The GetDPI Photography Forum
Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!
Difficulty Dupliating Flat Artwork
- Thread starter Craig Stocks
- Start date
Hello Craig,
I see two differences between your workflow and mine which I think might make a difference to your results.
I create my profiles using the SG Colorchecker. To create the profiles I'm using Phocus software as we use Hasselblads in my studio.
Using the SG color checker I get a delta E of 1.48 on average and a Max delta E of 4.4. This works pretty well for me.
Also, when I photograph originals I include a mini passport color checker in the photo.
When I bring it into Photoshop I open a curves layer, set the highlight eyedropper to 240 and click the white patch of the color checker then I set the shadow eye dropper to 35 and click the black patch of the color checker. This usually gives me a pretty accurate exposure and contrast range.
Hope this can be of some help to you.
Mike
Profiles Studio
Philadelphia
I see two differences between your workflow and mine which I think might make a difference to your results.
I create my profiles using the SG Colorchecker. To create the profiles I'm using Phocus software as we use Hasselblads in my studio.
Using the SG color checker I get a delta E of 1.48 on average and a Max delta E of 4.4. This works pretty well for me.
Also, when I photograph originals I include a mini passport color checker in the photo.
When I bring it into Photoshop I open a curves layer, set the highlight eyedropper to 240 and click the white patch of the color checker then I set the shadow eye dropper to 35 and click the black patch of the color checker. This usually gives me a pretty accurate exposure and contrast range.
Hope this can be of some help to you.
Mike
Profiles Studio
Philadelphia
Craig Stocks
Well-known member
I have and I get the same result, the ColorChecker photographs great but the artwork shows too little saturation in the yellows. I've tried sunlight, LEDs, and strobes. So far the strobes with UV and polarizing filters work best. (Some of the watercolor has optical brighteners that glow under UV.)
Wayne Fox
Workshop Member
I think that’s pretty normal for trying to repro water colors. I think the light colors get diluted by the light penetrating the paper which sort of lights it up from the back. Even if you put it on black (which seems to help sometime), there’s still a lot of light bouncing around in the paper base. Also the colors are so light and delicate they are toward the fringe areas of the gamut of sensor as well as of the paper.
I don’t think I’ve ever found a water color that didn’t take some work.
beano_z
Active member
I really haven't read all the replies in detail, but things that come to mind are:
1) Use an incandescent light source, such as a regular incandescent bulb or halogen, the LED's that usually are sold (even professionally) have a color rendering of less than 90%, especially missing the reds, the light spectrum simply isn't there to reproduce all colours. But I think the flash you're using should be fine as the flash tube is also an incandescent light source.
2) Try to expose more towards the left, as you'll loose saturation the more you expose to the right. As I can see from your example, the dynamic ranges of your subjects are quite limited, so even though I'm not completely sure about the technicalities of this approach, I still think given the circumstances, it's worth a try at least.
1) Use an incandescent light source, such as a regular incandescent bulb or halogen, the LED's that usually are sold (even professionally) have a color rendering of less than 90%, especially missing the reds, the light spectrum simply isn't there to reproduce all colours. But I think the flash you're using should be fine as the flash tube is also an incandescent light source.
2) Try to expose more towards the left, as you'll loose saturation the more you expose to the right. As I can see from your example, the dynamic ranges of your subjects are quite limited, so even though I'm not completely sure about the technicalities of this approach, I still think given the circumstances, it's worth a try at least.
dougpeterson
Workshop Member
(1) should really say “use a light with a high CRI/CQS and non-spiky spectrum”. While most LEDs have a very middling spectral quality there are LEDs with very high quality spectrums. The DT Photons, for example, are individually spectrally measured at the factory with a minimum value of 98 CRI and 98 CQS.
I only do this kind of work occasionally, but those are the suggestions that came to mind to me as well, especially after reading this:
Re: Using natural light/long exposure (assuming cool skylight?)
Christopher Campbell
New member
I am a successful artist, and a very serious photographer, and this is a subject I have worked on for many years, and on which I will be speaking at the upcoming Rijksmuseum 2and3D Photography symposium in June. As others have indicated, it's a very deep subject, but I will list a few critical points from my own experience.
Lens. A lens with a relatively flat field (controlled field curvature) is essential, and I understand that the Phase One 120mm macro is an excellent choice.
Light. Light used for copy work must be even, color constant, and repeatable, which rules out daylight. Tungsten lamps emit very few short wavelengths, and so can't properly differentiate subtle differences between various blues on digital sensors. That leaves either flash, or LEDs. I use the Paul C. Buff Einstein E640s, as they have a color constant mode (maintain the same color temperature as the output level is changed), but the AlienBees should be fine. LED sources that meet the highest TM-30-18 standards are also excellent, but still roughly 10x the cost of xenon flash. 45° is much too high an angle of incidence for the flash beam. I highly recommend purchasing a copy of the “AIC Guide to Digital Photography and Conservation,” which specifically cautions against 45°, and recommends 25°, which is the angle I use.
Polarization. Polarized light can be very effective in removing specular reflections, but it can also be problematic in the dark values. I am currently using Rosco #7300 linear polarizing filters on my flashes, setting their rotational angles in relation to a fixed Zeiss T* circular polarizer on the lens, but then dialing back the lens polarizer (called an analyzer) to 70°, which lessens the effect of the full 90°.
Flat-field correction. I use four flashes to produce a large, even field of illumination, but copy work can be done with two lights if you use flat-field correction. The LCC function in Capture One works extremely well, as does the flat-field correction available in Adobe Lightroom.
Parallel planes. One can't rely on depth of field to produce optimal sharpness in copying flat originals without diffraction limiting, so tools for establishing parallelism such as zig-align mirrors or (better) the Versalab Parallel laser are essential. I also use the Versalab — set at a distance — to set the parallelism of the standards on an Arca-Swiss M-line two (mf) camera.
Exposure. I use RawDigger to determine the point of absolute clipping for a digital sensor, and then set my working exposure level 1/3 of an f-stop below that point to optimize the signal-to-noise ratio (advice from Iliah Borg). My primary camera is a Phase One IQ3 100MP Trichromatic, and it has a full 16 stops of dynamic range (the maximum dynamic range I have measured in one of my paintings is about ten stops). I make my captures tethered to a MacBook Pro running Capture One, and find the "Focus Meter" to be superb for finding the point of optimal contrast, well beyond the ability of my eyes alone. I use the L*a*b readouts in Capture One to keep track of my exposure level, placing the diffuse white of the PTFE patch at L* 95 when determining exposure.
Neutral reference. I have measured many neutral reference patches (ISA Golden Thread, ColorCheckers of all kinds, WhiBal and QPC cards) and none of them are particularly neutral. The one material that is exceedingly neutral is sintered PTFE (teflon), which can be had in small sheets from Thorlabs. Next best is to wrap a Staedtler Mars white eraser with white PTFE plumber's tape (also advice from Iliah Borg)! The SNI (Spectral Neutrality Index) can be easily measured with a spectrometer such as an i1Pro2 and SpectraShop software from Robin Myers. The SNI for the ColorChecker SG (patch F5) is 73.3, whereas the Thorlabs PTFE is 99.0 on a scale from 1–100.
rmimaging.com
Camera profiling. The camera profile is one of the most critical components in fine art reproduction, and the “Flash – Fine Art Reproduction” profile supplied with all Phase One digital backs is very good. The only thing better is building a camera profile for your particular sensor, with your individual set up. Since the process of camera profiling is essentially using the sensor as a colorimeter to characterize its spectral response, it takes a fair number of color patches to do the job, and this cannot begin to be done adequately with a standard 24 patch ColorChecker. The ColorChecker SG, with 140 patches, is the least-expensive standard target. More refined targets are available from Digital Transitions, and I ended up building my own, using 50+ hues of oil paint in some 270 patches, because I wanted to know quite precisely how the Trichromatic responds to real cobalt blue and cadmium orange.
Profiling software. basICColor input pro 6 is one of the best applications for camera profiling, and it does a beautiful job, such that once I apply my profile and tone curve (in either Lightroom or Capture One), I never feel the need to adjust any individual colors. When the colorimetry of the profile is correct, lightness, hue and chroma are all corrected simultaneously without the need for manual intervention (Full disclosure: I am a basICColor Ambassador, and beta test for the developer, Franz Herbert).
heritage-digitaltransitions.com
The digitization workflow guide that Doug Peterson referenced above is excellent, and highly recommended, especially if you're new to camera profiling and flat-field correction:
heritage-digitaltransitions.com
Tools and tech support for this kind of work are hard to find, but the folks at Chromix are another excellent resource for targets, software and expertise:
Good luck!
Copy set up for flat originals up to 40 x 57" (101 x 148 cm)
\
Using basICColor input to evaluate the quality of a new profile:
Lens. A lens with a relatively flat field (controlled field curvature) is essential, and I understand that the Phase One 120mm macro is an excellent choice.
Light. Light used for copy work must be even, color constant, and repeatable, which rules out daylight. Tungsten lamps emit very few short wavelengths, and so can't properly differentiate subtle differences between various blues on digital sensors. That leaves either flash, or LEDs. I use the Paul C. Buff Einstein E640s, as they have a color constant mode (maintain the same color temperature as the output level is changed), but the AlienBees should be fine. LED sources that meet the highest TM-30-18 standards are also excellent, but still roughly 10x the cost of xenon flash. 45° is much too high an angle of incidence for the flash beam. I highly recommend purchasing a copy of the “AIC Guide to Digital Photography and Conservation,” which specifically cautions against 45°, and recommends 25°, which is the angle I use.
Polarization. Polarized light can be very effective in removing specular reflections, but it can also be problematic in the dark values. I am currently using Rosco #7300 linear polarizing filters on my flashes, setting their rotational angles in relation to a fixed Zeiss T* circular polarizer on the lens, but then dialing back the lens polarizer (called an analyzer) to 70°, which lessens the effect of the full 90°.
Flat-field correction. I use four flashes to produce a large, even field of illumination, but copy work can be done with two lights if you use flat-field correction. The LCC function in Capture One works extremely well, as does the flat-field correction available in Adobe Lightroom.
Parallel planes. One can't rely on depth of field to produce optimal sharpness in copying flat originals without diffraction limiting, so tools for establishing parallelism such as zig-align mirrors or (better) the Versalab Parallel laser are essential. I also use the Versalab — set at a distance — to set the parallelism of the standards on an Arca-Swiss M-line two (mf) camera.
PARALLEL
Versalab Parallel - laser enlarger alignment tool - precision enlarger alignment - darkroom printing - copy alignment - photographic darkroom
www.versalab.com
Exposure. I use RawDigger to determine the point of absolute clipping for a digital sensor, and then set my working exposure level 1/3 of an f-stop below that point to optimize the signal-to-noise ratio (advice from Iliah Borg). My primary camera is a Phase One IQ3 100MP Trichromatic, and it has a full 16 stops of dynamic range (the maximum dynamic range I have measured in one of my paintings is about ten stops). I make my captures tethered to a MacBook Pro running Capture One, and find the "Focus Meter" to be superb for finding the point of optimal contrast, well beyond the ability of my eyes alone. I use the L*a*b readouts in Capture One to keep track of my exposure level, placing the diffuse white of the PTFE patch at L* 95 when determining exposure.
Neutral reference. I have measured many neutral reference patches (ISA Golden Thread, ColorCheckers of all kinds, WhiBal and QPC cards) and none of them are particularly neutral. The one material that is exceedingly neutral is sintered PTFE (teflon), which can be had in small sheets from Thorlabs. Next best is to wrap a Staedtler Mars white eraser with white PTFE plumber's tape (also advice from Iliah Borg)! The SNI (Spectral Neutrality Index) can be easily measured with a spectrometer such as an i1Pro2 and SpectraShop software from Robin Myers. The SNI for the ColorChecker SG (patch F5) is 73.3, whereas the Thorlabs PTFE is 99.0 on a scale from 1–100.
High-Reflectance PTFE Sheets
Thorlabs specializes in the building blocks for laser and fiber optic systems. From optomechanical components to telecom test instrumentation, Thorlabs' extensive manufacturing capabilities allow us to ship high quality, well priced components and devices for next-day delivery. Optomechanics...
www.thorlabs.com
Robin Myers Imaging: SpectraShop™
SpectraShop™ software can measure and analyze spectra, help you examine your color management, provides a low-cost density meter and more! From Robin Myers Imaging.
rmimaging.com
Camera profiling. The camera profile is one of the most critical components in fine art reproduction, and the “Flash – Fine Art Reproduction” profile supplied with all Phase One digital backs is very good. The only thing better is building a camera profile for your particular sensor, with your individual set up. Since the process of camera profiling is essentially using the sensor as a colorimeter to characterize its spectral response, it takes a fair number of color patches to do the job, and this cannot begin to be done adequately with a standard 24 patch ColorChecker. The ColorChecker SG, with 140 patches, is the least-expensive standard target. More refined targets are available from Digital Transitions, and I ended up building my own, using 50+ hues of oil paint in some 270 patches, because I wanted to know quite precisely how the Trichromatic responds to real cobalt blue and cadmium orange.
Profiling software. basICColor input pro 6 is one of the best applications for camera profiling, and it does a beautiful job, such that once I apply my profile and tone curve (in either Lightroom or Capture One), I never feel the need to adjust any individual colors. When the colorimetry of the profile is correct, lightness, hue and chroma are all corrected simultaneously without the need for manual intervention (Full disclosure: I am a basICColor Ambassador, and beta test for the developer, Franz Herbert).
Search results for "basICColor" | DT Heritage
heritage-digitaltransitions.com
The digitization workflow guide that Doug Peterson referenced above is excellent, and highly recommended, especially if you're new to camera profiling and flat-field correction:
Digitization Workflows: Reflective (Print Edition) | DT Heritage
heritage-digitaltransitions.com
Tools and tech support for this kind of work are hard to find, but the folks at Chromix are another excellent resource for targets, software and expertise:
Good luck!
Copy set up for flat originals up to 40 x 57" (101 x 148 cm)
\Using basICColor input to evaluate the quality of a new profile:
Last edited:
Thanks for the breakdown of your process Christopher,
Being less than 1mm thick, what do you mount the PTFE sheet onto (in order to maintain its rigidity yet not interfere with its neutral spectral properties)?
thanks,
Alex
Christopher Campbell
New member
In the image of basICColor input, you can see a variety of targets. The large artist's oil paint target in the center is the one I constructed (oil paint drawdowns on polyester film), and it is flanked by an Image Associates/Golden Thread 2x target above. That's a fairly standard object-level target that many cultural heritage photographers include in every frame. I wanted something with a more finely differentiated gray scale, like the Munsell Linear Gray Scale, the 21-step target you see at right, so I constructed a set of my own object-level targets. They are made from 3mm Dibond (aluminum composite panel), on to which I laminated inkjet prints I made on a Canon Pro-4000 printer with a very accurate custom profile. I specified L* levels from 0 – 100 in L* 5 steps in Photoshop so that I could make fine adjustments to my images after applying the camera profile.
The Thorlabs PTFE sheet comes with a 3M adhesive backing, so it's extremely easy to simply cut a square with a sharp Olfa utility knife, and adhere it to the inkjet print/Dibond support. If you look carefully, you can see that the PTFE patch at far right, labeled "WB," sits proud of the rest of the target, and casts shadows from the four flash units. The other great advantage to having a set of matched targets (and the gray scale at the center of the oil paint chart is identical to the object level targets), is that I can place one at the center of the work being copied, and one at the edge, and quantify the amount of luminance vignetting I am seeing from a given lens, such as the Schneider Apo-Digitar L 5.6/72, which is one of my main lenses for copy work. With identical values (I created reference files for each target using a Konica Minolta Myiro-1), it's easy to see how well the flat-field correction is working, and whether it needs any fine tuning.
The Thorlabs PTFE sheet comes with a 3M adhesive backing, so it's extremely easy to simply cut a square with a sharp Olfa utility knife, and adhere it to the inkjet print/Dibond support. If you look carefully, you can see that the PTFE patch at far right, labeled "WB," sits proud of the rest of the target, and casts shadows from the four flash units. The other great advantage to having a set of matched targets (and the gray scale at the center of the oil paint chart is identical to the object level targets), is that I can place one at the center of the work being copied, and one at the edge, and quantify the amount of luminance vignetting I am seeing from a given lens, such as the Schneider Apo-Digitar L 5.6/72, which is one of my main lenses for copy work. With identical values (I created reference files for each target using a Konica Minolta Myiro-1), it's easy to see how well the flat-field correction is working, and whether it needs any fine tuning.
Craig Stocks
Well-known member
Thanks for the offer but I have switched to Prophoto strobes now.