Prosumer (post-Minolta/Canon) film scanner recommendations?

[jens_finke]

Hello people

 

Apologies for this very long post. I'd like advice on buying a high

resolution scanner for slides. I know that similar queries have been

posted countless times before, but that was before the recent demise

of Minolta. I've tried to be precise in stating my requirements, which

I'm sure are shared by other people. Models I've considered are

briefly covered at the end of this post.

 

Film stock to scan

 

Several thousand mostly 100 ISO transparencies, various makes, shot

over 20 years. Consumer stock; plenty of Kodak Elite, no Kodachrome.

No tripod used, but steady hands and bright lighting in general made

for fast shutter speeds and minimal camera shake; most of them should

make decent A2-size prints.

 

Scanner speed

 

Mostly irrelevant; image quality is far more important (I know, even

with thousands to scan...) Obviously, time would be a factor if, as is

the case with the Plustek 7200i, applying IR dust removal takes 40

minutes per scan instead of 2 minutes, rendering the Plustek's dust

removal option completely useless.

 

ICE/ICE4

 

The films were stored, mostly mounted, in plastic or paper sleeves on

a damp island (tsk tsk), with the unfortunate consequence that a

handful of films became infected with some kind of fungus. Whilst I do

have the time and willpower to clean dust and scratches manually (in

fact, I'd prefer this over ICE if it doesn't do an absolutely perfect

job), advice on the effectiveness of ICE4 in rescuing whatever image

data might remain *under* fungus would be welcome, or simply

clarification about Infrared dust removal, ie. does it just use IR to

detect dust, and then interpolate/clone the missing data? To repeat, I

don't mind doing dust removal manually, even if it means spending half

an hour on each shot.

 

Resolution

 

Scanner output should contain all of the original image data. Remember

that the slides are not pin sharp, but reasonably sharp nonetheless.

The ability to finely resolve film grain isn't important, I think; the

limit I suppose should be one or two pixels per grain in order not to

lose any image information. I have a feeling that 4000dpi lies just

below the ideal threshold for 100 ISO transparencies, assuming that

the scanner's focusing is accurate.

 

Effective dynamic range

 

One of the factors for choosing my current scanner, a Minolta Scan

Dual III (2820dpi), was the high *effective* dmax (not the ridiculous

4.8 claimed, but the real world ability to pull out both shadow and

highlight detail in the same scan). I didn't realize at the time that

I'd be needing to archive stuff, otherwise I'd have gone for a 4000dpi

model and presumably woudn't be posting here today!

In practice, however, the Minolta fails to deal effectively with

contrasty shots, of which I have many. As such, I'm pretty much

resigned to the fact that whatever scanner I use, I'll end up having

to make at least two scans of each contrasty shot, one optimised for

shadows, the other for highlights, then merge the two in Photoshop.

 

Noise

 

Given the problems that every scanner has with dynamic ranges, a

scanner offering low noise in shadow areas on scans with longer

exposures (precisely to pull out shadow detail) seems more important.

No point having so much noise in shadow areas that cleaning it up

removes the extra detail I want to extract.

 

Colour accuracy

 

I've never seen a slide scanner that sees colours correctly, so will

be using custom made profiles (using Faust's targets) for calibration.

As such, out-of-the-box colour accuracy shouldn't be *too* important.

However, given that colour correction will inevitably be needed after

scanning, whether via the profile or with curves etc, 12 or 16 bit RGB

colour channels (output as 16 bit TIFF) would be essential to avoid

too much data loss.

 

Price

 

Anything under $1000 is ideal, and I'd be prepared to go to $1500 if

warranted. Would have to be damn good though! Can't afford any more.

 

Warranty

 

My scanner of choice had been the Konica-Minolta 5400 II. As there's

still no real guarantee of useful support from Sony for Minolta's film

scanners other than in one press release ("including film scanners"),

I'm now in a quandary. The Minolta Scan Dual III was dead on arrival;

luckily I'd gotten a verbal guarantee from the shop about immediate

replacement, and got a new one a few days later. Now with production

on the 5400 II ceased, it might not be so easy.

Scanner needs to work perfectly for one year (with efficient

support/guarantees over that period); enough time to scan in the bulk

of my existing slides. More than that would be nice but not essential.

 

 

I've considered the following scanners, each with their drawbacks:

 

* Minolta DiMAGE Scan Elite 5400 II

from what I've seen on the net, appears as though a good chunk of the

extra resolution is gobbled up by dodgy focusing and possibly sub-par

optics. Still, should be enough for my needs (I guess it still reaps

the equivalent of 4000dpi). Dmax might be good, but I don't believe

the hype. Oh, and every Minolta scanner I've used gives slides a

horrible magenta cast.

 

* Plustek OpticFilm 7200i

from what I've seen

(http://www.ephotozine.com/equipment/tests/testdetail.cfm?test_id=403),

scans appear to be amazingly sharp, finely resolving grain; optics and

focusing would therefore appear to be excellent. However, low

effective dmax is putting me off (not too clever with shadow detail),

and I'm not sure whether the scanner outputs the full 12-bits per

channel, or cuts it all down to 8 per channel. Also, its

ICE-equivalent, iSRD, seems to be way below par, and way too slow - 40

minutes as opposed to 2 minutes, at 7200dpi, is unusable (manually

spotting would actually be quicker, and more accurate).

 

* Nikons

I'm not convinced that 4000dpi is sufficient to pull out all the image

detail in slides, even those without a tripod. All of Nikon's models

appear to have lower effective dynamic ranges than the Minoltas, but

as neither are perfect, I don't think this necesarily translates into

more work for me (ie. the same contrasty slides would have to scanned

twice irrespective of scanner make). Optics need to be perfect to get

the most out of the 4000dpi sensor.

 

Any suggestions would be very much appreciated.

 

Many thanks,

 

Jens Finke

[jtk]

In addition to being an orphan, the 5400II does not compare well to Nikon mechanically. Open them and look inside, or just consider why Nikon weighs twice KM.

 

KM's nominal 5400ppi doesn't out-resolve Nikon's 4000ppi, but it does make bigger files (waste data). I've owned and compared 5400II to my current Nikon V: they're equal, no more and no less, in terms of speed, detail resolution and image quality (Nikon has higher level Ice4, a minor advantage). Some say Nikon's autofocus and "manual" focus are superior, but others prefer Minolta's focus knob.

 

In fairly large prints (12X18) 4000ppi film scans record more visual detail than do D70/10D, and it's easy for a relatively personally skilled color photographer (ie. someone experienced in color negative darkroom printing) to quickly and easily do superb, precise color matches with both Nikon and Minolta, with their respective software, and with Vuescan.

 

Personal visual skills (trained color perception and individual acuity) are important, despite the lazy desire to buy rather than learn.

 

Many comparisons of actual prints have convinced me that D70/20D have disadvantages, as well as advantages, both with big prints and with little ones (eg 8.5X11) compared to 35mm: Fine 6 or 8mp digital do often seem to demonstrate more "sharpness," but they often lose the visual content between the points of sharpness (eg line pairs look sharper, actual images look less detail rich). From what I've SEEN, and to my taste, the film/digital crossover occurs with 5D.

[thomas_janik]

Jens:

 

I have the Minolta 5400 (1st version, no plastic) and a Nikon 9000. I have kept both because on a few slides (they must be very sharp, small grain, e.g. KM25) the extra dpi of the Minolta is significant. The Nikon will do a better job with very dark slides. It is reallly on a case-by-case basis. If you haven't used a tripod and no Kodachrome, I'd go with the Nikon

[geoff....]

I have been doing a bit of research as I plan to change my 5400 soon. Not overly happy with the sharpness. Had a Polaroid Sprintscan and looking at old scans was amazed at how much finer the detail was achieved with it. The Microtek 4000tf is the evolution of the Sprintscan. Those "few" that use it talk of the fine detail and ability to draw out detail from the shadows and highlights. The Dmax of 4.2 Microtek say is more honest and say good scans can be achieved with one pass, the other manufacturers achieve their Dmax of 4.8 through multi-scanning.

 

I am interested in other peoples opinion of the Microtek 4000tf.

 

Perhaps the answers might assist the original posters assesment.

 

Geoff Wise

www.wises.com.au

[thomas_janik]

Geoff:

 

Have you tied using manual focus with the Minolta? It can result in much sharper scans. See:

 

http://www.photozone.de/2Equipment/reviews/elitecoolscan.htm

[jens_finke]

Thanks all for the responses so far. It does seem as though the Nikons (especially the Coolscan 5000) would be best for my requirements, given the 100ISO with no tripod combination. Good comparison at http://www.bythom.com/coolscanv.htm

 

All I know of the 5400 is from reviews and online samples; the latter didn't appear to resolve any finer detail than the Nikons at 4000dpi, though I'm sure a lot of that has to do with focusing correctly (and the slides used for the comparison may not have been pin sharp). I'm used to Minolta's focusing system on the Scan Dual III, and know that the auto focus doesn't always get it right; manual focusing is often significantly better, but obviously takes rather longer to get right. Any opinions on the quality and acuity of the 5400's optics?

 

Les: your photos make interesting comparisons. There's a very obvious colour difference between versions scanned at 2700dpi and those at 4000 (eg. the orange-blue flower on Reala); I presume you scanned them at different times, as side by side, they're not aligned the same either. All a bit bluish (other reviews mention that too regards the Nikons), though easily corrected with calibrated profiles or curves. The cast is nowhere near as visible as the Scan Dual III's magenta cast, thankfully. Any feedback on colour casts on the 5400?

 

Useful thread, in case someone else is interested: http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=006mD6 (discussion of the 5000 and KM 5400)

 

Geoff: like Thomas, I'd be interested to hear about manual focusing versus auto on the 5400.

 

Dmax: as I wrote in my original post, I've long given up on expecting a film scanner to deliver the goods in terms of dynamic range, and am resigned to having to scan stuff twice or thrice, each optimized for different areas, then merging them post-scan. Still, seems as though both the Coolscan 5000 and KM 5400 (both 16-bit) are quite similar when it comes to hauling detail out of shadows, and for slides without dynamic extremes, should be fine just scanning them once.

 

Finally, one for Les: Chasing the Lizard's Tail? Yah, don't believe a word of it. It's online in its entirety (http://www.bluegecko.org/lizard) and for free, though if you want to avoid eye strain, I still have a ton of originals I'd like to get rid of. Just let me know (email address is all over the website). There are no photos in it, mind you, in case that's what you're interested in.

 

Thanks again

 

Jens

[jens_finke]

Les, the coolness I mentioned was seen on a calibrated TFT, nothing better. I won't be home for a month or two, so I can't say how they'd look on a decent CRT. I know TFTs aren't perfect for photos as far as colour goes (but are brilliant for seeing flaws and noise CRTs hide so well), as so much depends on the viewing angle, but the slight blue cast was evident throughout in comparison to other pictures I know well (ie. on my own system) that were colour balanced on a calibrated CRT, and subsequently viewed on several different monitors. Nothing scientific, and it might just be me, it's just that the slight bluish cast I saw did make sense in the light of some reviewers reporting the same. Remember you said your online scans had not been tweaked other than for the copyright text. Then again, colour calibration on the internet is a whole different aquarium of off-topic fish...

[geoff....]

In regard to manual focusing, that is OK if you are using the Minolta software, as a screen pops up to assist. If you use Vuescan, which I use for most shots except for unusual lighting, which Vuescan does not seem to interpret well, then manual focusing is a bit of a hit and miss.

 

In addition scanning with the manual focus button with Minolta software still has not given the same sharpness that the old Sprintscan did.

 

My wokflow with the Minolta is to turn it on for a half hour, then calibrate, then start scanning. I leave it for a half hour because this seems to reduce the green lines that appear on some scans, a issue reported extensively in the past.

 

What is interesting in Australia, is that there has been stock outages on the Nikon 5000 for a while now with no date when stock will be available, it was supposed to be April.

 

Perhaps they are reducing stock in advance of a new model.