Flare / Ghosting Problem with Minolta 5400

[edhebert]

I'm having a problem scanning a particular Velvia slide with my Minolta Elite

5400 (series I). The slide admittedly has a high dynamic range, but there's a

huge ghosting problem between a region of bright sky and a building that's in

relative shadow. The edge of the building is completely bled over and ghosted.

The original slide shows no trace of this problem.

 

Is there a scanning technique that you'd suggest for this image, or is there a

defect in my scanner?

 

Thanks

[mendel_leisk]

Could you post a crop? Maybe the grain dissolver would help, if you're not already using it.

[scott_brim]

Gee, I thought only the Nikons had this problem. In those scanners, it's the window that covers the CCD that causes the ghosting.

 

Try reorienting the slide 180 degrees, or even flipping it over and scanning it reversed. I've also done things like making two scans at different exposure levels, extracting what I can from the shadows using Photoshop shadow/detail, and then merging the two images using layers.

 

I've also discovered that Silverfast does a better job than NikonScan of retrieving shadow detail while reducing noise in the dense areas, and this has the side benefit of making it easier to deal with the Nikon's flaring problems.

 

Noise in dense areas and bleed over from high contrast zones are the two problems that prevent these desktop scanners from being decent performance competitors for the Imacons and the drum scanners.

[edhebert]

Actually, I think my problem might be more related to pushing a slightly underexposed chrome to the dmax limit of the scanner than any problem with the scanner itself. After my initial post, I tried rescanning the chrome twice, exposing once for sky and once for shadows. The more conventional exposure scan exposure showed no ghosting - the pushed exposure showed the ghosting. I merged the two in Photomatix, and was able to produce an acceptable image.

[mendel_leisk]

I'll see a little flare sometimes if I've got a mounted but mis-cut slide (in other words I'm scanning air), at the edge between the the gap in the film and and the slide frame. This is with the same scanner.

 

Yeah, sounds like something you just notice in extreme cases.

[jtk]

Velvia's a bad film for many purposes. Inaccurate color, tendency to big black areas that could be handled better by more professional films (virtually any other slide film).

 

I don't think Nikon has flare problems with well-exposed films, with the exception perhaps of Velvia, which can't well-exposed at extreme ranges. A goofy film. If one was photographing those scenes with B&W, one would expose and process N-1 or N-something. Happily Fuji offers several better films.

[edhebert]

John, thanks for your perfunctory conclusion about Velvia's "professionalism". How this aids the discussion, however, is a mystery to me. Funny, on the box, it doesn't say "less professional film". If I asked about the color accuracy of Velvia, you'd be spot on. If I had shot the image in Acros and developed in PMK, I'd have tons of latitude and wouldn't have this problem. But, then again, I wouldn't have achieved the hypersaturated, contrasted colorful result I had intended by pushing Velvia one stop to record this long exposure image to the point of reciprocity failure. I understand the subtleties of different emulsions, and exploit them to achieve the results I'm chasing. That said, the end result on my Velvia chrome is almost as I had intended. As my original question indicated, the problem lies in tranferring this image to my computer via a Minolta 5400. I was hoping someone that'd been down this road might offer some ideas. And I got some great suggestions from the forum, so thanks all.

[robert_k1]

Ed, it would help if you can provide the details of your workflow (i.e. scanner settings and PS conversions, etc.) and an example image.

 

Here are some summaries of what my experience with flares on Nikons and Minolta 5400:

 

- I tested a few Nikon scanners, old and new, and have found pronounced and offensive flares on all of them. I have not noticed any (or noticeable) flares on my Minolta 5400, after a few years.

 

- My above observations are based on accurately exposed chromes (i.e. highlight details not blown out), including Kodakchromes and Fujichromes. All scans are raw without scanner sw interventions.

 

- I have found scanning good shadow details are much more difficult on Velvia 50 and Kodakchrome 64 than on Provia or Astia.

 

Few (any other?) posters here have compared flares on both a Nikon and a Minolta. You may find these threads and links interesting:

 

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00IGyN

 

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=001A4q

 

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=004EWS

 

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00A2Sh

 

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00CTcF

 

http://www.vad1.com/photo/dirty-scanner/

 

http://www.pearsonimaging.com/ls5000cleaning.html

 

This member apparently was aware of the flares:

 

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00H558

 

Here's what he ended up with after buying a Nikon:

 

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00HCnM

[edhebert]

Robert-

 

Thanks for the detailed reply! Very helpful. I had seen some of these posts in my PN search, but hadn't yet seen the page about the dirty scanner issues. Interesting.

 

Anyway, to explain a bit about my workflow. I've calibrated the scanner using a 35mm IT8 target, and for the most part, slides scan accurately. This problem occurs scanning difficult slides in 16 bit using either VueScan or the DImage twain driver. The problem doesn't occur in all slides, just high contrast ones like the image I've included here. This was scanned 16 bit using the DImage Twain. The original chrome contains significant detail in both the tree bark and the sky. You'll note the CCD bloom in the tree trunk areas that border the bright sky. Again, this is an unretouched scan, and I'm able to bring back quite a bit of detail in post. Still, the scan doesn't come close to representing the original in this case.

 

Maybe the issue relates to dust/dirt in the scanner. Although the design of this unit is pretty robust and well enclosed. Any techniques for cleaning a 5400?<div></div>

[scott_brim]

There it is, a classic case of flare, even if it comes from a Minolta. As for my Nikon CS 5000, I've pretty much despaired of Nikon ever fixing the flaring problems in their desktop scanners. The market for these devices isn't what one would think it should be, probably because the learning curve of getting into digital film scanning and digital post editing is so steep. So there isn't much incentive to make evolutionary improvements or to fix what is an obvious design defect in the product.

[robert_k1]

Yup, that's what I'll call flares. Many of my slides have this kind of contrast, and the flares would show up on the Nikons. But I have never seen anything like this on my Minolta 5400 after hundreds of scans over three years. BTW, my 5400 sits in my rather dusty study without a cover. The Nikon problem may not be due to dust (alone).

 

I'm using Minolta's sw, and scan raw, i.e. with all corrections turned OFF. To scan raw, the preferences must also be set up correctly. Have you tried that?

 

If there are details in the tree trunks, I can't see any in the image. If the trunks are without details even after PS post processing, then correcting the flares in PS is simple. But if the details are there, then it is nearly impossible to get rid of the flares in PS (easily). This was the major headache I had with yet another vendor's scanner before I switched over to the 5400. I swore I would give up scanning if I could not find a solution to the flares. Lucky for me, I did my testing and found the 5400.