Negative Scanners

[Lizzard[MPE]]

Anyone have any ideas of good negative scanners?

quality ones? Maybe something that takes whole rolls as I used to ask them not to cut my negatives!

[Nowhere]

BTW this reminds of something (hope I won't hijack the thread too far)

For some time I'm sorting out the mess with my photos...most of them are digital but there are also few rolls of negatives (both from compacts and SLR)

So, for "proper archiving into digital format" (in the sense: it's unlikelly that I'll ever want to do it again, more properly; and might as well discard negatives) kind of deal, is trying to find a lab that makes PhotoCDs (NOT PictureCDs) and be done with it a good plan?

[degrub]

What do you intend to do with the images ? Is this just for yourself / family /hobby or for exhibition / professional work ?

If for the former , the Epson V750 Pro is probably the best deal.


If you don't mind cutting into strips, the Epson V750 is decent. The main issue is the resolving capability of the optics - limited to about 1800 - 2000 ppi, even though the sensor has many more elements per inch. Many people find it is ok for non-demanding 35 mm use and fine for scanning medium format film. Enlargements beyond 8x10 for 35 mm start to show loss of detail sharpness.

If you plan to do professional grade work, you need to consider at least a reasonable dedicated film scanner. The Nikon CS5000 or CS9000 are both very good for the price and resolve around 2700 ppi or so. THe main advantages are the ability of the scanner to pull out detail in the darkest (lightest for negs) areas of the frame with low noise. The scanner also does multisampling ( similar to what you did for the stacking of astro images) to further reduce random noise in the image and gain 1-2 more bits dynamic range. The Nikon does have a roll feeder. Also has ICE4 for defect removal. Good for enlargement to about 12x15 without loss of detail sharpness. Fairly easy to resell on ebay, but i would be cautious about buying a used one unless you get a solid performance guarantee from a reputable supplier.

Beyond that, you are into the drum scanners and the Creo type flatbed scanners and big $$.


With either of the above scanners, the vendor software may be adequate, but if you want more control of "developing" the image, try the demo versions of Vuescan and Silverfast. Vuescan is really pretty good and inexpensive. Silverfast is high end scanning software that gives you a lot of control over the scanning process once you learn to use it.

SOme basic reference sites:

https://www.scantips.com

https://www.computer-darkroom.com

https://www.luminous-landscape.com

Vuescan:

http://www.hamrick.com/vsm.html

Silverfast Ai Studio:

http://www.silverfast.com/show/silverfast-ai-studio/en.html

Scanner reviews ( reasonably done)

http://www.photo-i.co.uk/Reviews/interactive/Epson%20V750/page_1.htm

Hope this helps,

Degrub

[degrub]

Nowhere,

Unless you are just dumping the images to disk for display on a TV, i would stay away from PhotoCD. You should easily find a decent imaging shop to scan to tiff file and do basic corrections for a reasonable fee ( maybe 1 USD/ frame) . Or sometimes you can take a photo course at the university and borrow their scanning gear during the semester.

[Nowhere]

I see (but...university courses don't work like that here (unfortunatelly...)), will look into it.

BTW, from time to time I also think about repairing my old Zenit SLR and getting back into photography...but since times change, not only for darkroom (though using hands to control exposure of parts of photo always will be fun ), but...I'd also want to get a scanner. Those mentioned here are way out of my price range, and I don't think consumer-grade scanners are that great...but for some reason ("SCSI is sooo yesterday" mentality?...) it's quite easy to find unopened boxes (I noticed that you almost urge to get new scanner, degrub?) with Agfa scanners that are reasonably priced and seem to be "pro" enough... (like Agfa DuoScan HiD or Agfa Arcus 1200). Do you think getting something like that would work in "serious hobby" kind of stuff?

[degrub]

i believe that both of those are rebadged Microtek scanners. They are no better or worse than any other consumer scanner. THe optics are the limiting part of the equation. 1000 ppi will only let you enlarge 3-4x for a 35mm film. Fine for screen viewing. ANything above that is made up by the scanner driver. IF the price is right, they should be fine. i would look into using Vuescan since the manufacturer's software is usually pretty poor. Vuescan is cheap and quite decent.

[ND66]

Know someone that owns a Drum Scanner????

But......

...Even though I own a drum scanner I would clearly like to share the following story, showing how good low-end flatbeds can be. Before buying this scanner, I had a dealer send me prints made on an Epson printer from the same transparency, one was scanned with a $2,500 Agfa DuoScan, one from an Imacon FlexTight $14,000 CCD scanner, and the last was made by an ICG drum scanner, one of the most expensive made at $65,000.

The chrome that was scanned was of good quality and did not have a very large exposure latitude, probably 2.5 stops. The prints were only on 8x10" paper and since the chrome was on 6x7cm all the scanners could easily handle the dpi required with no interpolation at approximately 1500 dpi.

So what was the results? I have to admit, I was quite shocked. When looking at all three prints I could see slight differences in color gamut, but not sharpness. I concluded that none of the prints looked better than the others and any of these color differences, which were slight, could have been easily been corrected in PhotoShop. OK, I had to place my bets on which was which, so I placed them in 1,2,3 order with the best being in first place. When I turned them over and looked at their ID, the order I placed them in was, Imacon, Agfa, ICG. Go figure, right? But when associates were asked the same question, they picked different orders, so this is how I concluded that none of them looked better than the others, just a bit different.

.......

Read it all HERE:



.

[degrub]

Yeah, well if i was only scanning images with 2.5 stops of dynamic range, sure. Almost any of the consumer flat bed scanners will do that (normally about 8-12 f stops range). So he would be right for the case of some/many (who knows) slides (positives). At 8x10 i wouldn't expect to see a difference between the images. What he doesn't say is that the drum has the optical resolving capability to put the other two scanners to shame provided he knows how to control the aperture of the scanner. And he is correct, he doesn't know anything about scanning negs.

Not much of a test in my book. Certainly, his experience.

i'm surprised Michael has kept that on the site.

[ND66]

... What he doesn't say is that the drum has the optical resolving capability to put the other two scanners to shame .....

Absolutely correct.

Didn't notice he didn’t say that. What I wanted to point out that the lower end may be enough for someone needs.

Somehow I know Liz wants to go “poster” big ……. maybe I’m wrong.

.

[Nowhere]

bleh, I hoping for something better from those Agfa scanners... (after all, they look ridiculously large compared to "clearly consumer level" scanners, so... )

But I'm thinking...how usefull would be this astronomy software in this case?

[degrub]

ND66 - absolutely. There is so much marketing out there and not much fact that it is way easy to under buy or over buy.

Heh Heh , then Liz will be into drums fast !

Nowhere - i don't understand your question .

[Nowhere]

Rephrasing: what can be accomplished by using astronomy software (the one that stacks dozen "identical" images) on ordinary scans? (yep, assuming that you have time on your hands and made multiple scans)

Especially since you mentioned that Nikon CS5000 or CS9000 sort-of do that in hardware.

[degrub]

Well, using the multisample feature of the Nikons , for every two times you scan the image, you gain 1/2 bit of dynamic range further into the darkest part of the slide (lightest on a neg). THe maximum number of samples the Nikon will do is 16 so you get 2 full bits extra ( 14 bit range rather than a 12 bit range for example)
The way this works is that much of the noise seen in the darkest part of a slide that has been scanned is electronic noise since these devices are operating at 25-35 C, and as such is generally random in nature. So the ccd elements that change between the scans are picking up electrical noise. If it is truly random, averaging will remove most of the noise. If there is actual image data there, it will remain and you will be able to display it.

So for a 16 bit per RGB channel scan, instead of being able to see 12 bits of data, you can see 14 bits, which can be the difference in seeing any detail in the darkest portions of the slide.

The stacking software may not do this directly, but to get the better quality images, you average some together and then stack them. Another approach is to take a dark frame ( scope cover on) and then subtract out the pixels that have more than a "0" value of all of the frames. Then stack. THis works if the noise is not random.

Using stacking software can add photons to the pixels and bring up pixels that did not have enough data on the film single scan. It is difficult to do this for non-astro photos, since most pictures do not have little pinpoints of white to get the pixel alignment just right. However, in photoshop there is a way (i forget since i do not use it) to do this - i think it is by subtracting the two images from each other and the minimizing the difference that is displayed, to get the pixel alignment so that you can add the pixel values and not create noise.

Silverfast has a new feature call MultiExposure that will do this automatically. It seems to work ok for some images that have a wider dynamic range than the scanner can capture in one pass.

BTW, scanners that are not able to keep the ccd element fixed while multisampling have problems with pixel registration.; ie they have to take x passes with the ccd element. THe Nikons do not have this issue.

[chaoliang]

Nikon Coolscan LS5000

I use the LS5000 (=CS 5000?) to scan both negatives and slides. I've scanned about 3500 images. I'm very satisfied with it till now.

Its scan resolution is 4000 dpi (2700 dpi was with LS 2000, which I onwed before). Like degrub commented: Quality device with reasonable price. You can buy the optional roll feeder. For my purpose, I cut the rolles into 6-Frames-Stripes, then keep the stripes in slide archiv albums for easy viewing.

BTW: The NikonScan software is i.m.o much better than Silverfast and VueScan for this scanner.

chaoliang

[VJ]

One thing to consider when scanning negatives is digital storage...

With the current crop of harddisks and writable media, it might be less of an issue, but organizing the scanned data is also important...


Jörg