Pages: [1] 2 3   Go Down

Author Topic: macro lens vs film scanners for digitizing transparencies  (Read 31222 times)

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
macro lens vs film scanners for digitizing transparencies
« on: December 27, 2011, 04:04:36 am »

I have a large amount of 120 (mostly 6x6cm) and 135 (36x24mm) transparancies in a family archive that I want to be able to digitize with high quality. I am myself not too familiar with digitizing, and looking around it seems to be a bit of a hassle and/or very expensive to get better than mediocre quality. I can do with some hassle, but have a limited budget.

There seems to be three options: 1) flatbad scanner 2) dedicated filmscanner 3) digital camera with macro lens.

According to my research flatbed scanners have poor dynamic range and hugely overestimate their resolution, 2300 ppi is what a good scanner (epson v750, €700) can do in practice despite 6400 ppi claimed resolution. It is the cheapest and quite efficient solution, but I think I want better quality, both in resolution and DR.

Concerning resolution I'd like to sample also grain so large prints get analog look with visible grain up close. It seems about 4000 ppi is what is needed for this goal (?).

Dedicated film scanners are plentiful for 36x24mm, but very few exists for 120 film. Reflecta MF5000 is one of the few, which gives about 3050 effective ppi and a bit better dynamic range than the best flatbeds. The discontinued Nikon Coolscan 9000 still seems to be king among "affordable" scanners, but still much more expensive than the Reflecta. Better DR still and 3900 effective ppi.

The Nikon Coolscan seems to provide the quality I desire but is a bit too expensive and hard to find (€6500 is a price I have seen), with some compromise the Reflecta MF5000 (€1600) may do as well.

Finally, there is the option to take a photo of the transparency using a sharp macro lens. I would use a 21 megapixel 5Dmk2 with 1:1 macro lens which would in theory yield about 3900 ppi (stitching or lower res required for 6x6, would probably do lower res for casual digitizing and stitch when digitizing for a print). This option was the hardest to find information about, so I don't know what quality to expect. In theory dynamic range would be great, HDR also possible (as is with some scanners), and if it is possible to focus and get a sharp shot, resolution would be good too. No automatic infrared dust detection though, but I think/hope I can live with that.

Has someone experience with these methods and recommendations? I'm especially interested in the macro lens method since I have not been able to find any information of what quality one can expect from that compared to scanners.
« Last Edit: December 27, 2011, 04:09:15 am by torger »
Logged

dergiman

  • Full Member
  • ***
  • Offline Offline
  • Posts: 206
    • http://www.philippderganz.com
Re: macro lens vs film scanners for digitizing transparencies
« Reply #1 on: December 27, 2011, 04:57:24 am »

I am also planning to digitize the family archive. It is mainly slides (mounted in glass) and B&W negatives in 36x24mm. I think i will go the 5dmk2 with makro lens setup way. Taking photos is just a lot faster than scanning. Those old film images don´t contain that much information, so 21 MP should capture all the relevant information just fine. The resulting RAW files are also smaller than the TIFF files from a scanner.
Logged
Philipp Derganz Photography
www.philippderganz.com
Architektur Photo
Wien, Austria

mediumcool

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 770
Re: macro lens vs film scanners for digitizing transparencies
« Reply #2 on: December 27, 2011, 05:45:04 am »

I have a large amount of 120 (mostly 6x6cm) and 135 (36x24mm) transparancies in a family archive that I want to be able to digitize with high quality. I am myself not too familiar with digitizing, and looking around it seems to be a bit of a hassle and/or very expensive to get better than mediocre quality. I can do with some hassle, but have a limited budget.

Whatever you do, it will be a hassle: this is drudge work, so finding an efficient and effective way to work is the key. Have some nice music and bevvy to while away the waiting.  ;D

There seems to be three options: 1) flatbad scanner 2) dedicated filmscanner 3) digital camera with macro lens.

I would consider a flatbed for 120 film, but never for 35mm. I will return to the camera option later. FYI, I have scanning experience spanning many years, mainly for print reproduction.

According to my research flatbed scanners have poor dynamic range and hugely overestimate their resolution, 2300 ppi is what a good scanner (epson v750, €700) can do in practice despite 6400 ppi claimed resolution. It is the cheapest and quite efficient solution, but I think I want better quality, both in resolution and DR.

You’ve done some homework! The Epson has decent dynamic range, and there is nothing else available that is better in its segment (though Epson software sucks in my NSHO*). Occasionally large-format pro scanners (Fuji, Lino etc.) come up, but they are often hard to fix, and may need to be used with an older OS for compatibility. For perspective, my first medium-format scans were made on an Agfa Argus II scanner, enlargeable 200–300% for offset with excellent results. The Argus’ native resolution was 600 ppi.

Concerning resolution I'd like to sample also grain so large prints get analog look with visible grain up close. It seems about 4000 ppi is what is needed for this goal (?).

4000ppi at a minimum, yes; the 2700ppi resolution on many scanners turns grain into clumpy porridge.

Dedicated film scanners are plentiful for 36x24mm, but very few exists for 120 film. Reflecta MF5000 is one of the few, which gives about 3050 effective ppi and a bit better dynamic range than the best flatbeds. The discontinued Nikon Coolscan 9000 still seems to be king among "affordable" scanners, but still much more expensive than the Reflecta. Better DR still and 3900 effective ppi.

Not heard of the Reflecta; it looks like a re-badged Plustek. The Polaroid Sprintscan 120 scanner and its evil twin (joke) from Microtek come up occasionally on eBAY. Another option is a medium-format Minolta scanner, though they seem to be highly regarded, thus not cheap.

The Nikon Coolscan seems to provide the quality I desire but is a bit too expensive and hard to find (€6500 is a price I have seen), with some compromise the Reflecta MF5000 (€1600) may do as well.

There may not be enough difference between the Reflecta and the Epson flatbed (for 120) to justify the price differential.

Finally, there is the option to take a photo of the transparency using a sharp macro lens. I would use a 21 megapixel 5Dmk2 with 1:1 macro lens which would in theory yield about 3900 ppi (stitching or lower res required for 6x6, would probably do lower res for casual digitizing and stitch when digitizing for a print). This option was the hardest to find information about, so I don't know what quality to expect. In theory dynamic range would be great, HDR also possible (as is with some scanners), and if it is possible to focus and get a sharp shot, resolution would be good too. No automatic infrared dust detection though, but I think/hope I can live with that.

For best results, you would need a lens optimised for 1:1 when shooting 35mm slides; many macro and enlarging lenses are best at 1:10 or so. Is it possible for you to easily try this out? Many years ago I photographed a 5" x 7" Ektachrome transparency for a client who ran a Cibachrome printing lab—his largest enlarger was 4" x 5". I turned my LPL colour enlarger head upside down, put an improvised light box over the head made from a small polystyrene cooler (diffused with white perspex) and shot it with Ektachrome Duplicating film on my Nikon F w/55mm macro. While I never saw the result of my labours, the client said a printed 8 x 10 was very acceptable. If you go this way, you will be able to balance your light source much more easily than in the bad old days! There are a lot of slide copiers on eBay, including the ancient Bowens Illumitran. Here’s a cheaper one, but without the copystand by the look of it.

FWIW, Schneider Comparon enlarging lenses are optimised for around 1:5 whereas Componons are best at 1:10 and more; Comparons are much cheaper on eBay.

In conclusion, I would get the best flatbed I could afford for the 120 scans, and a dedicated 35mm scanner for the littlies. HTH

Just spotted on eBay: Polaroid SprintScan 4000 Plus (SCSI and FW). Oops, not working.

*NSHO = not-so-humble opinion.    ;D
« Last Edit: December 27, 2011, 05:48:02 am by mediumcool »
Logged
FaceBook facebook.com/ian.goss.39   www.mlkshk.com/user/mediumcool

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: macro lens vs film scanners for digitizing transparencies
« Reply #3 on: December 27, 2011, 05:46:57 am »

There's a lot of carefully made landscape shots among those transparencies so there's as much detail as the film can take there. I've looked at tests for macro lenses for the Canon, and it seems like the new Sigma 150mm could be a good candidate. Canon's own 100mm cheap has some barrell distortion and L version some chromatic abberation, I would probably go for the cheaper of those two since a slight barrell distortion is easier to work with than chromatic abberation I think.
Logged

mediumcool

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 770
Re: macro lens vs film scanners for digitizing transparencies
« Reply #4 on: December 27, 2011, 05:51:05 am »

There's a lot of carefully made landscape shots among those transparencies so there's as much detail as the film can take there. I've looked at tests for macro lenses for the Canon, and it seems like the new Sigma 150mm could be a good candidate. Canon's own 100mm cheap has some barrell distortion and L version some chromatic abberation, I would probably go for the cheaper of those two since a slight barrell distortion is easier to work with than chromatic abberation I think.

I would try an enlarging lens first, but of course you’d need bellows etc. Please keep us posted; this is a problem that many people face in preserving older work.
Logged
FaceBook facebook.com/ian.goss.39   www.mlkshk.com/user/mediumcool

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: macro lens vs film scanners for digitizing transparencies
« Reply #5 on: December 27, 2011, 05:54:34 am »

Whatever you do, it will be a hassle

Thanks for the detailed reply! I actually don't have a macro lens, but sooner or later I'd like to have one for macro photography so I would buy one, but would think more about having one suitable for copy work (low distortion, high sharpness, good corner performance, bokeh less important etc).

I don't know which distances these standard 1:1 macro lenses are optimized for though... the resolution tests are surely not made at 1:1 distance...
Logged

EricWHiss

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2639
    • Rolleiflex USA
Re: macro lens vs film scanners for digitizing transparencies
« Reply #6 on: December 27, 2011, 12:11:17 pm »

I just went through this myself.  Thousands of old family slides some dating back to 1950 most 35mm.  I looked at a nikon 5000 scanner with the auto loading feature but its a lot of money and I also looked at Scancafe.com.    I decided to do it myself with my MFDB and used one of the slide copier attachments which you can find on ebay.   This worked fairly well and I did several hundred.  I'm not sure how the digital capture would compare to the scan but it works well.  I did several hundred slides in a few hours.  Later I moved to a bowens illumitran with slide carrier and this was even faster to load and unload.  I tried a macro lens first then an enlarging lens but got decent work from both.

Here's what I can tell you from the experience:  The biggest problem you will have will be dust on the slides themselves and adjusting faded colors.  Some scanning software has very good color restoration software that can save a bunch of time.   

I still have about 2000 more to go.  I think for these I will just review in a slide projector and send the best to scancafe!

Logged
Rolleiflex USA

Rob C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 24074
Re: macro lens vs film scanners for digitizing transparencies
« Reply #7 on: December 27, 2011, 12:18:20 pm »

I've done it for 6x6 and 6x7 with my D700 and a 2.8/105 Micro Nikkor. It was relatively easy to sit a Kodak lightbox on its side, make a screen-sized black card mask for the two formats, stick that to the perspex and then just switch off the room lights and expose via the transmitted light of the box, which is supposed to be as close to daylight as matters. Okay, I didn't use anything other than my eye and the grid lines in the camera, but had I hung onto my old Durst enlarger stand...

You obviously only get to use a small square out of a D700 frame, but 6x7 does better.

If this isn't for pro use, then it should be perfectly fine for domestic use. I've some such images on my website, but have never gone to the lengths of printing anything up, so can't claim anything there for the technique. Yes, if you can fit an enlarger lens, you'd probably do better at the required magnification.

Were 120 film scanners reasonable in price, I'd be back with 120 film! However, they are not, and I suspect that, medium long-term, film will vanish completely, so not worth spending big bucks (for me).

Rob C

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: macro lens vs film scanners for digitizing transparencies
« Reply #8 on: December 27, 2011, 05:04:02 pm »

Hi,

I have a decent MF film scanner (Minolta Diamge Scan Multi Pro) and it is a pain in the neck. Definitively good for great 70x100 prints, but you need to work a lot to achieve excellent results. Scanning is slow, don't count on more than 2 slides per hour including post processing. Colors may be weird and image quality may be far from what you expect from a DSLR.

Shooting with a DSLR and a macro lens may be worth trying.

Best regards
Erik


I've done it for 6x6 and 6x7 with my D700 and a 2.8/105 Micro Nikkor. It was relatively easy to sit a Kodak lightbox on its side, make a screen-sized black card mask for the two formats, stick that to the perspex and then just switch off the room lights and expose via the transmitted light of the box, which is supposed to be as close to daylight as matters. Okay, I didn't use anything other than my eye and the grid lines in the camera, but had I hung onto my old Durst enlarger stand...

You obviously only get to use a small square out of a D700 frame, but 6x7 does better.

If this isn't for pro use, then it should be perfectly fine for domestic use. I've some such images on my website, but have never gone to the lengths of printing anything up, so can't claim anything there for the technique. Yes, if you can fit an enlarger lens, you'd probably do better at the required magnification.

Were 120 film scanners reasonable in price, I'd be back with 120 film! However, they are not, and I suspect that, medium long-term, film will vanish completely, so not worth spending big bucks (for me).

Rob C
Logged
Erik Kaffehr
 

roskav

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 254
    • http://www.roskavanagh.com
Re: macro lens vs film scanners for digitizing transparencies
« Reply #9 on: December 27, 2011, 06:37:28 pm »

Sold my Imacon last year knowing that I would come up against this problem at some point. When asked for an unscanned image from the archive -  I used an Aptus 75 back on a hass V with extender on a tele lens.  Results were quite good for the purpose I was scanning for ... easy to get colour right from a colour neg and is also quick in comparison to scanning.  It also fits in right away with your workflow by running the files through your normal raw proccessor.  I set up a vertical black frame to hold the negs and then about 1.5 m behind had a white background with 2 strobes trained on it on either side to give a strong light source.  I then shot it horizontally.  If you were looking at it from the side you would get from right to left:  Camera - frame  - strobe then white background.  If you made a robust frame with easy setup you could get through quite a bit of scanning in a short time.  The problem then is the usual dust and hair removal.   Another drawback is that it may be difficult to set up profiles on any camera that would do the same job as scanner software... hence any negs are shot in real colour .. then converted in photoshop after... You would have more immediate feedback when scanning trannys.

R
Logged

mediumcool

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 770
Re: macro lens vs film scanners for digitizing transparencies
« Reply #10 on: December 28, 2011, 03:34:13 am »

Sold my Imacon last year knowing that I would come up against this problem at some point.

Crazy man. I confess that I’d forgotten about the Imacon in my post (have almost bought a few on eBay, but they always went over my reserve). Each passing year makes a dedicated scanner less of an option. Unless you want to be like the last guys processing Kodachrome!
Logged
FaceBook facebook.com/ian.goss.39   www.mlkshk.com/user/mediumcool

Stefan.Steib

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 471
    • HCam - Hartblei Pro Photography solutions
Re: macro lens vs film scanners for digitizing transparencies
« Reply #11 on: December 28, 2011, 05:38:15 am »

If you can use your existing setup of DSLR with a Makro lens + maybe extension rings you´ll be fine for a painless setup, for lighting you could use something like a long TTL Cable and your flash, Tripod + some gaffer tape and and some white opaque Plexi.
Buying a lot of equipment (scanner) may only be interesting if you really want to dig into this and the amount of slides is really big.
And finally: do yourself a favour, put up all these images on a light table, gather your family and sort out what is really of interest (will be a fun evening!)  and if you have  a certain amount sorted out bring it to a lab and let somebody else do it, if this is only 100 or maybe 200 slides this is the best way...... :)

Greetings from Munich
Stefan
Logged
Because Photography is more than Technology and "as we have done it before".

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: macro lens vs film scanners for digitizing transparencies
« Reply #12 on: December 28, 2011, 01:02:34 pm »

I did a quick test comparing CanoScan FS4000US with vuescan multipass scanning (HDR) on a 36x24 transparency, and also made a shot with the camera (5Dmk2). The camera needs HDR to match the multipass scan, and it is required or else the dark parts are excessively noisy. So it will be messy.
Logged

Rob C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 24074
Re: macro lens vs film scanners for digitizing transparencies
« Reply #13 on: December 28, 2011, 01:42:38 pm »

I did a quick test comparing CanoScan FS4000US with vuescan multipass scanning (HDR) on a 36x24 transparency, and also made a shot with the camera (5Dmk2). The camera needs HDR to match the multipass scan, and it is required or else the dark parts are excessively noisy. So it will be messy.


That's what I have, too; unfortunately, it seems to have stopped speaking to my computer. It might be that it no longer loves PS6, but it used to... I'm hoping it's only a damaged USB cable, but since the kit one has a thinggy on the cable and my other cable does not, and as I'm not sure what that thinggy really does (imagine it cuts out interference?), I'm nervous about trying it.

When it worked, I loved it.

Rob C

Ben Rubinstein

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1822
Re: macro lens vs film scanners for digitizing transparencies
« Reply #14 on: December 28, 2011, 01:49:43 pm »

I did a quick test comparing CanoScan FS4000US with vuescan multipass scanning (HDR) on a 36x24 transparency, and also made a shot with the camera (5Dmk2). The camera needs HDR to match the multipass scan, and it is required or else the dark parts are excessively noisy. So it will be messy.

I was wondering about that.

How does the detail/sharpness compare?
Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: macro lens vs film scanners for digitizing transparencies
« Reply #15 on: December 28, 2011, 02:52:35 pm »

I was wondering about that.

How does the detail/sharpness compare?

I'm no macro photographer (yet) so I actually don't have a macro lens. I put a 2X teleconverter on a 24mm TS-E to at least get some magnification (only covered a quarter of the frame still), which of course is no test for resolution. Before running to the store and buy a macro lens I thought I should try out dynamic range, so that was what I did.

Transparencies can be really dense it seems. I picked out photo of a dark scene as a test. There is a lot of information in those dark areas I could see when making the longer exposures with the camera. To get that information with reasonable quality HDR is required, just like multipass is required to get sufficent quality with the canoscan, but that is automatic with vuescan, HDR with the camera is more cumbersome. With the camera and HDR one can exceed the canoscan dynamic range a little, but without HDR the scanner clearly delivers better shadow quality (in multipass of course). I thought multipass was HDR but it seems to just be averaging, it does improve noise though.
« Last Edit: December 28, 2011, 02:54:34 pm by torger »
Logged

Camdavidson

  • Full Member
  • ***
  • Offline Offline
  • Posts: 185
    • http://www.camerondavidson.com
Re: macro lens vs film scanners for digitizing transparencies
« Reply #16 on: December 28, 2011, 05:19:42 pm »

scancafe.com

http://www.scancafe.com

Scanned in India - super secure UPS shipping.  Very fairly priced.  I had over 1000 aerial chromes scanned last year and was very happy with the results.  ( order the higher rez scans)

They scan and clean the files.  Scans are made on Nikon 9000 scanners.

Also look at Jaincotech.com  I had the selects and sisters for two Geographic stories plus my Over Florida book scanned by Jainco and they were fantastic.

They scan on Imacon scanners in the US and India.  Also fairly priced. (but higher than ScanCafe.com)


Logged

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: macro lens vs film scanners for digitizing transparencies
« Reply #17 on: December 29, 2011, 03:46:18 am »

A professional service is probably best for bulk scanning. There's some fine art stuff there too, and then it would be kind of nice to be able to it myself. If macro lens + HDR (+stiching for medium format) just works it could be worth the effort for fine art prints, I usually spend 12 - 16 hours of tuning those anyway, an hour or two in digitizing would not be too heavy in that process.

Now I just need a macro lens and a good film holder to test the concept fully...
Logged

Rob C

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 24074
Re: macro lens vs film scanners for digitizing transparencies
« Reply #18 on: December 29, 2011, 03:55:19 am »

The greatest problem, outwith a purpose-built rig, is keeping parallel to the slide/print being copied. On small blowups it isn't so important, but anything over about 600x600 pixels is going to show shortcomings, both in sharpness over the field as well as in the shape of the rectangle or square you will be printing/transmitting up...

Rob C

mediumcool

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 770
Re: macro lens vs film scanners for digitizing transparencies
« Reply #19 on: December 29, 2011, 05:00:18 am »

Logged
FaceBook facebook.com/ian.goss.39   www.mlkshk.com/user/mediumcool
Pages: [1] 2 3   Go Up