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

[ErikKaffehr]

Hi,

Thanks for the info! Could you tell us a little bit on colour reproduction, like do you use a specific colour profile or are you tweaking individual colours for correct reproduction?

Best regards
Erik

Love this subject! Great reading thru all the posts, good info for anyone just getting involved in this biz for sure
I have dedicated a large portion of my life and experience to art repro capture, and run a very successful group of studios one of which is entirely dedicated to this very subject. Not our bread and butter, but enough to employ 3 highly skilled and experienced craftsmen, hundreds of thousands of dollars worth of equipment and studio space.
My firm belief is that info gleaned from art repro capture research is immediately applied to our commercial work and vice versa.

I don't think the initial question "Best camera setup for art repro...?" is a serious one.
Analogous to a  question like "what is the best car for driving on mountain roads?"
-Are we racing? Is speed an issue? whats our budget? Will there be weather? who are our passengers? Is comfort an issue? Are they paying for the trip? will there be sand, or water, or rocks on our path? How much gas do we need?...etc

We have many clients for art repro capture and reproduction but most are private collections, large museums, some really high end well heeled artists, and insurance companies.
One of the posts earlier in this discussion stated that $1000 for a single image was very high. I can assure you that $1000+/image is not considered crazy expensive, in fact, its fairly commonplace. We just completed a single capture for a museum reproduction for over $10k. Mind you we are not just one photographer with a bitchin camera and a computer, but a highly skilled team of specialists who have all been doing this for major museums and artists since the 70’s and film days. So though $10K might seem like a bundle, it is very small compared to value of the painting we were hired to capture...and we were hired to do it twice! (before and after restoration) The restoration work on this single painting alone cost over $100K. Just the scaffolding costs were close to that.

We have most if not all of the gear mentioned in this thread but over 50% of the time we have found that a highly skilled and experienced operator who knows how to properly light, expose, post-proc using C1 with an 80MP CCD MFD back and corresponding lenses is our go to system.

Recently (Fall 2014) we went up against a skilled competitor who uses the Sinar system and we bid using our Credo80 system and our team. We both had to show accuracy and competence on a smaller piece about 6ft x 9ft (valued at $12 million!)
In the end, we won the contract based on a variety of factors but our color accuracy was said to be the most important factor.
So although, on paper, I agree that a true-color type system with multi-pop or scan back should be more accurate, there are many more factors that go into the equation and should be considered.

For instance, we tested out the Hassy 200MP MS and though on paper it looked like a winner, in real world testing, the 6pop multistep tech did not work so good, suffered from misregistration, and was extremely susceptible to even the most minute vibrations. Even when bolted to a concrete floor we had problems with it.

[egor]

I’ll try
For printing its a mixed bag as well, and factors are if we are printing in-house or on someone else’s equipment.
We print profile using a combination XRITE ProfileMaker hardware/software and sometimes older legacy Heidelberg PrintOpen software comboned with tweaking of profiles using ORIS ColorTuner software
Basically it depends on whether we are targeting a standard web or sheetfed press that uses AM screening or an inkjet FM press that uses stochastic (FM screening) and High Fidelity ink sets.
In either case, we profile/fingerprint the press. In-house we print on a Epson 11880

The camera sensors are a different story. We have tried using a wide variety of profiling software and hardware but have found the best results for MFD and scan backs is using certain canned profiles and tweaking results directly in C1 or ACR.
The canons and small dslrs are easy to profile but the big guns just aren’t.

We output drum scanned files into Linocolor CieLab, and most camera captures into Adobe 1998 or ProPhoto RGB colorspace. Sometimes 16bit (when there is a reason to take highlight detail or shadow detail that needs bumping.

In the end, we do our best to profile thru the whole process but it really is just a start point because each job is different and has different colors and densities that are targeted. We almost always go 3-4 rounds of tweaking with a high dollar job for reproduction even with all the profiling and calibrated systems, its just never an “exact science” it seems.

We had a job that required a huge capture and then reproduction on a single piece of canvas over 12ft x 15ft. We had to take a machine that ordinarily prints billboard signage that size, and re-calibrate it to our giclée print standards. That took 2 months! Then we had to have a single roll of ink jet receptive canvas made just for this machine and its special latex ink sets and profile all over again. That took an extra month. There were so many variables you just wouldn’t believe it, but in the end, came out great and the client was extremely happy. One for the record books because much of what we did from capture to finished stretched piece required adventure into uncharted territory.
The attached iPhone shots of what we did don't really do it justice. But gives some idea of scale. The 2nd iPhone shot is of us doing the actual capture. The 1st image is over 3months after the capture when the original and our reproduction were , for a brief moment, side by side as they took down the original and put up our giclée reproduction.
The capture was done using 6 frames on a Sinar 4x5 using our venerable Credo80 back and C1 LCC frames and color profile targets for each and every frame/exposure. Each exposure was a minute suspended over 25 ft in the air on rickety scaffolding. I only bring it up because the capture part of the job was every bit as hard as the print part, but we only had one shot at it in one evening and no second chances! Pretty cool, huh?

[ErikKaffehr]

Hi,

Thanks for sharing! Some good insight in high level reproduction work!

Some small reflections on your writing:

- No real surprise that tweaking is needed. A profile is always generic and there will always be metameric issues with natural pigments, I guess. There was a very entertaining interview here at LuLa with Ray Maxwell, who is a colour scientist, I think. He sad, that for correct reproduction a profile is needed built patches using the same pigments as in the original. That would reproduce the painting ideally in the same light as used for taking the picture.

- Something that came as a bit of surprise that 8-bit colour can be used with wide gamuts. It may be that we use 16 bits a bit habitually, but using a wide colour space in 8-bits may not be the ultimate sin I would have believed.

Best regards
Erik

I’ll try
For printing its a mixed bag as well, and factors are if we are pronting in-house or on someone else’s equipment.
We print profile using a combination XRITE ProfileMaker hardware/software and sometimes older legacy Heidelberg PrintOpen software comboned with tweaking of profiles using ORIS ColorTuner software
Basically it depends on whether we are targeting a standard web or sheetfed press that uses AM screening or an inkjet FM press that uses stochastic (FM screening) and High Fidelity ink sets.
In either case, we profile/fingerprint the press. In-house we print on a Epson 11880

The camera sensors are a different story. We have tried using a wide variety of profiling software and hardware but have found the best results ofr MFD and scan backs using certain canned profiles and tweaking results directly in C1 or ACR. The canons and small dslrs are easy to profile but the big guns just aren’t.
We output drum scanned files into Linocolor CieLab, and most camera captures into Adobe 1998 or ProPhoto RGB colorspace. Sometimes 16bit (when there is a reason to like highlight detail or shadow detail that needs bumping.

In the end, we do our best to profile thru the whole process but it really is just a start point because each job is different and has different colors and densities that are targeted. We almost always go 3-4 rounds of tweaking with a high dollar job for reproduction even with all the profiling and calibrated systems, its just never an “exact science” it seems.

We had a job that required a huge capture and then reproduction on a single piece of canvas over 12ft x 15ft. We had to take a machine that ordinarily prints signage that size and re-calibrate it to giclee print standards. That took 2 months! Then we had to have a single roll of ink jet receptive canvas made just for this machine and its special latex ink sets and profile all over again. That took an extra month. There were so many variables you just wouldn’t believe it, but in the end, came out great and the client was extremely happy. One for the record books because much of what we did from capture to finished stretched piece required adventure into uncharted territory.

[egor]

He sad, that for correct reproduction a profile is needed built patches using the same pigments as in the original. That would reproduce the painting ideally in the same light as used for taking the picture.

Hahaha...My favorite thing to tell clients with super high expectations of reproductions is: "If you want a perfect copy, then let us make two copies and throw away the original!"

- Something that came as a bit of surprise that 8-bit colour can be used with wide gamuts. It may be that we use 16 bits a bit habitually, but using a wide colour space in 8-bits may not be the ultimate sin I would have believed.

As previously stated, I am less interested in what "should be better" and test to see what actually works more often than not. 16bit has its uses in some cases but is rarely needed, imo. It can be a crutch for poor exposure or proc skills like some friends of mine always ranting about needing more and more dynamic range so they can pull detail out of under or overexposed areas. I think it can be of help in many situations but in a studio controlled lighting and exposure situation I don't think its needed. And it uses up a lot of disc space and proc power. Now the conversion from 16bit to 8bit is very important I have found, but thats another variable...

[ErikKaffehr]

Hi,

Comments, see below!

Best regards
Erik

Hahaha...My favorite thing to tell clients with super high expectations of reproductions is: "If you want a perfect copy, then let us make two copies and throw away the original!"

Erik: I really liked this one!

As previously stated, I am less interested in what "should be better" and test to see what actually works more often than not. 16bit has its uses in some cases but is rarely needed, imo. It can be a crutch for poor exposure or proc skills like some friends of mine always ranting about needing more and more dynamic range so they can pull detail out of under or overexposed areas. I think it can be of help in many situations but in a studio controlled lighting and exposure situation I don't think its needed. And it uses up a lot of disc space and proc power. Now the conversion from 16bit to 8bit is very important I have found, but thats another variable...

Erik: Conventional wisdom, or at least my reading of it, says that more bits are needed to represent a larger colour space as each change of bit represents a larger shift in hue and density. I got this from one of the writings of Karl Lang, cannot find it right now, of course. Well, if 16 bits are not needed for ProPhoto RGB than we can just work with smaller files, a good thing. 

[egor]

I feel it depends upon output. 8bit vs 16bit or Lab vs sRGB vs ProPhoto RGB vs Adobe RGB gamuts are all well and fine if your output is RGB transmissive light like a well calibrated computer monitor. But if you are printing something, It all becomes subtractive light model for reflective and standard print gamuts of CMYK or Hex or some other variation. Basically, imo, you are taking a color range the size of Nebraska and fitting it into a coke can. So unless you have some detail that really needs exaggeration so it shows on a print somewhere in this process, what is the point?
OTOH, if you are shooting for detective, forgery, or restoration work or something that requires serious pixel peeping on a monitor by experts, it can be totally justified.

[Theodoros]

Love this subject! Great reading thru all the posts, good info for anyone just getting involved in this biz for sure
I have dedicated a large portion of my life and experience to art repro capture, and run a very successful group of studios one of which is entirely dedicated to this very subject. Not our bread and butter, but enough to employ 3 highly skilled and experienced craftsmen, hundreds of thousands of dollars worth of equipment and studio space.
My firm belief is that info gleaned from art repro capture research is immediately applied to our commercial work and vice versa.

I don't think the initial question "Best camera setup for art repro...?" is a serious one.
Analogous to a  question like "what is the best car for driving on mountain roads?"
-Are we racing? Is speed an issue? whats our budget? Will there be weather? who are our passengers? Is comfort an issue? Are they paying for the trip? will there be sand, or water, or rocks on our path? How much gas do we need?...etc

We have many clients for art repro capture and reproduction but most are private collections, large museums, some really high end well heeled artists, and insurance companies.
One of the posts earlier in this discussion stated that $1000 for a single image was very high. I can assure you that $1000+/image is not considered crazy expensive, in fact, its fairly commonplace. We just completed a single capture for a museum reproduction for over $10k. Mind you we are not just one photographer with a bitchin camera and a computer, but a highly skilled team of specialists who have all been doing this for major museums and artists since the 70’s and film days. So though $10K might seem like a bundle, it is very small compared to value of the painting we were hired to capture...and we were hired to do it twice! (before and after restoration) The restoration work on this single painting alone cost over $100K. Just the scaffolding costs were close to that.

We have most if not all of the gear mentioned in this thread but over 50% of the time we have found that a highly skilled and experienced operator who knows how to properly light, expose, post-proc using C1 with an 80MP CCD MFD back and corresponding lenses is our go to system.

Recently (Fall 2014) we went up against a skilled competitor who uses the Sinar system and we bid using our Credo80 system and our team. We both had to show accuracy and competence on a smaller piece about 6ft x 9ft (valued at $12 million!)
In the end, we won the contract based on a variety of factors but our color accuracy was said to be the most important factor.
So although, on paper, I agree that a true-color type system with multi-pop or scan back should be more accurate, there are many more factors that go into the equation and should be considered.

For instance, we tested out the Hassy 200MP MS and though on paper it looked like a winner, in real world testing, the 6pop multistep tech did not work so good, suffered from misregistration, and was extremely susceptible to even the most minute vibrations. Even when bolted to a concrete floor we had problems with it.

It must have been a faulty 200MS... did you check it for malfunctioning?   

[egor]

Hi Theodoros
Yes, we considered that possibility at the time. I felt bad for our Hasselblad rep at the time (who was fantastic).
In the end, there were other system-wide implementation considerations that led us to choose PhaseOne/Leaf over the Hasselblad solution.
I couldn't help but feel that if that problem occurred with a demo model and rep right there, that it could, and would, occur to us on the job at some point.
Because of the nature of how this work is performed in real-world scenarios, I did not see enough advantage over the existing scan back solution to justify the purchase; whereas I did with the PhaseOne/Leaf systems. That having been said, the Hasselblad MS is a hell of a nice machine and system and it was a close call.

[ErikKaffehr]

Hi,

I guess I see your point. I made a small experiment, taking one of my P45+ shots of an IT8 target in Prophoto RGB 16 bits. I opened the image in Photoshop CS and saved as a TIFF, than I converted it to 8-bits and back to 16-bits and save that image. After that I read both images using the scanin program from Argyll CMS and compared the resulting CGATS files using BabelColors Patchtool.

The errors resulting from the roundtrip to 8 bits were ignorable (DE*) <= 0.55 and all differences occured in the darks.

Learning all the time...

Best regards
Erik

I feel it depends upon output. 8bit vs 16bit or Lab vs sRGB vs ProPhoto RGB vs Adobe RGB gamuts are all well and fine if your output is RGB transmissive light like a well calibrated computer monitor. But if you are printing something, It all becomes subtractive light model for reflective and standard print gamuts of CMYK or Hex or some other variation. Basically, imo, you are taking a color range the size of Nebraska and fitting it into a coke can. So unless you have some detail that really needs exaggeration so it shows on a print somewhere in this process, what is the point?
OTOH, if you are shooting for detective, forgery, or restoration work or something that requires serious pixel peeping on a monitor by experts, it can be totally justified.

[Theodoros]

Hi Theodoros
Yes, we considered that possibility at the time. I felt bad for our Hasselblad rep at the time (who was fantastic).
In the end, there were other system-wide implementation considerations that led us to choose PhaseOne/Leaf over the Hasselblad solution.
I couldn't help but feel that if that problem occurred with a demo model and rep right there, that it could, and would, occur to us on the job at some point.
Because of the nature of how this work is performed in real-world scenarios, I did not see enough advantage over the existing scan back solution to justify the purchase; whereas I did with the PhaseOne/Leaf systems. That having been said, the Hasselblad MS is a hell of a nice machine and system and it was a close call.

 I guess with so much reputation in your work, you also own a Sinarback eXact.... I wonder why you didn't choose that instead... You see... as I said before, after Yair refused me to compare the Leaf 12II (which you choose to use) with my (at the dates) 528c, I did try the Aptus out of another owner (who had the back after the original owner died at the age of 42 - Yair knows the case) and I must say, although lenses where different, the results where so much in favor of the 528c that even a "blind" man could see... OTOH, the Sinarback 54H is even better than the 528c....

   I wonder... surely if you do such a great  job, have all these studios and charge all that money, you surely have a web page... what is it? You may also contact me privately via PM with your mail address... We are both at the same subject with not much (serious) competition around and knowledge on that (very difficult) part of photography is developing all the time...  We may benefit both after we get to know each other...  


P.S.... Aaaah! ...welcome to the forum!

[egor]

I guess with so much reputation in your work, you also own a Sinarback eXact.... I wonder why you didn't choose that instead... You see... as I said before, after Yair refused me to compare the Leaf 12II (which you choose to use) with my (at the dates) 528c, I did try the Aptus out of another owner (who had the back after the original owner died at the age of 42 - Yair knows the case) and I must say, although lenses where different, the results where so much in favor of the 528c that even a "blind" man could see... OTOH, the Sinarback 54H is even better than the 528c....

   I wonder... surely if you do such a great  job, have all these studios and charge all that money, you surely have a web page... what is it? You may also contact me privately via PM with your mail address... We are both at the same subject with not much (serious) competition around and knowledge on that (very difficult) part of photography is developing all the time...  We may benefit both after we get to know each other...  


P.S.... Aaaah! ...welcome to the forum!

Sorry, Theodoros, you sound kind of defensive, and I am not a fan of peeing contests. Forgive me if I am wrong in that assessment.
If you disagree with anything I have said, thats fine. I'm good with that. I am just a working pro who's been doing this a long time.
Thanks for the welcome to the forum! I joined recently to answer a question for someone else who asked me to. I saw this subject and thought I'd chime in. I have no interest in debating what and who is better at any particular aspect of art repro. There are many factors that determine success in this field.
Yes, I have multiple web pages, but no interest in advertising here.
I am; however always interested in how other people do the same job, as well as new ways to accomplish the same goals.
From what I gather, you like to use the Sinar 54H and have had good results with it that you feel are superior to a Leaf Aptus (why the Aptus?, btw?, I never mentioned using an Aptus. I use a Credo80....or is that what you are challenging Yair with?)
My retort is simple. A scan back like a Betterlight or PhaseOne scan back, or multi shot MFD are marvelous tools for art repro capture. On paper, they are better color and resolution than single pop modern single capture backs like the Credo and IQ250 CMOS backs. But in my real world experience it is more about skill, lighting, knowledge of the client's goals, budget and experience. The "big picture" pardon the pun.Just my 2 cents.

[Theodoros]

Sorry, Theodoros, you sound kind of defensive, and I am not a fan of peeing contests. Forgive me if I am wrong in that assessment.
If you disagree with anything I have said, thats fine. I'm good with that. I am just a working pro who's been doing this a long time.
Thanks for the welcome to the forum! I joined recently to answer a question for someone else who asked me to. I saw this subject and thought I'd chime in. I have no interest in debating what and who is better at any particular aspect of art repro. There are many factors that determine success in this field.
Yes, I have multiple web pages, but no interest in advertising here.
I am; however always interested in how other people do the same job, as well as new ways to accomplish the same goals.
From what I gather, you like to use the Sinar 54H and have had good results with it that you feel are superior to a Leaf Aptus (why the Aptus?, btw?, I never mentioned using an Aptus. I use a Credo80....or is that what you are challenging Yair with?)
My retort is simple. A scan back like a Betterlight or PhaseOne scan back, or multi shot MFD are marvelous tools for art repro capture. On paper, they are better color and resolution than single pop modern single capture backs like the Credo and IQ250 CMOS backs. But in my real world experience it is more about skill, lighting, knowledge of the client's goals, budget and experience. The "big picture" pardon the pun.Just my 2 cents.

No defense at all (how did this come to you?)... I don't mind at all to disagree with people, I'm still interested to see your page... I'm only doing "small jobs" myself (I'm currently doing the Academy of Athens (5200 pieces), then I have to do the Gennadios library (2400 pieces) and then the Acropolis Museum this year) in a small country, that its monuments are of minor importance for world culture... So what do I know?

But what I do, I have no problem in posting (because ....I do it) ...see? I do believe that you have nothing to learn from me, but please, let me have a taste on that page of yours... If you don't want to share it with the rest on the forum, please do it privately... I promise not to reveal it.... It's the word of a pro, that finds the Sinarback 54H far superior than Leaf's 12ii back... (which I believe has minor IQ differences than Credo?) ....regards, Theodoros.

[yaya]

Just wondering if any of the experts on here is coming to the 2D/3D conference at the Rijkmuseum in April? We could all sit around the dinner table and discuss this subject until we drop...

And welcome to LuLa Eric! Hope you brought your helmet along?

BR

Yair

[Bart_van_der_Wolf]

Just wondering if any of the experts on here is coming to the 2D/3D conference at the Rijkmuseum in April?

For those interested:
https://www.rijksmuseum.nl/en/2and3dphotography

Cheers,
Bart

[egor]

Just wondering if any of the experts on here is coming to the 2D/3D conference at the Rijkmuseum in April? We could all sit around the dinner table and discuss this subject until we drop...

And welcome to LuLa Eric! Hope you brought your helmet along?

BR

Yair

Thanks, Yair
I would absolutely love to attend such a conference and am very fond of Amsterdam!
Unfortunately, won't be able to attend for usual reasons like work, family, and budget.
I'd send one of my assistants but she'd just hang in the cafes and get stoned all day....She told me as much.
Unlike some others here, I seem to have my hands full with only a few hundred art captures per year
Our next fun project is some cave art with no available electricity or platform to shoot from. Should be fun!

Thanks for the welcome to LuLa, I only have my bicycle helmet, I hope it is enough

e

[ErikKaffehr]

Hi Egor,

I also wanted to thank you for sharing your experience! Carry on, Sir!

Best regards
Erik

Thanks, Yair
I would absolutely love to attend such a conference and am very fond of Amsterdam!
Unfortunately, won't be able to attend for usual reasons like work, family, and budget.
I'd send one of my assistants but she'd just hang in the cafes and get stoned all day....She told me as much.
Unlike some others here, I seem to have my hands full with only a few hundred art captures per year
Our next fun project is some cave art with no available electricity or platform to shoot from. Should be fun!

Thanks for the welcome to LuLa, I only have my bicycle helmet, I hope it is enough

e

[engelye]

Hello,

I constructed a combination to reproduce the artworks which are oilpaint on wood whose dimensions are changed between 10x15 cm and 100x200 cm. What do you think, is this set up sufficient? Is there any equipment that I have to add or change? Nevertheless, is there any equipment that incompatible with each other?

Sinar rePro RC or Sinar p3-df

Apo-Sironar digital 55 mm f/4,5 with eShutter + polarizer filter + UV filter

Betterlight Super 8K-HS Digital Scanning Back

North Light HID Copy Lights (2 X 600 + 2 X 300) + polarizer filter

Sinar parallel (for parallel alignment)

X-Rite Digital ColorChecker SG Card

sekonic l-758dr light meter

Sekonic Prodigi Color C-500 Color Meter

Betterlight  ViewFinder Digital Scanning Camera Software

EquaLight 3.1.1

Mac Pro + Eizo CG276

Thank you

[engelye]

Hello,


I constructed a combination to reproduce the artworks which are oilpaint on wood whose dimensions are changed between 10x15 cm and 100x200 cm. What do you think, is this set up sufficient? Is there any equipment that I have to add or change? Nevertheless, is there any equipment that incompatible with each other?

Sinar rePro RC or Sinar p3-df

Apo-Sironar digital 55 mm f/4,5 with eShutter + polarizer filter + UV filter

Betterlight Super 8K-HS Digital Scanning Back

North Light HID Copy Lights (2 X 600 + 2 X 300) + polarizer filter

Sinar parallel (for parallel alignment)

X-Rite Digital ColorChecker SG Card

sekonic l-758dr light meter

Sekonic Prodigi Color C-500 Color Meter

Betterlight  ViewFinder Digital Scanning Camera Software

EquaLight 3.1.1

Mac Pro + Eizo CG276

Best regards

[Chris Valites]

The colorchecker is a good choice, but many people we work with have built in color swatches directly into the table/shooting surface. That way you can have the color reference in every frame, and then just crop it out later. Saves you the hassle of having to get out a passport or worry about that.

[Theodoros]

Hello,


I constructed a combination to reproduce the artworks which are oilpaint on wood whose dimensions are changed between 10x15 cm and 100x200 cm. What do you think, is this set up sufficient? Is there any equipment that I have to add or change? Nevertheless, is there any equipment that incompatible with each other?

Sinar rePro RC or Sinar p3-df

Apo-Sironar digital 55 mm f/4,5 with eShutter + polarizer filter + UV filter

Betterlight Super 8K-HS Digital Scanning Back

North Light HID Copy Lights (2 X 600 + 2 X 300) + polarizer filter

Sinar parallel (for parallel alignment)

X-Rite Digital ColorChecker SG Card

sekonic l-758dr light meter

Sekonic Prodigi Color C-500 Color Meter

Betterlight  ViewFinder Digital Scanning Camera Software

EquaLight 3.1.1

Mac Pro + Eizo CG276

Best regards

If you insist on the Betterlight, it is best to use a 5x4 camera, with the P3 (or the repro) it is best to use one of the multishot  Sinarbacks... I wouldn't use a WA lens either...