Review: Nikon Super Coolscan 5000 ED

[stephanie_mckeon]

all very good - but what about black and white, for arguments sake a film that

can take ice like portra. Is that better - are the ski tones smoother? I have

found no to be the answer. WIth the cs4000. the only plus point is using gem

to reduce grain size. Multi sampling x 16 has shown no improvement over x2.

Is THIS improved with the cs5000. otherwise I shall just commit everything to

a high quality CD from my lab- despite its limitations, the scans made while

developing the flim are sharp and smooth, and now dedicated software can

pixelate thme upwards.

 

Steph

[michael_brousseau]

Steph, unfortunately Digital ICE cannot be used with monochrome film. That�s one of

several side notes in the Nikon manual. Another problem area is ROC, where they

warn about the possibility of unwanted color being added if you try to apply this filter.

DEE has no effect in monotone shadows or highlights if the details have been totally

obliterated, it will just leave them a uniform color/tone. When I was researching an

alternate software program for my old LS-2000, I found some interesting information

at the VueScan site, some of which was applicable to B&W. I�m not a B&W shooter, but

suspect it may be more a function of software rather than hardware. I say that for two

reasons; first because in my own research on the net I�ve come across information on

B&W scanning (sorry I can�t be more specific, it�s not what I shoot), and second if your

getting your work processed by an outside lab they�re using some type of software to

generate your B&W scan. I would think with some research you might be able to find a

more acceptable working solution. Multi-sampling should not be an issue with B&W,

depending on your intended output size it should give a better result than you would

be able to achieve with interpolation. Hopes this helps.

[stephanie_mckeon]

Michael - I do not use traditional black and white films - only ones that can be

developed via C41 process, usually portra 400bw. Ice works absolutely fine

with these films and that is not my issue. I originally upgraded to the Ncs

4000 from the IV because the mult sampling feature is supposed to reduce

noise in the scans. However I have not found this to be the case at all, and

stopped using the multi sampling as it took so long to perform a scan, and in

many cases caused my very powerful MAC to close down photoshop. I am

interested to know whether black and white C41 film scans "smoother" with

the 5000. I always scan in colour and desaturate afterwards so funny colours

have never been something I have paid heed to.

[michael_brousseau]

Steph, sorry I have very little working knowledge of Portra; I�ve only tested it one time and have never attempted a scan of the negatives. It�s good then that you have the ability to take advantage of ICE. If you are working primarily with Portra, or for that matter any other print film, I would recommend that you try operating your LS-4000 using VueScan software, before making any decision on the LS-5000. VueScan allows you to download a free demo version prior to purchase, which would give you the opportunity to make a direct comparison to the Nikon Scan results. One of the limiting factors to producing good scans is software. When it comes to negative film, Nikon Scan gives the operator only two selections, "Negative Color" or "Negative Mono", were as VueScan gives the operator about two hundred options for negative film, including a number of different ones just for Portra. When I was forced to have to change my software from Nikon Scan to VueScan in order to be able to operate my LS-2000 on my new G5 MAC it turned out to be the best thing that could of happened. Everything about the scans was greatly improved by comparison to what I was getting when I was using the Nikon Scan software. When you get to the performance level of these two machines, the quality of an image has less to do with the hardware.

 

Another variable when trying to achieve goods scans is dynamic range. The greater the dynamic range the greater the ability of a scanner to pullout the shadow detail. The LS-4000 is already capable of producing 4000 dpi scans, which is the same maximum dpi one will achieve with the LS-5000. You may want to compare the difference in dynamic range between the two machines to see what the gain would be. They are also advertising improvements to the optics, which would merit some comparison.

 

I've experienced to same problems with PhotoShop either closing down or freezing after scanning was completed in Nikon Scan, but prior to the file being saved into PhotoShop. It's happened both when I tried running Nikon Scan as a stand-alone program and as part of PhotoShop. I first went to Nikon to see that I had the most recent version of Nikon Scan 4 (v 4.0.1). Afterwards I performed a complete uninstall of the software, and then reinstalled the program. I reset my preferences with some minor changes, and all seems ok so far. The most significant change I made to the preferences has been to set "File Saving", "Use Alternate Creator Code" where I selected PhotoShop under "Choose Application". This is only available with MAC. My thinking was that when you select Nikon Scan through PhotoShop the file saving defaults to saving in PhotoShop, so it makes some sense to me that if you run Nikon Scan as a stand-alone program it should save to the same application, not the default, which is Nikon Scan 4. I've tested the program, using multi-sample both with PhotoShop and as a stand-alone and it's function properly. If the problem persists I'm sure I'll be one of those individuals looking to have Nikon Tech Support sort it out, but for now all is well.

 

I totally understand your frustration; when software is not doing what it's designed to do it can be very annoying. Assuming the Nikon Scan software issue can be resolved, multi-sample is truly the only way to increase quality at the highest resolution. The process is very straightforward and involves making multiple passes over the image, which are then averaged. The results reduce the effects of electronic noise, producing more accurate reproductions with smoother changes in tone. If I understand you correctly these are the things that are most important to your workflow. I not aware of any situation where multi-sample works with one image but not with another; multi-sample is purely a function of mathematics. It's not unique to Nikon, all manufacturers of high-end scanners feature multi-sampling in one form or another. If I read your comments correctly it's what influenced you to go to the LS-4000. I would reserve any comparison of the effects of sampling to an actual side-by-side printout of an image. Good luck, hope this is of some value.

 

[jos_roost]

Stephanie,<br>

your last two posts leave me confused: what is it that you want?<br>

Porta is quite a grainy film. If you then ask "are the ski(n?) tones smoother",

'using GEM' should be the answer. However, GEM does not "reduce grain size" as

you state, it just smoothes the image in a smart way so that the grain structure

is less visible. If you don't want to see the grain, you e.g. have to scan at a

lower resolution. (from hear-say: the 5400 dpi from Minolta produces a 'prettier'

grain structure)<br>

Multi-sampling is NOT the solution for smoother tones and/or less grain.

It is a tool against noise, meaning against scanner CCD noise: this occurs only

in very dark parts of your image, where the scanner, being at the limits of its

sensitivity, has to decide between "I don't see anything here" or "I can just

see something here". A wrong decision will result in a noise pixel.

Multi-sampling means performing such a decision multiple times so that a better

'final decision' can be obtained.

[stephanie_mckeon]

Michael - thank you for your input - ver y helpful. I will try vuescan.

 

Jos - it is not my experience that portra black & white pro film is grainy. When

developed and printed in the darkroom it is very smooth and has great tonal

range which makes it ideal for portraiture , especially in context of the

lighting situations I need to use for my clients (natural, often low light). - are

you sure we are talking about the same film? Are you referring to the colour

films? These I have never used. Thank you for clarifying the matter of noise -

I must say I have found it quite confusing as I do associate, it seems

incorrectly, noise with grain.

 

I have tried, very briefly, the minolta 5400 and sent it back. It does not work

correctly with the latest OSX system, and although I was able to scan, the

scans were consistantly out of focus - no doubt this is a MAC issue.

 

On a different note, bearing in mind that the scanner can freeze/crash when

performing large scans, how can it scan a entire roll of film without doing the

same? Steph

[jos_roost]

Stephanie,<br>

I have to confess that I only tried Portra color film some time ago, and then

decided that I liked Fuji NPH400 better.<br>

However, you stated that you used GEM to counter grain in the Coolscan 4000 scans.<br>

You also mentioned that "<i>the scans made while developing the film

are sharp and smooth, and now dedicated software can pixelate them upwards."</i>.<br>

This still makes me think you have a grain problem, because you consider the

lower-res (I assume) "scans while developing" smoother than your 4000ppi

Coolscan4000 scans, which may be because of these lower-res scans-while-developing

do not resolve the grain.<br>

Furthermore, it is often said that the LED lightsource of the Nikons can result

in more pronounced grain, whereas the more diffuse cold cathode lightsource of

e.g. Minolta or Canon may give a softer (smoother?) grain image.<br>

Finally, about "pixelate them upwards": I don't think that is a good idea.

If you somehow obtain e.g. a 1000x1500 pixel scan, of course with e.g.

PhotoShop you can up-sample it to 4000x6000 pixels. However, the actual image

information of this large image will be not more than that of the original

low-res scan. If you want to retrieve the maximum real image information from your

negatives, you have to scan them at a high resolution directly.

[stephanie_mckeon]

Jos - I must admit I do find the digital arena confusing, especially from the

point of view - what words do I use to describe what I mean?

 

My lab supplies high quality scans of an enitre roll while developing it on cd

for me when I ask - these go to about 240 dpi at 10x8 inches. I really do not

know how to describe the resolution in any other terms. I use extensis

Smartscale to increase the resolution when needed. This software is

outstanding and, more importantly, EASY to use. I have noted that my labs

scans are "smoother, yet much sharper' than the best from my nikon cooscan

4000. The nikon has bags of info, and too much grain in the skin tones (or is it

noise? Or grit? Or is it simply info?) Whatever it is, it gives a definite digital

look to my film images. Having said that, I am a TOTAL perfectionist and as

long as I keep my mouth shut, none of my clients notice. But I do.

 

Recently I have used my lab more and more, but my last job's scans from

them were way off (I used the new neopan cn400 which has a much higher

contrast than portra and as nice as it is, it needs careful scanning). This is

why I am participating in this forum. really it seems I need to control every

step of my workflow.

 

A digital camera is on the cards, but I am sure I will continue with film for a

long time. Having to focus fast, meter fast in B&W, at fast speeds, while

running after two year olds has a lot to do with it.

 

I appreciate your time, being self-taught and very busy, it is invaluable to me

(www.stephaniedelengphotos.co.uk). You know, I called up Nikon

Professional of which I am a member and asked if I could send them a neg to

scan for me - just to see what the scan was like ...... and I am still waiting four

weeks later. Thanks again Steph

[michael_brousseau]

Steph: My complements on your web site; you have some nice work on display. If one-steps back slightly from the detail of the above posts, the majority of issues being discussed are software issues. With the proper workflow (scan, post-process, and print), there is no reason for your images should take on a digital look (I assume you mean pixilated). Your scanner however should not be thought of as a mini photo processor, where you scan, color/tone correct, then print. With your quality expectations, I making the assumption that you finalize your work using PhotoShop, and that you�ve taken your printer, paper, and ink into consideration.

If you�re getting a freeze/crash on a single scan multi-sample rest assured this is a software issue. As a MAC owner/operator, Nikon Scan 4 has given me the same headache. As I explained above, after performing uninstall then a reinstall, with preference adjustments, all seems ok to date. If your issue is freeze/crashes when scanning a roll of images then it is possible you�re running out of RAM. If you have plenty of available RAM, then it�s possible that the program allocation would need to be increased. You�ll need to evaluate the particular circumstance of your situation to better pinpoint the problem.

The out-of-focus issue you experienced with the Minolta 5400 is the same as I�ve experienced with the LS-5000. I have zero knowledge of the workings of the 5400, so I�m have no idea on that machine how the focus issue would be resolved. On the LS-5000, Nikon Scan 4 includes a �Calibrate� feature in their software, and for that situation recalibration has been effective.

I have to agree with Jos regarding his comments on using interpolation to increase your file size. Even with programs designed specifically for that task, it should be a measure of last resort. With digital being subtractive, you�ll get a better result sizing-down a large file as opposed to sizing-up a smaller file. If you stretch the envelope beyond its capacity you will see pixels in your final printed image. Your newer 6 & 8 color printers are looking for 360 dpi, and will interpolate whatever you send them to that requirement.

I standby my original statement on multi-sampling; �The results reduce the effects of electronic noise, producing more accurate reproductions with smoother changes in tone�. True, in the shadows this means a reduction in CCD noise. Tonal transition however occurs throughout an image, whether it is in the highlights or shadows. If grain is an issue GEM is one way to address it. Post-processing in PhotoShop is another way. PhotoShop is widely recognized as the standard for professional photographers working in digital. If you�re looking to sharpen your skills in the area of PhotoShop, or for that matter digital imaging in general, I�d recommend the UK author Martin Evening. His book on PhotoShop for Photographers is well worth considering.

[michael_brousseau]

There seems to be a stability issue with the Nikon Scan software that operates the LS-5000 when used on the Mac OS X platform. Since installing it on my G5, I�ve experienced a number of freezes on completion of the scan. After the scan operation is finalized, the program fails to convert over to the file saving application, such as PhotoShop. It will just sit there idle, doing nothing until you have to perform a force quit in order to exit the program. When I did the initial install, all ran well for a few scans then the problem surfaced. I thought I had solved this with a reinstall, but it appears that I was overly optimistic. I�ve now tried, more than once, to uninstall then reinstall with the same results. After doing a bit of net research, it appears I�m not the only one with the same issue, though it hard to tell how wide spread it may be. I�ve reported the problem to Nikon Tech Support, and am waiting their reply. When things are running right the scanner does a beautiful job. The work-around for me would normally be to use VueScan as the operating software, but at this time there�s an issue with multi-sampling that I�m waiting for them to iron out.

[paul_lines]

I am considering buying either the 5000ED or the LS-50 (V). Are there huge differences between the two? One is twice the price of the other and I need to know whether the 5000 is worth the price tag.

[nick_dantona]

First I'd like thank all those who have contributed so much information and knowledge to this forum. I have read the thread with great interest and it has resulted in my purchae of the 5000 ED. After several days of scanning I have found bi-polar results. That is, when scanning Ektachrome images from 20-30 years ago the scanner produces excellent results without much, if any, manipulation. The same cannot be said of the Kodachrome images from the same period. I'm referring primarily to Kodachrome 25, 64 and 200. The scans of these films produce images with highly compressed shadow and midtone values. Additionally, they are deficient in chroma and carry a blue-ish patina. With over 10 years experience in color correcting motion picture film and telecine-to-video transfer experience, I'm rather adept at this and have to work very hard at salvaging a workable image. As I have not picked up much complaint about this in the forum, I suspect I may be doing something fundamentally wrong. I have experimented with ICE, DEE, SIE, GEM and the like, but nothing seems to compensate for the extreme contrast delivered into these images. I have to push the curves, histogram and analog controls to the max to dig out any detail in the mid to low range. Has anyone else experienced this? Is it the film stock or am I nuts?

[edbustya]

Not sure if this will help, but I was getting strange results with Kodachrome until I discovered that there is a selection for Kodachrome in the "film type" drop down menu in Nikon Scan. This is the first film scanner I've owned that had such a choice. It made a big difference for my scans. I'm very happy with the scanner and with Digital ICE. Digital ICE doesn't get everything, but it saves me a lot of hand dusting in Photoshop.

 

The rest of the ICE4 features: Digital ROC, Digital GEM, and DEE, I find do not work as well as their equivalent Kodak ASF plugins, and it is so tedious to have to do a prescan to see the results of every setting change (correct me if I'm wrong, but I don't think these features are tied to the hardware, so it seems like Nikon Scan could show the effects of the setting changes without doing a prescan every time). I've found that the scanner software's Digital ROC doesn't work very well - I get odd colors (off my 40-year old faded-to-orange Ektachromes). The Kodak Digital ROC Pro plugin for Photoshop can be fine-tuned and gives better results. The scanner's Digital GEM seems to make the grain and noise worse on my 40-year old Ektachromes. But I haven't found any other miracle cure for grain and noise. The scanner's DEE is much more difficult to use than Kodak's Digital SHO plugin.

 

This is a great scanner. I'd use the automated image enhancing ROC, GEM, and DEE for a high-volume batch job, like converting a carousel of slides to a low-res DVD slideshow. But for quality prints from slides, I'd skip the enhancements and use Photoshop to get it right.

[nick_dantona]

Thanks, Ed. I have tried the Kodachrome selection and found that it made little improvement. I should point out that although the chromes I refer to are 20-30 years old, they are in excellent condition. One other thing that I don't believe I mentioned is: during installation I chose ADOBE RGB (1998), instead of the recommended sRGB, so I would be consistent with with the my color space environment in PhotoShop. I like your tip on the Kodak Plugins, but I'm going to continue to try and work this out. I am undaunted and not discouraged. I've only been working with this unit a few days. As I mentioned, of the slides that do scan well, the results are very impressive. I'm just baffled as to why such a distiction between Ektachrome and Kodachrome. To the naked eye, the characteristics of the original images are very similar. I'm still up for any other suggesions.

[Stock-Photos]

For Nick regarding scanning Kodachrome, from my <A HREF="http://www.saugus.net/Photos/scanning.shtml">extensive

experience</A> with the Coolscan 4000, Kodachromes, especially

under-exposed Kodachromes, always reveal less detail in the shadows

than does Ektachrome. Just a denser emulsion in my opinion.

Exposure latitude seems narrower with Kodachrome. Although projected

(slightly dark) Kodacromes look acceptable, scanned Kodachrome

seems to lack detail in dark areas compared to other film types. Also, as you probably know, Kodachromes will have a blue cast when flash is used.

[nick_dantona]

Thanks, Jim. Two lovely reproductions on your web site. Some helpful information and links (love the monitor/gamma info by AIM)as well. Looks like some challenging work ahead on the old Kodachromes. Any suggestions on what film emulsions for outdoor/landscape photography will scan best on the Nikon 5000?

[gord_cameron]

Carl (or others), I am very interested in the comments in this thread about scanning Kodachrome film, because I am in the market for a new scanner and the vast majority of my pre-digital work, which I intend to scan for archival purposes, has been on Kodachrome.

 

A few comments in this thread appear to criticize the quality of Kodachrome scans per se (on the 5000 or other scanners), while other comments are directed at problems when the scan is run through ICE processing.

 

The Nikon literature only mentions a problem with Kodachrome when ICE is used, but if there is a general problem with scanning Kodachrome, I would like to know, because if so it would make sense for me to wait until the technology advances to the point where this problem is overcome.

 

Another (expensive) option is to go for the Nikon 9000ED, which is described in Nikon's literature as being able to handle Kodachrome with ICE. This appears to be possible because of a combination of a new ICE program exlusive to Nikon ("ICE Professional") and a "rod dispersed" light source, i.e. where the LEDs are directed into both ends of an optical rod that disperes the light along its length.

 

The 9000ED would have the advantage of being able to scan 5 slides or two strips of negatives at one shot, but I don't really need that convenience or the medium format capability. So it would be for the Kodachrome capability that I would go to the 9000ED. Any comments?

[nick_dantona]

Gord: As I have mentioned, scanning Kodachrome has resulted in compressed, slightly tinted images whether or not ICE is used. I am becoming more certain that it is not a function of improper set-up, but rather the typical response of this emulsion. That said, I color correct the image "flat" (pre-final scan)and make adjustments in Photoshop with accetable results. The scan quality itself is excellent as far as resolution goes. If timing is not an issue, and the bulk of your slide library is Kodachrome, I suggest that you wait to hear from others on this forum and elsewhere. My impression is that either I'm the only one experiencing such severity in my scans or I used Kodachrome a lot more frequently than those in this forum.

Nick

[mike_nunan]

Hi Nick and Gord,

 

I used to have an LS-30 and experienced the problems of scanning Kodachrome. I now have a Minolta 5400 does a superb job of these, and I believe the difference is all to do with the light source. My theory is that the "spiky" spectral distribution of the LED emissions is an unfortunate match for the absorbtion characteristics of the Kodachrome dyes.

[mitchell_gusat1]

A while ago Enrico posted here two factual observations about flare and 'chromes'.

 

1- "CCD leakage": the CCD cells near a bright area tends to record more light than expected, so you may have something like a "flare". This flare is noticeable only when a very bright area is close to a deep dark area. The Minolta also is affected by this problem, but less (about 50%). 2- Kodachrome: the 5000 works not very well with this film, but the results are surely acceptable. Minolta is better because it records more color informations with less posterization.

 

While (2) is an old issue of Nikon scanners, attributable to their light source and optical system, my take on (1) is that speed hurts quality. Increased 'leakage' could be the consequence of using two sensors in parallel within 18mm of each other. I speculate that such design is intrinsically exposed to

 

a) optical interference

b) electronic crosstalk

c) elevated noise floor.

 

This theory could probably be tested with a computer generated film pattern that uses the two sensors in various phase combinations, from synchronous to anti-phase (one half clear, the other black, read raw data in VueScan); then the result compared against the V, 4000ED and Minolta 5400. (I'd wish the reputable review sites wouldn't have abandoned scanner testing...)

 

The high speed of 5000ED's stepper doesn't help either the quality; some sites report the new model noisier (more vibrations) than the 4000ED. See www.medienarchiv.com/Service/Fotografie/NikonCoolscan5000ED.htm , use Google tools for translation - the Fazit contains educative conclusions, e.g., the 5000ED is not worth the upgrade from 4000ED. Neither was the DOF and peripheral focus improved.

 

However, probably I'll end-up buying the 5000ED for its productivity, and accept its inherent compromises. Ultimate quality is mutually exclusive with speed, which was the design point of this unit; otherwise I'd go for the Minolta 5400 or wait for another rev. It's just too bad that despite its pricing policy and reputation, Nikon can't deliver yet the option to slow down the scan, de-activate one sensor and perform 'ultimate' scans.

 

Mitch

[iqbal]

I am novice in digital photography area. I need your help in answering some key questions:

1)In terms of quality print out 8x10, will it be wise to buy DSLR (say canon D10) OR continue with the traditional 35m. film SLR plus buy a Film Scanner? I already have 1000 of B&W negatives.

2) In terms of quality, which film scanner is better: Nikon super coolscan 5000 or Minolta Dimage 5400? I know Nikon is more expensive.

3)For B&W negative films, will ICE and some other features make a difference to improve the quality?

Thanks.

<Falsafay>

[duncan_smith1]

Thanks to all for some really insightful comments. I shoot 75% of my work in monochrome (Ilford Delta - ISO 100, 400 and 400 uprated to 1600), with the remainder on Velvia 50. For me, optimum image clarity and detail is of paramount importance, whereas a rapid scanning time is a nice-to-have.

 

Whilst I could feasibly expect improved scan results by applying, say, a 4x multi-scan to the Ilford negs, my understanding is that the Dynamic Exposure Extender software will not render improved shadow detail in monochrome images.

 

From previously-posted comments, I am not sure whether using VueScan rather than NikonScan will get around this - so is there an authority on the B&W scan subject out there?!

 

I will happily invest in the extra capabilities of the 5000ED but I need to be convinced that I'll get a discernable improvement in B&W image scan, as opposed to simply running a CoolScan V and VueScan software.

[nick_dantona]

Thanks, Mike. Your observation on light source mismatch for the Kodachrome dyes makes sense. As the bulk of my photo library is Kodachrome, I may consider your suggestion of the Minolta 5400. I first have to figure out what to do with my Nikon 5000 ED!

[simone_degan]

These days I have worked with the Nikon 5000ED and the Minolta 5400. I am not an expert and these are my thoughts.

 

Both scanners are very well built.

Nikon software gave me no problems for the installation, whereas with Minolta I had to reinstall it three times before being operative (scanner not connected, was the error message).

Nikon showed to have some focusing problems with almost all the pictures scanned. They were no sharp. I had to use the unsharp mask in PS to obtain good results. No problem with Minolta. Perhaps the Nikon unit I have used has got a problem, but it was brand new.

Nikon has a better tonal range, but at about 500% magnification the pixels are more visible than in the Minolta scans. I think this may be referred as the posterization effect as Enrico previously stated for the Nikon.

Nikon picks up slightly more details in the shadows.

Nikon is surely faster than Minolta.

 

In conclusion I think that Nikon is an excellent scanner but suits more the professional lab (not my case) for tons of slides to be scanned, with a balanced ratio of quality/speed. Overall I would say that quality is slightly better with the Minolta 5400. I am thinking to keep the Minolta since here in the UK it costs 400 pounds (715 USD) less that the Nikon and it has the Firewire connectivity as well.

Here I show a test, but I repeat I am not an expert and I do not know if the following comparison is OK for this purpose. The small pictures at the bottom are about 500% of the original image using PS. Nikon is on the left and Minolta on the right. The Nikon picture was treated with the unsharp mask to increase sharpness. The original was not sharp at all.

What do you think?<div></div>

[alex_jaimes1]

Hello,

 

Has anyone figured out how to san old 35 mm negatives? Im not even sure "35 mm" is the correct description: the film strip is the same size as 35 mm and fits perfectly into the feeder. But the holes are different from today's 35 mm and the image size is slightly differnt also. If I insert it manuually into the feeder by opening it up, the image can comfortably fit inside the scan window. However:

 

Feeding the negative doesn't work. The scanner rejects it.

If I put it in manually, the software acts as if nothing had been inserted so I cannot scan it.

 

The negatives are about 35 years old and I have tons I want to scan. With the 5000 of course.

 

Thanks,

alx