I think that a autofocus is not that important because there is time to do that manually.
In most cases manual focus lenses are actually
preferred to autofocus lenses. Most autofocus lenses tend to sag/slip in focus when pointed straight down; most manual focus lenses do better in this regard.
The ideal is either a manual focus lens with a physical lock, or an autofocus camera specifically designed to point straight down such as the
Phase One iXH.
If that is not in the cards, then a manual lens with a rubber band is the next best option. Using an AF lens should be done only after very carefully checking its focus-drift and focus-settling characteristics.
Note that all this is in regards to
The main question is; Do I have to use flashes or LED lights? Until now we were using flashes to black out the room light. But for flash I have to use a lens with a shutter. I think that a Broncolor LED F160 with a CRI of 97+ would be a good decision (used with a big soft box like 30x180 (1'x5.9') or 30x120 (1'x3.9').
Over the last decade most of our
museum/library/archive/gallery have migrated to high CRI/CQS LED lights over strobe. Most of those are now using the DT Photon or DT Photon XL lights we make specifically for heritage reproduction imaging; it has the highest CRI/CQS and the best spectrum for low DeltaE profiling. The Broncolor F160 is also a good option if budget does not allow for Photons.
The main reasons for this transition from strobe to LED is:
1) LED outperforms Flash for color accuracy once both have been carefully profiled. This was not at all the case 10 years ago, but LED technology has improved over time and now outperforms strobe.
2) Flash can be ergonomically bothersome if doing high quantities of imaging (many of our clients have stations running thousands of shots per day)
3) LED can be used with fully electronic shutters which lowers vibration and eliminate shutter wear/tear (again, the context here is that many of our clients shoot a million captures a year where wear/tear is a major factor).
4) LED has the same illuminant for capture and proofing; strobe has a modeling light that is entirely different than the flash tube used for capture. Therefore you cannot use flash for proofing, and a dedicated proofing station will also be using a different illuminant than the flash, so even that is not a great option.
5) Narrow, flat, uniform lights (LEDs like the DT Photon) are more conducive to geometrically accurate and even polarization (for situations where polarization is appropriate). You can, of course, polarize any light source, but when doing it with strobe it's more practically challenging to keep the polarizer parallel and flat; when doing it with a DT Photon it slides in as an accessory, immediately in front of the diffuser – totally flat and parallel.
The main exception to this is object photography where strobes are still heavily preferred due to their prevalence of light shapers, and for very very large artwork (e.g. 12-foot wide paintings) where the absolute power output of high-watt strobes is beneficial.
Note that CRI is a very flawed metric; it's very easy to cheat as a manufacturer. CRI-14 and better yet CQS are better points of reference, but ideally you should be looking at the spectrum itself holistically, both how close a match it is to daylight but also how smooth that spectrum is (which determines how easy it is to create an accurate ICC profile for). More here:
https://heritage-digitaltransitions.com/continuous-lighting-for-cultural-heritage/https://heritage-digitaltransitions.com/cultural-heritage-webinars/lighting-cultural-heritage/I would also emphasize that the differences/improvements being discussed here are pretty small. Very good quality color quality can be derived from a strobe system with proper calibration (we suggest
BasIIColor but there are several good options there). So while (the very best) LEDs are better than strobe, in many contexts that won't matter. The context in which these small improvements matter is when you're talking about strict compliance with, or even exceeding the FADGI 4-star or ISO19264 standards, or are working with color-accuracy-problematic material.
Regarding the Factum: Do you think it is possible/useful to shift the back to get around a 100mp image. Can I reach the FADGI/ISO/METAMORFOZE levels with a stitched image?
Main two issues here are color cast and lens quality.
The CFV uses the 7.5 year old Sony 50mp sensor that produces color cast when light strikes the sensor at a strong angle. This shouldn't be an issue for centered images with a 120mm lens, and probably won't be an issue with 10mm of movement with a ~100mm lens. Where there is meaningful color cast there is also a slight loss of sharpness, which doesn't matter in most cases but when you're trying to keep 92% or higher sampling efficiency in the corner of the frame every little bit counts. But definitely make sure to consider this in your testing from both a practical perspective (do you have to deal with LCC in post?) and a quality perspective (are you still hitting FADGI 4 for color and sharpness in the effected-and-corrected parts of images where you're doing LCCs). I think the chances are pretty good that this won't be problematic, but anytime someone mentions one of the 50mp systems and "shift" in the same sentence I think it's sensible to mention it.
Lenses like the 90HR-SW have an image circle more than large enough to allow you to stitch with super sharp corners (one of the best performing landscape lenses I've ever tested), but are very poor performers for flat artwork due to a modest field curvature which brings the (very sharp) plane of focus forward at the edges of a capture area such as an A0 piece of artwork. That is, those lenses are super duper sharp a few inches in front of the artwork in the corners, which means the art is not sharp.
Apo-Macro-Sironar digital 120 mm f/5,6
No. Not good enough at larger field sizes for your pixel pitch and ISO19264/FADGI-4-star compliance.
The ALPA Makro Switar 5,6/105 mm is just for Alpa - right?
There is no lens that is "exclusive" to Alpa – those are all just rebrandings of Roddy or Schneider lenses. The Rodenstock 105HR is *exceptional* at the macro magnifications it was designed for; no good for larger material.
Schneider Apo-Digitar 4,5/90 mm
Good lens for this. We used it for years on our DT BC100 system. High quality for material in the A3 to A0 size range. No longer available new, and if you buy used you should be very careful to test performance before you accept as we've seen many that are not sharp (when they stopped being available we tried for a while to source them by buying them used; that effort was not long lived).
Schneider Apo-Digitar 5,6/120 mm
The ASPH version of this lens (120mm ASPH) has the image circle and sharpness, but to my knowledge DT is the only entity that still sells this lens new – we custom commission runs of this lens directly from Schneider. Very few were made so finding them used is hard, and if you do make sure to test before accepting.
The N version (120mm N) is pretty good for your use case. Can't be bought new and not that many were made.
The M version (120mm M) is great for macro but not for larger art.
There is also a new FADGI/ISO target:
https://www.youtube.com/watch?v=AC6agcSOXa0Does anyone know that
https://deltae.picturae.com will update to that target?
[/quote]
It will not be updated. It has not been updated in many years, and the developer left the company that publishes that website.
Golden Thread is the best commercial option for evaluating that target. If budget does not allow/justify that, OpenDICE published by Library of Congress is free; just know that it can be a bit wonky to install, use, and that there is no commercial support.
---
You didn't ask, but if you haven't taken our
Digitization 101, Intro to Modern Digitization course I strongly suggest it. We're coming up on 1000 attendees from museums, libraries, archives, and galleries from around the world so even if you don't learn anything new (though you very likely will) in the very least you'll be exposed to the terminology, philosophy, and techniques used in those institutions. Many of our clients who previously provided art reproduction services to individual photographers have migrated to providing digitization services (same actual service; just a different word used in that market) to museums, libraries, and archives. Many artists are now fine with an iPhone picture of their art so there is simply less demand for professional art reproduction than 20 years ago. In contrast, the amount of digitization work being done in museums, libraries, and archives has been going up sharply over that same period.
---
Re stands themselves I'd encourage you to consider our stands such as the
DT Element or
DT Atom. Our systems are US Made and designed and calibrated for modern 150+ megapixel capture at FADGI 4-star compliance levels of sharpness. We used to sell the stands of a legacy manufacturer (I won't sully their name) but found they were using basically the same design as they did in the film era and that their design (soldered joints, unidirectional joiners) was problematic for maintaining sufficiently precise parallelism. Our benches also integrate with the capture software, provide finer vertical movement increments (more important for macro or film scanning than for large material). Finally, our benches are designed for use with a wide range of accessories to facilitate a range of capture modes such as objects (on a product sweep with integrated transillumination), film scanning, book scanning, and stitching for larger material.
We also have a new laser alignment system that we'll be announcing soon (couple hundred bucks) that outperforms the accuracy of the 3rd party systems we've previously used.
I'm the Head of R+D so would be glad to help with any technical questions on our systems or on digitization in General.
