min. for one scan in 300%scaling and 300dpi MEx4 and ICE.
im working on a catalogue and have no time to waste.

im very angry about whats comin out.
on some contrast-edges i have this tiny lines in red, blue and green.
so, i will not and cant use this feature any more.

please make your software work fine and prepare a working update as
fast as you can.

im losing time and money with this peace of software.
thanks in advance.

[rnwhalley]

I too have experienced similar problems with Multiple Exposure. The problem seems to occur with high contrast slides when the ME setting is sample 4 times. Particular issues have been when scanning sunsets on Velvia 50 slide film. Sampling only x2 seems to improve results as does scanning from lower contrast slides. I have however stopped using this feature as I can't rely on it. A shame because its one of the two reasons why I upgrade from the default software that came with the Epson v700.

I have found the ICE setting too slow when set to High Quality but acceptable when set to Lite. With the Quality settings the software appears to identify dust as being detail on the scan. I now scan everything with just ICE Lite and the results are excellent.

Another problem I have found is that when using the multiple sampling setting I get blurred images. When I view slide through a Loup or sample the scan just once the images are pin sharp. When I make mulltiple samples the images appear to have "ghosting" as if the multiple scans don't quite line up. The ability to do multiple sampling of a scan was the other reason I upgraded and neither seems to work reliably.

The software is much better than the Epson software that shipped with my scanner but at present I can't see why I paid out to upgrade as I don't see either of the key features working properly. The SE Lite version that Epson shipped does everything else that I wanted.

Some feedback from Silverfast regarding these problems would be helpful.

[LSI_Heidorn]

Hello,

first of all, when experiencing those "color-seam" Problems i would strongly advise to use 2x MEX ( 4xMEX doesnt give a much higher boost in dynamic range, takes longer and is more prone to problems, we'll drop it soon for a different mode ).

If you get better results with 2xMEX, please stick to it until we replaced or dropped the 4xMEX mode.

I think you are a bit harsh by saying that MEX is useless, it does miracles on a lot of images for a lot of Users, but like any brandnew feature it has to evolve some more steps to satisfy everyone...

Greetings,


Nils Heidorn, R & D, LaserSoft Imaging AG

[viermalfuenfinches]

"I think you are a bit harsh by saying that MEX is useless, it does miracles on a lot of images for a lot of Users, but like any brandnew feature it has to evolve some more steps to satisfy everyone..."

i have to say that i want to work with that feature and dont want to be a beta-tester.

i updated to the newest version and its still not working well.
i got stairs-effect on lines and ice doesnt seem to work fine with 2xME.

greets

[LSI_Heidorn]

Dear all,

as we have many working test scans and many Users where MEX works just fine you seem to have specific Problems.

Either you use images that are hard to align or your scanners may have mechanical problems that produce drastically misaligned scans.

Could you tell me whether the "stair" artifacts occur on all of your images or only on a certain type ?
( rmwhalley already wrote that it only happens to high-contrast images which is strange since alignement should be easier on those ).

Referring to the mechanical problems:
Some Users reported everything working after the simply taped the slide(s) to be scanned in MEX mode because due to the vibrations of the sled movement the image simply moved after each scan...

In any case i'd ask you to report your bugs through our Bug reporting System at www.silverfast.com as we are not speaking about general issues here...

Greetings,


Nils Heidorn, R & D, LaserSoft Imaging AG

[viermalfuenfinches]

i must agree,
the stairs-effect only shows up on high-contrast parts of images.
where light meets shadow..... the part with a thin halation.

i will give the tip with tape a try, but i think its not the problem.
otherwise the whole image would be affected.

could it be that the slide arches during the scan ?

[rnwhalley]

Sorry for the delay in responding as I hadn't checked this posting for a while. I had already come up with the theory that the slide holder might be moving on the glass so repositioned my scanner onto a very solid surface and weighted the lid shut. This has improved performance greatly having rescanned a problem slide around 20 times I only had 2 problems.

The two problem scans are similar to before where a halo appears along the edge of a tree branch where a bright sky is the background. The problem is however much less obvious and I can remove it with a bit of Photoshop. A further point however is that not all the tree branches are affected which I would have expected if the slide had moved.

I will continue to monitor for problems but at the moment I am making headway.

Thanks for your continued development of the software (I love NegaFix by the way - great for scanning my Holga films)

[LSI_Heidorn]

Dear rnwhalley,

thanks for the "thumbs up" on that issue, we really were startled why some of you should have so much problems where others ( and we ) experienced good results in the overwhelming cases.

We are however also working in getting those "2 out of 20" cases all right that you mentioned.

As a general thing i will ask for our webmaster to add a few lines to the MEX article stating that you should really use holders or tape to fix your images !

Greetings,


Nils Heidorn, R & D, LaserSoft Imaging AG

[viermalfuenfinches]

maybe it helps to see some examples from my results with and without Multi-Exposure.

I scanned a middleformat slide 4,5 by 6 cm.
Both scans are HDR with ICE, scaled to 517,2% and 355dpi.
I taped the mask, so it cant move.

First cutout scanned with ME and the second without.
Why is it still impossible to get a perfect result ?

[LSI_Heidorn]

Dear viermalfuenfinches,

you're right, the image looks like an image that should be nearly perfect for the autoalignment routine, hmmm...

Did you ( or could you ) try to disable ICE and to report the result, still alignement problems ?!

If ICE on or off influences the result, that would be a much more narrow part to search for that problems.
Obviously we did test MEX with ICE but we have no reproduce to test all possible combinations of settings / motives you may use in the "real world".

You probably wrote it in another posting, but could you just (again) let me know wich Scanner it is that you use adnd what ( if any ) special filters ( usm ?) you used for that scan or if those settings doesn't make a change for the unsatisfying MEX results you get...


Greetings,


Nils Heidorn, R & D, LaserSoft Imaging AG

[viermalfuenfinches]

hi nils,

i use a Epson Perfection 4990 Photo.
all filters are off.
i scanned the same image with Multi-Exposure... ICE is off.

i choosed a smaller cutout.
take a look at the eyes of the skull, there are still red and green shadows
and the whole image looks kind of unsharp.



greets

[LSI_Heidorn]

Dear viermalfuenfinches,
that is strange, so turning off ICE definately changed something but the result is not perfect...

You have a personal Mail, i need some more contact data to feed to our support team to contact you for replicating the problem here.

Greetings,


Nils Heidorn, R & D, LaserSoft Imaging AG,