CCD flare with Nikon Coolscan 5000ED

[josip_loncaric]

Store exchanged my Nikon Super Coolscan 5000 ED for another brand new unit, which has exactly the same problem: flare in high contrast areas. My old Minolta at less than 1/3 the price doesn't have this problem, although its resolution is lower and it lacks some Nikon features. After two bad Nikon scanners, I must conclude that this is a design defect.<div></div>

[robert_k1]

Josip, flares in scans from an older scanner drove me nuts. After much research to replace it (see my earlier posts in this thread), I chose a Minolta 5400. For two years, I have not been bothered by flares in hundreds of Kodachrome and Fujichrome scans from the 5400.

[lex_harris]

"After two bad Nikon scanners, I must conclude that this is a design defect."

 

Yep, I agree totally.

 

Josip, at least your supplier did the right thing and exchanged your scanner, mine would neither replace it nor give a refund.

 

Currently I'm stuck with a $2000 scanner that is useless for scanning my Kodachromes. I'm still interested in trying Erik's suggestion of removing or coating the cover glass on the CCD. Does anyone know the exact device type number for the CCD array in a 5000ED? I'm thinking of just buying a spare CCD and getting the cover glass coated to see if this reduces the problem. I'm hoping the CCD will be mounted in a socket rather than soldered into place and will therefore be relatively easy to replace. Can anyone confirm this?

[robert_k1]

My post in the following thread has a link about disassembling and cleaning a Coolscan 5000ED. Disclaimer: I don't own a 5000ED and have never disassembled one. Good luck.

 

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

[robert_bohl]

The flare problem, especially with Kodachrome, has been around for a long time, unfortunately with no solution in sight. See some other postings at:

 

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00Bzm8

 

In my experience Nikon seems to be totally avoiding this problem, and I've about given up with trying to get a decent scan from my Kodachrome transparencies, which are about 80% of my collection. I've had no problem with my few color negatives however, probably due to their lower contrast.

 

As an experiment, I took my new Nikon D200 and copied some Kodachromes using my old slide-duping outfit (micro lens, diffused color head, bellows, etc.) The results (RAW) where generally superior to comparative scans with the Coolscan 4000, no flare, better color and range, less dirt -and it was a lot quicker! The only drawback lower resolution (although I do have the option to crop extraneous detail). Those of you with hi-res DSLRs may want to try this "scanning" method as a quick fix, especially if the final reproduction size is small.

 

Frankly, I don't understand how the flare problem has gone on so long without more s___ hitting the fan. I think some manufacturers have a lot to answer for!

[josip_loncaric]

My Nikon Super CoolScan 5000 ED returned from service. Nikon replaced the main PCB and the CCD board, adjusted the CCD board, did general check & clean, and even replaced the bottom cover. Post-service, flare is only marginally reduced, at the expense of increased sensor noise and loss of shadow detail. Nikon support people really tried to help, but the overall result is not that great: To match the original scan color balance without turning on ROC, I've had to apply significant amount of color correction, because the repaired scanner produces about 10-30 levels darker images.

 

Despite Nikon's help, I'm not sure that repair of this problem is even possible without scanner redesign.<div></div>

[lex_harris]

Josip,

 

Frankly I'm not surprised that even the best efforts of your local Nikon service people couldn't eliminate the problem. Gathering together all the evidence, I've concluded it's definitely either an inherent design defect or an optics quality issue, neither of which can easily be corrected by a service facility.

 

For me the solution will probably be either to go to a better quality professional grade scanner, or, as Robert Bohl has suggested, to a high resolution digital camera with slide duping attachment. Like Robert I've given up trying to get good scans from Kodachrome transparencies with a 5000ED. The scanner is simply not up to it. Each scan requires an enormous amount of time with PS to try and fix the flare defects the scanner introduces, and often the defects are so bad they simply cannot be repaired.

 

To follow up on the CCD cover glass issue, I recently stripped my scanner to find out what we are dealing with. The photo shows the CCD from my scanner. The cover glass does have a slightly matte appearance and reflects a little less light than a polished glass surface, but it reflects a lot more light than a properly coated optical surface. Also the reflected light is not coloured as is usually the case with anti-reflective coating. It seems to me that the glass does not have a proper anti-reflective coating and this would go a long way in explaining the flare I am seeing in my scans.

 

One final comment, it's interesting that no-one from Nikon has posted any comments. When I first approached Maxwell Optical about this problem, I suggested to the service manager that he simply view this thread to see the examples. His reply was along the lines of "I never look at forums, they're not worth looking at". Speaks volumes I think.<div></div>

[erik scanhancer]

Yeah, that's the bastard allright! Will you tell us if you have managed to remove the cover

glass? I'd like to see your pinhole test repeated. I bet you'll get the same improved results as

the guy in my sample.

[scott_brim]

Uhhh... Erik ... Might you be thinking of selling an after-market, upgraded version of this component? If you are, what might it cost? (It never hurts to ask. Really.)

[lex_harris]

"Will you tell us if you have managed to remove the cover glass? I'd like to see your pinhole test repeated. I bet you'll get the same improved results as the guy in my sample"

 

Erik, unfortunately I don't have the facilities to remove the glass cover. I'm not prepared to risk damaging the CCD unless I have a spare on hand (which I don't) plus the relevant service literature describing any setup/adjustments that need to be made if the the CCD is replaced (which I don't have either).

 

 

"Uhhh... Erik ... Might you be thinking of selling an after-market, upgraded version of this component?"

 

Looks like a nice little marketing opportunity here for someone... I for one would be interested in a coated or un-glassed retrofit if it can be shown to eliminate the flare problem. Coated would be preferrable I think, un-glassed seems a little risky in an un-purged environment.

 

Physically removing and replacing the CCD is no big deal, it's a little fiddly but not rocket science. I suspect that fitting a replacement CCD might require some electronic adjustment, but without service literature I can't say for sure. Mechanical alignment does not seem much of an issue, in my case just a matter of small X-Y adjustment to centre the image in the NikonScan preview window.

 

So, anyone interested in producing an aftermarket kit containing modified CCD and installation instructions?

[erik scanhancer]

Scott, I'd really like to offer CCD's with the glass removed, but that is way out of my league.

I'm a fulltime photographer/filmmaker, who has already invested too much time in the

shortcomings of film scanners. Perhaps this might be something for a retired person with

lots of spare time on his/her hands.

[scott_brim]

... Scott, I'd really like to offer CCD's with the glass removed, but that is way out of my league .....

 

It's way out of my league, too, but I do wonder what it would cost Nikon to have a specialized version of this CCD manufactured without the window --- or with a better quality of optical surface if the window has to stay there --- and what the difference in price would be to the consumer. $20? $40?

[geoff_graham3]

I have been using a coolscan 5000ED for 6 months now and have to deal with this problem

in about 50% of my slides. This makes it a true lemon, because there is no way you could

miss this flaw if you tested your product at all. The glory days for this brand are over, keep

your eye out for more junk from them as they farm out more parts and engineering to

china.

[lex_harris]

Geoff, I can only agree. Nikon products are off my shopping list too. Apart from the flare, the build quality of my 5000ED left a bit to be desired as well. Take a look at the sheared screw on the back, it was even marked by the factory with a red dot (that's how it came out of the box) but the unit was still shipped anyway. And the front plastic cover has two securing clips, top and bottom, but the bottom one had been broken off (presumably in the factory during assembly) and it was just glued on with some kind of superglue instead. The local distributor, Maxwell Optical, was not interested when I raised these issues, they refused to replace the scanner under warranty.<div></div>

[lex_harris]

Well, it's been a long time coming but here is a final postscript to the above discussion.

 

After struggling with the Nikon for far too long I finally decided to stop wasting my time and bought an Imacon

Flextight 646. The skeptics may be interested to learn that the awful artefacts illustrated above are completely

absent in my Imacon scans of the very same transparencies. Dust on mirrors? Bad scanning technique? Defects on

film image?...... I don't think so. The Imacon breezes through my toughest Kodachromes without any trouble at all.

Excellent edge to edge focus, superb shadow detail and no "flare" even with the highest contrast images.

 

It's not all bad news for the Nikon though, the 5000ED is pretty fast and makes rather a nice proofing tool if you have

a large number of 35mm slides to get through. Viewing a 4000dpi image on your screen is better than any light

table, you can very quickly assess image quality and decide if it's worth scanning on the Imacon.

 

I will be happy to post comparison scans if anyone is interested.

[janrsmit]

Lex,

 

I have a Nikon Coolscan LS50, suffering from the same flare problem as highlighted in this thread.

Following the advice of my fellow countryman Eric, i found a firm capable of removing the protective glass from CCD

sensors. After dissassembling the scanner i had the protective glass removed and applied two more techniques to eliminate

or reduce any cause of flare.

To cut a long story short, THE FLARE ISSUE IS COMPLETELY GONE.

The effect is quite similar to the differences as shown in Eric's contributions to this thread.

The images also have more snap, the details in darker parts are better visible, which is quite logical as the local or micro

contrast is improved.

 

I am working on how to get some images uploaded to this thread, sofar not successful, but i will get it done.

 

Jan

[lex_harris]

Jan,

 

Can you tell us the name of the firm that removed the cover glass? Was it expensive to do?

 

"and applied two more techniques to eliminate or reduce any cause of flare"

What were the other two techniques you used?

 

I agree, the cover glass is the culprit and is causing the flare/ghosting as well as loss of contrast. But I think without a sealed cover the CCD surface will be very vulnerable to contamination and damage. It would be desirable to fit a new cover made of high grade optical glass, coated on both sides. This could be glued into place, preferably at a slight angle to the plane of the sensor so that any reflections from the underside of the glass would fall off the sensor array. The whole operation would need to be done in a clean-room I think.

 

BTW the sensor in my Imacon 646 DOES have a cover glass. Since there is no evidence of flare or multiple reflections in any of the scans from the Imacon I can only conclude the cover glass must be of considerably better optical quality than that of the Nikon.

 

Lex

[eldon_k.]

I'm also dying to know where the sensor glass removal was done and for how much money. What are the other things you did besides

removing the sensor glass, and how much do you think those things contributed to the reduction in flare? I'd love as many details as

you're willing to provide.

 

I wonder if a place like this would work on a scanner sensor:

 

http://www.maxmax.com

[robert_k1]

Lex, thanks for the update. My Minolta 5400 is still going strong without any flares. Only those of us who experienced or noticed the flares know what a pain they can be.

[janrsmit]

Lex,

 

The modifications i applied are:

Removal of the protective window, this was done by a firm in The Netherlands as a gesture to me, not on a commercial basis.

A company that does this, on a commercial basis, is Eureca in Germany: www.eureca.de, the person i contacted was Juergen Beckers. The risk is there, they estimate it at a 1 out of 10 chance that the sensor gets somehow damaged or becomes unusable. Obviously they are on the safe side. Thy can also put some other glass, like coated glass etc on the sensor.

 

Blocking of any entry of light in the "optical block", this is the plastic housing that holds a mirror, a lens and the sensor. This is covered by a thin flexible plastic cover. The lens is a metal barrel kept in place in a rectangular tunnel in the optical block, leaving openings that can let light into the chamber between the lens and the sensor. I blocked these with a putty that stays flexible, so it can be easily removed if needed. The plastic cover is kept in place and firmly fit with some self-adhesive tape.

 

Covering of the chamber walls with a light absorbing material.I found a fabric with a very high light absorption. The light that is projected by the lens on the sensor is not all absorbed by the line-sensor, a portion gets reflected back into the chamber bounces around and at some point must fall on the sensor again, be it much less intense. The walls of the chmber are dark grey but not really absorbing light. The fabric used, a cotton velvet, reflects only about 0.47% of incident light, just looking at it really redefines black! If you put a piece of this stuff on top of anything else that you considered black, all of a sudden this becomes grayish in you eyes.

 

All in all some work to get it done, i reckon that 60-80% of the result is achieved with removing the protective glass.

 

The results are quite rewarding.

 

Jan R. Smit

[janrsmit]

Addition.

I am somehow unable to post pictures as well, this weekend i will work on getting a site up so pictures can be downloaded.

I also had one of my slides, from the front side of the Karlskirche in Vienna (a must visit whenever you are in Vienna) scanned at a professional labo, they used a flexscan Flextight II, an old model, and the result showed flare similar to the results i get from the coolscans ls50 and ls5000 and shown here in some examples in this thread.

Actually my current scanresult is better. Next time i visit this labo, they do all my film development, i will compare the results with them.

 

Alex, i would love to see results from your flextight 848, it should be way beter than the flextight II, do you have a site where i can donload from?

 

Jan R. Smit

IQ is technology, PQ is YOU

[lex_harris]

Jan,

 

Thanks for sharing your experience.

 

Below is a comparison of scans from my Nikon 5000ED and my Imacon Flextight 646. The source image is a reasonably well exposed KR 64 taken about 7 years ago with a Canon A-1 and FD 35-105mm zoom. This image contains some bright white rails in front of a deep shadow area which is a good test. The flare and ghosting show up really well in the Nikon scan but do not exist in the Imacon scan. The Imacon scan also has vastly better shadow detail

 

The Nikon scan was done in NikonScan v4.02 with Kodachrome setting, 16 bit, 4000dpi and 16x multi-sample. All other settings were neutral, ICE, ROC, GEM, DEE all off, analog gain 0.

 

The Imacon scan was done in Flexcolor v4.86, 16 bit, 6300dpi with my own Kodachrome profile settings which include adaptive light and shadow enhance plus colour correction.

 

Apart from cropping and applying auto levels in PS, no other adjustments have been made to either scan.

 

The images below are:

(a) the full size scans

(b) crops of the white rails in front of the shadow area

© the same two crops with 50% shadow adjust applied in PS (just to prove that there is nothing lurking in the shadows of the Imacon scan)

 

Lex<div></div>

[lex_harris]

OK, looks like I can only upload one image per post, sorry. Here's the Imacon whole image:<div></div>

[lex_harris]

The Nikon crop:<div></div>

[lex_harris]

The Imacon crop:<div></div>