Jump to content

CCD flare with Nikon Coolscan 5000ED


Recommended Posts

<p>"I must say, I absolutely hate the Imacon."</p>

<p>I have several customers with expensive analytical instruments who criticise these instruments because they constantly have problems with them. These same customers never want to spend any money on maintenance and they refuse to buy a service contract. So the instruments are chronically neglected, they fail or perform badly as as result, and the customers blame the instrument.</p>

<p>It's not your fault I know, it's the lab manager who won't look after it ;-)</p>

Link to comment
Share on other sites

  • Replies 93
  • Created
  • Last Reply

Top Posters In This Topic

<p>"I'm not allowed to take it apart... *sigh*. "</p>

<p>This is an absurdly easy job.</p>

<p>Two screws to remove the main cover. 4 screws to remove inner cover. The lens is now staring you in the face. Clean the IR filter at the top end of the lens. Reverse the procedure. Cleaning a Coolscan mirror is advanced rocket science compared to this.</p><div>00S6nt-105131584.jpg.13130c5da196841f32b4d5745d7e29a2.jpg</div>

Link to comment
Share on other sites

<p>Yes let me rephrase what I said previously to: "I must say, I absolutely hate the Imacons <em>that I do not own</em> ".</p>

<blockquote>

<p>Cleaning a Coolscan mirror is advanced rocket science compared to this.</p>

</blockquote>

<p>Haha. Yet, like I've said, I've done it 3 times without incident :)</p>

<p>I wouldn't even know where to begin in proposing that I do it myself to the supervisor... by signing a contract that says 'if anything goes wrong, I'll pay for the full replacement?' :-/</p>

<p>Looking at how simple the design is, one wonders why this device costs $13k to $19k. I realize the lens is expensive, but, geez.</p>

<p>Any of you know how the Peltier element is cooled in these Imacons? Typically a heatsink + fan could be used... but I don't hear no fan. Which makes me question just how effective the Peltier element is and if its own heating up is causing the CCD more noise...</p>

<p>Actually, Lex, why is there an IR filter above the lens? What's its purpose?</p>

<p>Could be any of a number of things, but my hands are tied. Which is why I'm so frustrated.</p>

Link to comment
Share on other sites

<p>Rishi, heat from the peltier cooler appears to be dissipated into the CCD carriage via a heat transmission plate. Probably not a lot to dissipate, I believe the CCD is maintained at about 18C.</p>

<p>Re the IR filter, I guess maybe the light tube emits some IR which might otherwise affect the CCD in some way. Being a curious type I tried a few scans without the filter but couldn't see any difference in scan quality.</p><div>00S6sq-105157584.jpg.aee9995b8c5cc356c5c8c5ae8f835820.jpg</div>

Link to comment
Share on other sites

<p>Lex,</p>

<p>Just read your bit about using the black plastic with holes in it... "To try and understand the problem a little better, I scanned a piece of black plastic (containing several pinholes) mounted in a standard plastic slide frame."</p>

<p>Great idea!</p>

<p>Guess I could do that with some overxposed Velvia leader that had some holes poked in it, yes?</p>

<p>Will try that on my Nikon LS-4000... on which I hardly notice any flare.</p>

<p>Then I'll try it on the Imacon 848.</p>

<p>I'm really at a loss as to explain why most of the flare occurs in a direction normal to the CCD axis... as you yourself have mentioned above, this indicates that the flare occurs at a certain spot <em>before</em> the bright spot is lined up with the CCD! Which does point to bad optics. But why are the ghosts spaced a certain distance from feature, etc.... these are strange & interesting questions.</p>

<p>Lex, you took apart your scanner & accessed the CCD -- did you ever look into if it is possible to coat the glass yourself with an anti-reflective coating? If my local eye-glass shop can do it for my eyeglasses... can't be that hard to do it to the CCD cover glass, can it?</p>

<p>Rishi</p>

Link to comment
Share on other sites

<p>Actually, in that thread that is linked to above, a 'guru' (Kennedy) says:</p>

 

<blockquote>

<p>However, if it is just the presence of the CCD window that is causing the effect then you will see multiple images in your scan. If it is scattering from any optical surface in the scanner, including the CCD window, then you will see flared images.</p>

</blockquote>

<p>Since, with the Imacon, I'm not <strong>just</strong> seeing flared images but, rather, multiple images (look at the '1' replica to the right of '1' in the Imacon scan of the HCT target above), wouldn't it indicate that the ghosting is resulting from reflections perhaps near the CCD window? Not just from a dirty lens?</p>

Link to comment
Share on other sites

<p>Hmm, if only your lab guy would let you spend 10 minutes cleaning the scanner we could save ourself hours of speculation and discussion... :-/</p>

<p>Yes it looks to me like a combination of the two, flare and ghosting, but mostly flare. It does not seem normal for Imacon, I don't get anything like that. <br>

<br>

Coating just the outer surface of the window may help but I don't think it will eliminate it completely because you still have a reflective underside. Remember that the surface of the CCD device can reflect light as well, and light rays will have a lovely time bouncing backward and forward between the two. This explains offset ghost images I think- any light rays not perfectly normal with the plane of the sensor surface (and window) pass through the window at an angle and could hit the sensor surface at a point not actually on the active array then reflect back to the underside of the window and subsequently back to a point on the array itself. If you had really good collimation you could suppress such ghosting in the direction normal to scan movement I guess, but you would still have the possibility of the diverging rays from the lens resulting in transverse ghosting. </p>

Link to comment
Share on other sites

<blockquote>

<p>If you had really good collimation you could suppress such ghosting in the direction normal to scan movement I guess</p>

</blockquote>

<p>Funny, now that I actually think about it -- doesn't that mean that more harshly collimated light sources would give you <strong>less</strong> ghosting/flare, whereas diffused light would give you <strong>more</strong> ghosting/flare? I must be missing something here...</p>

Link to comment
Share on other sites

<p>Interestingly enough, my Nikon LS-4000 is now pretty damn good with regards to flare/ghosting. After cleaning the mirror and sealing that little gap (though I should probably unseal it again & see if it affects the scan to be scientific), take a look at this:</p>

<p><img src="http://staff.washington.edu/rjsanyal/ScanningWoes/LS4000_ScanhancerInSA-21_NoIR.jpg" alt="" width="800" /><br>

<a href="http://staff.washington.edu/rjsanyal/ScanningWoes/LS4000_ScanhancerInSA-21_NoIR.jpg">Link to Full-Size Image</a></p>

<p>Note the small amount of flare in the vertical direction, which is along the axis of the CCD. Is this known as <strong>CCD blooming</strong> , or is it mostly optical flare? This is a worst-case scenario, given that the white you see above is the <em>absence of film</em> . I just purposefully misfed the SA-21 feeder. So this is about as bad as it could ever get.</p>

<p>Note also the <strong>lack</strong> of horizontal flare or ghosting along the feeding axis (normal to the CCD). Lex, basically, contrary to your LS-5000 and my Imacon 848 results, there is not ghosting or flare <em>before/after </em> (temporaly) a sharp contrast boundary.</p>

<p>Actually the lack of flare in the horizontal direction but the presence of flare along the axis of the CCD could just be dirty/bad optics, it's just that the former doesn't happen because the light source is more of a line and so doesn't illuminate the regions before/after a sharp contrast boundary... in which case the diffuse/broad fluorescent lighting of the Imacon would actually hurt it in terms of flare/ghosting.<br>

<br /> Does that make any sense?</p>

Link to comment
Share on other sites

  • 5 weeks later...

<p>Hey Guys... yes, a little off topic again... but after chasing a few 5000 ed's i gave up and then chased a MINOLTA SCAN ELITE 5400 MK11<br>

and although paying too much for it. ( more than original price! ) I at least have something on it's way now.<br>

Pertty excited as it sounds like a good scanner to work with??</p>

Link to comment
Share on other sites

  • 1 month later...

<p>I almost gave up on the Nikons, but a friend let me use his 4000 ED - first scan was horrible, so I started searching the net. After reading this thread and many others, I opened the 4000 ED and cleaned the mirror (major grime layer!), added some black gaffer tape everywhere in the light path where it was possible, damped the chassis with some audio-related damping sheets, and set the scanner flat on a concrete slab before covering it with about 5 pounds of play sand in zip lock bags.<br>

The result was amazing - the scanner now clearly outperforms my old Polaroid Sprintscan 35 Plus (which it didn't before these steps). The mirror cleaning was by far the most effective step of this procedure, though.<br>

I was ready to go into the scanner and remove the CCD glas myself (buddy who owns it gave me green light - he hated the results he got with it), but I think I won't need to do that. Another mod I was contemplating is to upgrade the power supply with some top shelf audio capacitors and better diodes (cleans up power and reduces noise), but that would be rather radical and require an external power supply housing. Given I am not very impressed with the speed of the 4000 ED, I may just go for a 5000 ED after all and then start over with all these mods on that unit... </p>

 

Link to comment
Share on other sites

  • 2 weeks later...

<p>I am interested in purchasing the 5000ED to scanner to scan Kodachrome slides, and like everyone else in this forum, I'm concerned about the flare apparently caused by the cover glass over the scanner's CCD sensor, as well as from potential light leaks. I originally had the Minolta 5400II, but it died on me. Minolta (Sony) could not repair it but offered to buy it back for about $225. I took them up on the offer.</p>

<p>I spoke to Precision Camera in Enfield CT (They are a Nikon factory authorized service center), and asked them if they would remove the cover glass from a 5000ED scanner. They said that they are not able to since it would change the unit's original specification, since the cover glass acted as anti aliasing spatial low pass filter. I am waiting to get a quote from Eureka in Germany.</p>

<p>I have a question for the users who have removed the cover glass from their scanner. Have you seen any evidence of aliasing, even though the flare is gone? Norman Koren has a good tutorial on aliasing on his website, an provides some examples. However I am curious to know if this is a significant problem in scanning real world slides compared to resolution targets. http://www.normankoren.com/Tutorials/MTF2.html</p>

<p>If aliasing is a problem, maybe an AR coated version of the cover glass would solve the problem (I think it was mentioned earlier). I suspect that Nikon has included the cover glass refractive index and thickness in the optimization of their lens design, so I wouldn't aberrations are a problem. On the other hand judging by some of the the images posted without the cover glass, the aberrations don't seem to get worse.</p>

<p>Sincerely</p>

<p>Don</p>

Link to comment
Share on other sites

<blockquote>

<p><a href="../photodb/user?user_id=1340662">Don Perrault</a> <a href="../member-status-icons"></a>, Apr 22, 2009; 08:08 a.m.</p>

<p>...since the cover glass acted as anti aliasing spatial low pass filter.</p>

</blockquote>

<p>Well, maybe, maybe not. I don't have a spec sheet for the sensor in the 5000ED, but from memory it's around an inch or so wide. Let's say an inch, 4000 pixels, = 157pixels/mm. So the Nyquist frequency is 157/2 or ~80 cycles/mm. Above that aliasing is a risk, but few films would have significant MTF response beyond 80 cycles/mm. (The 5000ED scans the frame along the long axis, so the film width is also ~1" i.e. ~1:1 relationship).</p>

<p>Personally I would be very reluctant to permanently remove the cover glass. The risk of sensor damage due to dust, humidity, mechanical damage etc etc would be just too high. Apart from any possible filtering function, the glass is there for a reason - to protect the sensor.</p>

<p>If you did get Eureka to remove it, I would suggest looking into an AR coated replacement. Otherwise I would not expect the sensor to last very long outside a clean room. Imacon scanners use a Kodak KLI-8023 sensor. It has a flat, clear cover glass with AR coating on BOTH sides. This is the spec for the cover glass:</p>

<p><img src="http://www.lexharris.net/documents/KLI-8023_cover_glass.jpg" alt="" width="560" height="367" /><br>

I don't think a normal AR coating will turn the glass into an anti-alias filter. Such a filter would need to be some special material that spreads each optical point into a cluster points. Maybe there are special coatings that do that in addition to being AR, I'm not sure.</p>

Link to comment
Share on other sites

  • 6 years later...

I have a Nikon 5000 Coolscan. All the discussion about the flare being caused by dust, or the camera in the optic block...I

still had flare after cleaning mirrors, lenses etc...I finally opened my Nikon, desoldered the sensor (Sony ilx142a) and

ground off the glass using 220 carbide, on a glass sheet (wet sand slurry)in my kitchen sink. You have to be careful as

soon as the glass begins to cut through as fragments will cut the gold trace wires. I switched to 220 water proof paper and

switched to width wise strokes. Dip the sensor in water (face down) and tilt to allow glass bits to fall out. Next, carefully

pot the wires with epoxy, allowing it to drain onto the wires. Minimize amount to prevent running out into sensor surface.

Prop on end to let epoxy set without running out in surface of sensor. After one end is done, do the other end. I used a

10x loupe to check progress. Finally, clean the surface of sensor by flooding with detergent, water jet and then wiping

with a wet foam makeup applicator. The sensor itself is surprisingly robust. I soldered machine pins in place so I could

easily plug the chip in and out as needed. The performance of the scan is amazingly improved. No flare at all.

Pinhole tests had shown multiple reflections above and below main image of each pinhole. After deglassing, totally

eliminated artifacts. Now I can take advantage of the 16 bit AD and boost shadows without leakage of highlights.

Link to comment
Share on other sites

  • 11 months later...

The flare is caused by reflections from the protective glass cover on the sensor chip. I removed the glass from mine and

all the flare issues are gone. Prior to this, point sources (lights a night) had a single primary ghost and several additional

fainter ghost images. I desoldered the sensor, installed a plug mount. I ground the glass off using a loose silicon carbide

water slurry on a piece of glass in my kitchen sink. I found extra OEM replacement Sony scanner chips (used in this

Nikon model) just in case.

Link to comment
Share on other sites

  • 5 months later...

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...