Best camera setup for Fine Art Repro? Scanning backs? High MegaPixel DSLR?

[torger]

Here's a H4D-50 multishot available from UK, seems to be a decent price, £10k buy now: http://www.ebay.com/itm/281585779024

[BobDavid]

A used Hasselblad CF-39 MS is a good choice. So are the Sinar MS backs of the same vintage. Michael Ulsaker at Ulsaker Studio in Connecticut is a great contact. Stay away from Phase. Their present generation of single-shot cameras are fine, if you want to spend a zillion dollars. Most museums, archives, etc that use cameras are using multi-shot systems.  A used CF-39 MS back should not be more than $7K.

[tom_l]

Of coure they're better,
but reality isn't really like that. We're not always in our basement studio.
We carry our equipement on archeology sites,  to old collectors in their even older castles, with centuries old wooden floors. Sometimes I just drive to a gallerist to take 5 paintings, in 1 hour and to drive back by time for the next shooting. So the 1-shot solution is a very good one. Last year I did a huge chinese porcelain collection in 16 hours. The collector and his gallerist were a bit sceptical: "Sir, last time, the photographer did 16 shots to a picture." Yeah, sure, but then I have to stay a week.
The perfect solution would be to have scanning back, a multi-shot, a 80MP back, a view camera and a DSLR available, all the time.  For a 1 person studio, this is simply not possible. A 60-80 MP back can get a lot of thing done.

Tom

[Theodoros]

Using multishot or microstep can't be more than 5% of the time need if one was to use single shot and that applies only to captures... Setting up the capture is (obviously) the most time consuming part of the job.

In return, the time needed for processing the files to achieve colour accuracy, is much less with multishot, to an extend that one actually works much faster and has far superior results by using it (especially if one uses a Sinarback).

I have 100% success with all my 16x shots when using my Sinarback 54H on my Fuji GX-680 and (off course) with my Contax 645 too... Same happened with my previous Imacon 528c and same applies with my 4x shots with my CF-39MS... Learning to work around vibration and environmental limitations comes with the job... One must learn to work around the limitations and find solutions on how to overcome them.

[8x10 user]

Using multishot or microstep can't be more than 5% of the time need if one was to use single shot and that applies only to captures... Setting up the capture is (obviously) the most time consuming part of the job.

In return, the time needed for processing the files to achieve colour accuracy, is much less with multishot, to an extend that one actually works much faster and has far superior results by using it (especially if one uses a Sinarback).

I have 100% success with all my 16x shots when using my Sinarback 54H on my Fuji GX-680 and (off course) with my Contax 645 too... Same happened with my previous Imacon 528c and same applies with my 4x shots with my CF-39MS... Learning to work around vibration and environmental limitations comes with the job... One must learn to work around the limitations and find solutions on how to overcome them.

+1

I set the time between shots to 9 seconds on mine but that is just because I want to lower the draw that my powerpacks have on the wiring in my studio. With better wiring I'd be able to do a 16 shot image in less then a minute. Betterlight systems take much longer.

[EricWHiss]

A more promising route is to capture more than three color planes.

I think this is a good way to go, very much like the lucky people (mostly women) that are Tetrachromats and have 4 cone classes instead of the normal three.  Couldn't this be done with a sensor without a color filter array and either a filter wheel or a set of different illuminants?  I think MegaVision is doing something like this now -  I visited them recently - and doing it this way they time the capture of each color independently such that they get a full well.  If I'm not mistaken they also use UV and IR illumination too if needed.   

[ErikKaffehr]

Hi,

The Bayer pattern has a pair of green pixels for each blue and red pixel. Those greens may or may not be different. I have the impression that raw converters have provisions for handling different G1 and G2 channels.

Sony had some models having green and "emerald" as G1 and G2, but I think the concept has failed.



CFA response curves. I don't know if those contain sensor spectral response and IR filters. Here is the link.

#2 Is Kodak CFA
#3 Is Philips (may be similar to DALSA, as DALSA has taken over Philips CCD development"
#5 Is Foven
#6 Is Sony Super HADII

Best regards
Erik

A more promising route is to capture more than three color planes.

In 1992, Michael Vrehl, a student of Joel Trussell at North Carolina State  presented a paper at the SPIE Imaging Conference. I'm sorry I can't find a link to the paper itself, only one to the abstract.

"The quality of color correction is dependent upon the filters used to scan the image. This paper introduces a method of selecting the scanning filters using a priori information about the viewing illumination. Experimental results are presented. The addition of a fourth filter produces significantly improved color correction over that obtained by three filters."

I remember being quite impressed with the improvement in color accuracy afforded by the addition of the fourth filter.  The common term for cameras that have more than three filters is multispectral cameras. The ones I've seen are for scientific use. You can turn a monochromatic camera into a multispectral camera with a color wheel.

[torger]

Leaf has had a color wheel camera if I remember correctly. Today you can buy Sinar CTM which I think is a dual filter + the ordinary bayer sensor. Sinar eXact multi-shot + CTM maybe is the highest end among commercially available systems today, but yes the workflow is considerably more elaborate than firing away a singe shot...

/Anders

[eronald]

Erik,

 For completeness you should add in the Cone sensor made by Evolution 
 
Edmund

Hi,

The Bayer pattern has a pair of green pixels for each blue and red pixel. Those greens may or may not be different. I have the impression that raw converters have provisions for handling different G1 and G2 channels.

Sony had some models having green and "emerald" as G1 and G2, but I think the concept has failed.



CFA response curves. I don't know if those contain sensor spectral response and IR filters. Here is the link.

#2 Is Kodak CFA
#3 Is Philips (may be similar to DALSA, as DALSA has taken over Philips CCD development"
#5 Is Foven
#6 Is Sony Super HADII

Best regards
Erik

[ErikKaffehr]

Hi Edmund,

It is number one...

Erik

Erik,

 For completeness you should add in the Cone sensor made by Evolution 
 
Edmund

[Theodoros]

Suggestions on how things may improve on "true colour" capturing are of course very interesting... but i think the O/P is asking for current solutions as to what can be achieved with available equipment.
I guess the single shot solution some suggested has been proved that is to be taken with some salt as being adequate... To my surprise though, there was no talking about lighting, or experience shared on how different lighting affects the process. The Cruse scanner is using fluorescent with high CRI (>97) valves at 45' for the job.
IMO, the Cruse is a good device, but can be beaten by a well thought and executed photographic process out of an experienced specialist photographer.

[[email protected]]

Theodoros
I tried to talk about the lighting...... but maybe my comment was not interesting
or out of topic.
Ben

In my opinion the light is one of the biggest problem to reach accurate color.

Of course, the sensor is important, but on a certain sensor level other problems become more important.
Out of my 12 Bron G 3200 I don't have 4 lamps with an identical flash color. They are all a little bit different.
All have the same protecting glass, all powered by Grafit A4 packs, but I think the age of the flash tube
determines.
Even the polarisation filters I use for certain pictures in front of my lamps change the colors a little bit.

BTW: When the floor in a studio hasn't the same color as the ceiling, it's not worth talking about the sensor of the camera......

[Theodoros]

Theodoros
I tried to talk about the lighting...... but maybe my comment was not interesting
or out of topic.
Ben

It is neither out of topic, nor out of interest... The thing is that especially with multishot sensors, constant lighting temperature is one of the fundamental requirements. You'll rarely find satisfactory results with strobes... In fact, if strobe use is combined with single shot capturing, there will be different interpolated processing from the processor between shots.... Strobes and single shot combined, maybe (just) acceptable for other still life work (jewelry, food, ...etc) where color is tuned to the likes of the recipient, it may also be applicable with statues or other art capturing that is not as important as far as color accuracy is concerned (though multishot is still much superior for all still life work). But for paintings, you'll find that the best photographers that are highly praised for their work. are not using strobes....

[torger]

Fun thing, here's a new multishot camera in m4/3 format(!):

http://www.imaging-resource.com/PRODS/olympus-e-m5-ii/olympus-e-m5-iiA.HTM

[[email protected]]

It is neither out of topic, nor out of interest... The thing is that especially with multishot sensors, constant lighting temperature is one of the fundamental requirements. You'll rarely find satisfactory results with strobes... In fact, if strobe use is combined with single shot capturing, there will be different interpolated processing from the processor between shots.... Strobes and single shot combined, maybe (just) acceptable for other still life work (jewelry, food, ...etc) where color is tuned to the likes of the recipient, it may also be applicable with statues or other art capturing that is not as important as far as color accuracy is concerned (though multishot is still much superior for all still life work). But for paintings, you'll find that the best photographers that are highly praised for their work. are not using strobes....

Thanks, Theodoros
I didn't know that. With my work normally the print office and/or the graphic artist also have influence on the color,
so spending too much time on that would be a waste....
When I don't misunderstand, you say that the strobes are may be consistent but the interpolation of the light from the back can differ?
(Bron + Profoto praise themselves for absolute color accuracy with their most expensive power packs.) 
How about polarization filters in front of the light? I use them for old oil paintings or paintings behind glass to fight unwanted reflections.
Is this the way to go?

[bwphotog]

I just noticed there was an old similar discussion in the forums on using a Gigapan ($900) system for this, lots of good discussion to learn from there - notably the issue of using a panoramic head and panning/tilting for tile detail shots can distort pixels and leave some sections of the art out of focus (unless autofocus is used?). Taking flat direct photos by moving the camera or piece of art for each tile photo seems to be a "more perfect" yet also "more difficult and expensive" process. The detail from these Gigapixel images is amazing though!

  http://www.luminous-landscape.com/forum/index.php?topic=43793.40

I just added the following note:

Follow up to note that people are using the Gigapan system to document fine art:
http://gigapan.com/cms/about/success-stories-capturing-a-moment-in-time

The Gigapan EPIC Pro seems to be in use by the Cultural Heritage Science Open Source (CHSOS) organization in and around Italy to record the old artwork there and elsewhere. They teach classes on all of their techniques to train other art preservationists. http://chsopensource.org

Also the comparable Dr. Clauss systems were also highly recommended, and much more expensive:

Perfect Art Reproductions with RODEON Multi-Head solution
http://dr-clauss.de/en/applications/repro

[eronald]

35mm multishot might be good enough for some uses

http://www.dpreview.com/articles/9437515928/olympus-announces-om-d-e-m5-ii-with-40mp-high-resolution-mode

The world is changing.

Edmund

[Theodoros]

35mm multishot might be good enough for some uses

http://www.dpreview.com/articles/9437515928/olympus-announces-om-d-e-m5-ii-with-40mp-high-resolution-mode

The world is changing.

Edmund

I believe Torger posted the same earlier.... Olympus method applies only on resolution increase which is the minor benefit with multishot... It still uses interpolation to "produce" colour.

[torger]

The Olympus is a m43 (17.3x13mm sensor), not a 135 camera. It's shoots 8 shots with 0.5 pixel steps so of course you get improved colors too, not only resolution, but yes luminance resolution increase is larger than color resolution increase, but afaik the moire issues are gone. Here's a video showing in which orders shots are made https://www.youtube.com/watch?v=nejL7ZMPmUI

[Theodoros]

Thanks, Theodoros
I didn't know that. With my work normally the print office and/or the graphic artist also have influence on the color,
so spending too much time on that would be a waste....
When I don't misunderstand, you say that the strobes are may be consistent but the interpolation of the light from the back can differ?
(Bron + Profoto praise themselves for absolute color accuracy with their most expensive power packs.)  
How about polarization filters in front of the light? I use them for old oil paintings or paintings behind glass to fight unwanted reflections.
Is this the way to go?

Strobes are never consistent... even the best ones (Bron, profoto) can differ up to 100 (+/- 50) degrees of temperature between shots. OTOH, if colour balance is interpolated, different subject can produce a different balance between colours that doesn't (exactly) obey the same colour profile as before. Polarising filters on both lighting and/or lens are only used to control unwanted reflection (another part with painting reproduction that demands high skills to avoid), their use is sometimes necessary even if there is a "soft room" that has been set up to enclose the subject within, depending on the kind of varnish and the reflectability of it.

As far as output profile is concerned, my opinion is that there should be a perfectly calibrated monitor that should be used. My suggestion is that the monitor should be calibrated so that two different materials (a fine art matte paper and a high quality canvas) print consistently when they are printed on the same printer (automatic profile on the printer for the material used). There is no point in choosing more than two particular materials, as material profiling that are supplied from makers for each different printer are not as consisted as they should be. Personally, I use Hannemule Daguerre and Epson enhanced matte paper, both on MK black on Epson 9900 printer. Then, one can assure that if a printing lab or Offset are using their automatic profiles, results will be as accurate as possible! If they aren't, ...then it's not his fault and he can prove his case by providing his own prints for comparison.

I cannot help as to mention once more that Sinar's multishot backs, are in a class of their own for setting up a capture profile for the back and that they can output a Fogra 27 or 39 CMYK profile ready to be used on offset printing. To sum it up, lighting is very important for art reproduction, but if one uses an interpolation colour back/camera, perfect lighting isn't going to save things. There has to be a multishot back for the capture and then one has to work as to maximise the result with lighting.