Cover Image: DTDCH.

com
Ansel Adams by Edward Weston 212-529-6825
From the Center for Creative Photography 35 W 35th St.
Version - 1.0.0 Floor 2
New York, NY, 10001

Colophon & Acknowledgments

This document was prepared by the Cultural Heritage teams at Digital Transitions and Phase One.
For feedback and general inquiries please contact info@dtdch.com. Special thanks:

Doug Peterson, Digital Transitions, Lead Author

Peter Siegel, Digital Transitions, Review
Rudy Lorejo, Digital Transitions, Screen Grabs
Arnab Chatterjee, Research and Review
Spencer Zidarich, Digital Transitions, Research and Review

Liability
Digital Transitions and Phase One disclaim any and all liabilities from the interpretation and use of the information
presented in this document. All content and images are provided on an “as is” basis. Any images and/or information
used or obtained from this document is done at your own risk and you will be solely responsible for all damage, including
but not limited to damage to your computer system, hardware or loss of data.

Trademarks
DT Photon ®, DT Atom ®, and DT CulturalHeritage™ are trademarks of Digital Transitions. Capture One and Phase One are
trademarks of Phase One A/S in the European Union and/or other countries. All other trademarks are the property of their
respective owners.

© 2018 Digital Transitions & Phase One. All rights reserved.

2
Table of Contents
Introduction Method: Content Reproduction
1.1 The Goals of this Document 5 6.1 Clone Variant 37
1.2 Tools Used in this Document 5 6.2 Set Mode 37
1.3 Feedback or Questions? 5 6.3 Set Curve 38
6.4 White Balance 38
Needs 6.5 Set Exposure 39
2.1 Compared to Reflective 7 6.6 Set Crop 39
2.2 Stakeholder Needs 8
2.3 Historical In-Person Uses 9 Method: Speculative Artist’s Intended Rendering
7.1 Clone Variant 41
Challenges 7.2 Research the Artist’s Intentions 41
3.1 The Film Ecosystem 13 7.3 Adjust Color & Tone 42
3.2 Lost Aesthetic Context 15
3.3 Lost Technical Context 16 Batch and Automatic Processing
3.4 Big Challenges Going Small 17 8.1 AutoCrop 47
3.5 Conservation Issues 20 8.2 Modify Crop & DT Crop Control CH 49
3.6 Handling Concerns 21 8.3 Auto Levels 51
8.4 Copy and Apply 52
Digitization Strategy 8.5 Naming and Folder Structure 52
4.1 Multiple Derivatives 25 8.6 AppleScript 53
4.2 Metadata 26
4.3 Presentation 26 Advanced Techniques
9.1 Stitching 55
Method: Object Reproduction (PDO) 9.2 Reflective Content 56
5.1 Select Lens & Extension 29 9.3 Capture Multiple Variants 57
5.2 Set Camera Position & Focus 30
5.3 Set Mode, Profile, and Curve 32 Appendix A: What PPI is Best?
5.4 Set Sharpening & Noise Reduction 32 A.1 PPI Selection Strategies 59
5.5 Capture LCC Raw File 32 A.2 What do the Standards Say? 59
5.6 Analyze LCC 34 A.3 Practical Considerations 60
5.7 Apply LCC 34
5.8 Set White Balance 35 Appendix B: Additional Resources
5.9 Set Exposure 35 B.1 List of Written Resources 63
5.10 About the Crop 35 B.2 List of Individuals 63
5.11 Ready to Capture 35
Appendix C: Checklist Summaries
C.1 iXG, DT AutoColumn, & DT Photons 65
C.2 DT RCam, CMOS, DT AutoColumn 66
& DT Photons

3
 1
Introduction

4
1.1 The Goals of this Document?
This document discusses the theory and practice of the preservation-grade digitization of transmissive materials. While
the discussion of theory is informative regardless of the hardware and software used, the discussion of practice is in the
context of solutions provided by the Digital Transitions Division of Cultural Heritage.

The end-product of preservation-grade digitization is a Preservation Digital Object (PDO) that can serve as a surrogate to
the original object, replacing most needs for physical access to the physical original, aiding its preservation, broadening
access to its content.

Many transmissive material collections are in a precarious physical state. That places great urgency on their digitization.
But while the technology available to digitize transmissive materials has radically improved in the last decade, most film-
scanning references were written well before the modern era of instant film capture (a.k.a. camera-based digitization).
In this void we saw a need for a document that covered both the theory and practice of transmissive digitization written
with modern-technology. Digital Transitions is a member of the ISO Technical Committee 42. There we will work with
other stakeholders across the globe toward the goal of an international standard for the digitization of transmissive
materials. However, such ISO efforts can take years and produce documents that must be brand/equipment agnostic,
which limits their practical in-the-trenches utility. Therefore, we’ve undertaken to create this document, in hopes that it
can immediately serve as a useful reference for the community.

While we are a commercial entity, our first and primary mission is to further the cause of Cultural Heritage Digitization.
We hope you find that this document helps you fulfill that mission for your institution’s transmissive material collections.

1.2 Tools Used in this Document

Capture Device: DT RCam, Phase One iXG, or Phase One XF

Station: DT Atom, DT Versa, DT Element, or DT XY Film Scanner

Transillumination: DT Photon Cultural Heritage LED

Software: Phase One Capture One CH

Targets: ISA Standard Film Targets

Validation: ISA Golden Thread

Even Fielding: White matte Plexiglas, modestly larger than the largest film format to be digitized

Raw File Sample Set: DT CH Transmissive Raw File Sample Set, with contributions from many CH institutions
Available for download at dtdch.com

1.3 Feedback or Questions?

Please do not hesitate to contact us (212.529.6825, info@dtdch.com) with questions, feedback, or to arrange an in-
person or remote training on the topics covered in this document. We hope the document fosters vibrant discussion; the
more feedback we receive the better we can make this document in the future.

5
 2
Needs

6
2.1 Compared to Reflective
To explore the questions raised by transmissive digitization we begin by discussing the relatively straight forward matter
of reflective digitization. The digitization of film poses philosophical and aesthetic questions that are largely absent in the
digitization of most reflective paper materials.

For the purpose of this document we use the goal of digitization defined by the Digitization Program Planning Guide by
Digital Transitions Division of Cultural Heritage:

The goal of digitization is the creation of a Preservation Digital Object (PDO) which can act as a
surrogate for most in-person use of the original object.

Digitization which meets this standard serves the dual goals of improving both access and preservation. Access to the
material is improved by providing remote viewing. Preservation is advanced, both in that less frequent physical access
exerts a lower wear-and-tear toll on the original object, and in that the PDO is a failsafe in the case the physical original
is lost to time or catastrophe. Finally, preservation is also advanced in that a PDO documents the state of the object at a
given moment in time; in the case that the item shows deterioration over time or receives conservation treatments that
PDO is the only record of the state of the object at that moment.

For reflective objects such as a handwritten note on paper the operative phrase “as a surrogate for most in-person use”
has a reasonably obvious definition, namely that of viewing of the material in a reading room on a table, book support, or
held in the hand. Therefore the goal for the reflective digitization of a handwritten note is that someone should be equally
advantaged looking at the PDO on a screen as looking at the physical original.

There are some worthwhile exceptions to this simplified explanation, but they are minor in impact. There are technical
nuances between emitted light (e.g. from a computer screen) and reflective light. There is reasonable debate about how
bright reading room lights are and therefore how bright the surrogate digital screen should be. Moreover this simplified
explanation pretends that paper and similar materials are two dimensional, which ignores the potential benefit of
techniques that enhance or unveil the surface texture of the substrate. There are also uses that fall outside “most in-
person use.” For example, detailed chemical analysis or carbon dating cannot be done using a PDO, though multispectral
imaging can produce a PDO that provides some insight into what pigments, dyes, or materials are present in the object.

Still, generally speaking, preservation-grade digitization of reflective materials simply needs to produce a PDO that looks
like the object sitting in the reading room does. The particulars of how to precisely and consistently achieve that goal can
(and do) take up entire volumes. But the goal itself is obvious and simple.

The same cannot be said of digitizing transmissive collections. If the goal of digitization is the “creation of a Preservation
Digital Object (PDO) which can act as a surrogate for most in-person use of the original object” we must define “most in-
person use” in the context of transmissive originals. In other words we need to understand the needs.

7
 2.2 Stakeholder Needs
Depending on the nature of the material and the nature of the stakeholder there are three broad categories of needs that
one might have from the Preservation Digital Object of a transmissive original.

Object Reproduction
Sometimes the stakeholder has needs that require viewing the original material as an object unto itself. For example, a
researcher may want to examine the notch pattern on a 4x5 negative to determine what emulsion it is. This has nothing,
directly, to do with the content of the image the negative contains, but does inform the researcher’s interpretation of that
content. As another example, a conservator may want to judge the condition of a strip of black and white 35mm strip
film. To have an accurate understanding of the state of that film’s physical condition requires a color image of the object
even though the content of the image is monochromatic. The color image will allow the conservator to see if there are
stains of residual processing chemicals on the film (which might warrant the film being washed) as these stains often
show up as brown.

Content Reproduction
Sometimes the stakeholder has needs that require viewing the content of the image the film contains, devoid of and
separate from the physical object (the film) that contains that content. For example, if a member of the general public
conducts a search for “photos of the Yalta Conference” they are unlikely to benefit from seeing the sprocket holes that
surround a relevant photographic image. In fact, seeing these sprocket holes may confuse the viewer, causing confusion
or misunderstanding. As another example, a photographic negative contains a tonally-inverted (“negative”) image.
When viewed as a Object Reproduction the image remains tonally-inverted which makes it very difficult to understand
the content of that image. A Content Reproduction of a photographic negative presents the image as a positive, greatly
improving its legibility. Take, as a final example, a 35mm slide that contains a very dark/dense image, which obfuscates
a proper understanding of the subject matter in the image. An Object Reproduction must faithfully reproduce the object
as-is, dark image and all, whereas a Content Reproduction would adjust the image based on the content of the image.

Speculative Artist’s Intention

The third category of stakeholder needs is the most challenging; it is the need to see the image as the artist would have
intended. This is only relevant when the object in question was created by an artist with the intention of sharing an image
with others. As discussed in “The Artist’s Intention” section, the film itself is only one component of how the image was
intended to be presented to viewers. A curator hoping to exhibit a collection of Ansel Adam’s work may want to exhibit
Content Reproductions of his work based on the film itself, but may want instead, or in addition, to present the final
images as closely to how Ansel Adams would have presented them. If the relevant original photographic prints (made by
Ansel Adams or commissioned by him contemporaneously) exist and are available for exhibition then they are the more
direct way to accomplish that goal. However, if one or more of the images only exist on film then it may be beneficial to
create a Speculative Artist’s Intention asset.

This is an inherently subjective process and the precision of reproduction depends heavily on the amount of useful
reference material available. Sometimes a great deal is known about the period, artist, general body of work, or specific
work. For example, one of the artist’s original prints may be available, or notes may accompany the transmissive object
that indicate how the translation of that transmissive object into image was to be handled (e.g. an artist’s darkroom
processing notes written for a professional darkroom printer). Or more general information may be drawn on such as
research that indicates the photographer almost always printed their work on a warm-tone paper, or almost always
added contrast during the print process.

Summary Table of Stakeholder Needs

Object Reproduction Content Reproduction Speculative Artist’s Intention
Main Audience Conservators Researchers Researchers
Researchers General public General public
Curators Curators
Art Authenticator
Insurance
Unique Information Object physical condition Easily readable Shows the image the way the artist
Object type Content discovery intended it to be viewed.
Artist’s capture method

8
2.3 Historical In-Person Uses
With complex and varying needs from different stakeholders, how did film collections historically facilitate those needs
via in-person use? Put differently, when a user came to a film collection, what did they typically do with the film?

How do these uses inform the goals and methods of the modern digitization of transmissive collections? They reassure
us that there has never been a single use-case for transmissive items. One reason modern film digitization is so involved
is because there has always been a wide range of uses of film originals.

Lightbox viewing
One “in-person use” was the placement of the transmissive object on a light table or other transmissive light source to
view the image with the naked eye or through a magnifying loupe. This was especially common with slide film or other
positive transmissive media where the image is tone-normal (rather than tonally inverted, i.e. negatives).

However, it was also used when viewing negatives, especially when the viewer was seeking out technically-oriented
information such as how bright or dark (“thin” or “dense”) the in-camera exposure was. Notably, the brightness and color
temperature of such light boxes varied throughout history and even in a given time period could vary significantly from
one make/model to the next. Because the viewer was looking directly at the media itself their eye could adapt to the
overall exposure of a given frame, making the experience of viewing an especially bright (thin) or dark (dense) frame less
visually stressful and problematic than in other “in-person use.”

Projector
Another “in-person use” was the use of a film projector to project the image held by the film onto a viewing surface
such as a white wall or a dedicated projector screen. This was often done in the context of a presentation or other group
viewing setting. The brightness and color temperature of the image from a projector generated image depended on the
make/model of the projector, the distance between the projector and the surface, and the nature of surface on which it
was projected.

Contact Print or Contact Sheet

A darkroom contact print or contact sheet were commonly requested from some transmissive archives, especially
for large-format negatives and glass-plate negatives. When creating a contact print the negative was placed directly
in contact with photographic paper and exposed to light, resulting in a reproduction of the same size as the negative.
Dodging and burning was possible by means of waving a hand or tool between the light source and the negative, or by
using chemistry to effect tonality.

Enlarger Print
A darkroom print made at an enlarger was a common requested use for some film archives. This was available as an
option for negative materials since the early days of film photography. The advent of Cibachrome also allowed this option
for positive slide film without requiring a cumbersome interneg to be generated. With an enlarger print the film was
placed in a carrier in a photographic enlarger which used a lens and light source to project the image onto photographic
paper. This allowed reproductions larger or smaller than the size of the film. Dodging and burning was possible by means
of waving a had or tool between the light source and the negative. Notably, most of the processes used to print slide
positives increased contrast and/or saturation of the image.

Film Duplication (a.k.a. “Slide Dupe”)

Using a machine or manual process the image from a piece of film could be duplicated onto another piece of film. While
there was some inherent loss in image quality through this process, high-quality film duplication was possible.

Microfilm transfer
When the stakeholder did not require a reproduction with accurate color, tone, or detail, a microfilm transfer allowed
large quantities of film content to be economically carried or shipped. A microfilm duplication could not hold much detail
compared to a larger format original (e.g. a 35mm slide contains more detail than will fit on microfilm) and the transfer
process was often not well controlled, leading to additional loss of image quality. The resulting microfilm was then most
commonly viewed by means of a Microfilm Reader, an optical device that optically magnified the content of the microfilm
onto a screen or into an eyepiece.

Condition Check
Film is widely misunderstood to be inherently archival in nature. In fact all films can and do change over time, in both
9
 subtle and catastrophic ways. Conservators would occasionally survey a transmissive collection to evaluate their current
condition and asses the collection’s ongoing stability or deterioration. This included visual, mechanical, and olfactory
inspection (the latter useful for detecting Vinegar Syndrome).

Catalog Check
Institutions periodically audit their collections to make sure that their collection is correctly described by their Collections
Management System or other database or catalog. Transmissive collections would be audited for completeness and
accuracy. Such audits are always time consuming and tedious, but transmissive collections were especially tedious to
audit due to the difficulty of viewing transmissive material as quickly and easily as most reflective material.

10
11
 3
Challenges

12
3.1 The Film Ecosystem
As an artistic medium a piece of physical film was rarely the end product; usually film was one component in a
ecosystem of tools and techniques that were used to create and present a final image. The object of film cannot be fully
understood without considering the rest of that ecosystem.

Ansel Adams, the Zone System, and the Utilitarian View of the Film Image
Ansel Adams’ (American, b. 1902) images of the American West are iconic. But as significant as his photography was, his
contribution to the theory and practice of film photography may have been even more influential.

Adams considered a photographic print to be the goal, with physical film only an intermediate step. Along with Fred
Archer he crafted the Zone System, a technical framework to producing a photographic print with the exact range of
tones, contrast, and subject rendering they desire. In the Zone System the final photographic print is the artistic object
of merit and the content recorded on the photographic film is a carefully controlled means to that end. In fact, the Zone
System often requires the photographer to purposefully skew the exposure on the film in a way would normally be
considered a “thin” or “dense” incorrectly (or at least “abnormally”) exposed negative.

When viewing a piece of film produced by a photographer who adhered to the Zone System it is wholly inadequate to
consider the film itself as the complete and final indication of the artist’s intention. Ansel himself said that “The negative
is the equivalent of the composer’s score, and the print the performance.” That analogy is especially illuminative in terms
of digitization theory. It highlights that the artwork itself is not fully preserved by a neutral digitization of the film (the
“score”); something analogous to the print must also be produced. It simultaneously explains why we cannot dismiss
the value of having a neutral and faithful reproduction of the composer’s score. Both the score and the performance hold
value; one speaks to the intention and the other speaks to the methods used to achieve it.

The Rest of the “Film System”

The “Film Ecosystem” included many technical and creative use of a variety of tools and techniques. A partial list
includes:
In-Camera Choices Film Choices Enlargement Choices Print Choices Presentation
Film Emulsion Push or pull processing Paper Type On-Print Retouching Frame
Lens Filters Processing Chemistry Paper Size Chemical Toning (e.g. sepia) Mat
On-film retouching Crop Chemical Dodge & Burn Placement
Optical Dodge & Burn Printing Chemistry Illumination
Unsharp Mask

We say these constitute an ecosystem in that these tools and techniques are deeply interrelated and were developed
in symbiotic and concurrent ways. For example, the choice of Film Emulsion, Processing Chemistry, and Paper Type
were often tied together by a brand; a photographer might have shot Kodak color-negative film stock, processed that
film in Kodak chemistry, and then printed it onto Kodak paper processed in Kodak chemistry, which would result in a
different image than if printed on a Fuji paper. Such an ecosystem of matched components (film+paper+chemistry) often
imparted a non-neutral final rendering of the image; linear scans of that same film won’t look the same, because part of
that look is lost without the rest of the ecosystem.

Purposefully-Mismatched Processes
The challenge is even greater when the artist was purposefully using “mismatched” components. For example, some
photographers would purposefully process color positive film (“slide film”) through color negative chemistry such as
C-41. This would result in imagery with a purposefully non-standard color rendering. How that resulting film would
interact with various other parts of the film ecosystem was hard to predict even at the time, and even harder to
reconstruct after the fact.

13
 After-Dinner Games. 1947 Still Life with Watermelon, 1947 Salad Ingredients, New York, 1947

14
3.2 Lost Aesthetic Context
Time will erode the mechanical and technical knowledge of photography, but it will also slowly shift our visual
understanding of it. There is an unpredictable shift in how viewers interpret visual art due to evolving aesthetic norms.
For example, take the shift toward darker luminance among Hollywood films between 1935 and 2010 documented
in a Cornell paper (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3485803/). Aesthetic shifts of this sort make it
impossible for a modern viewer to experience the art in the context of a contemporary viewer.

This is true of viewing nearly any type of art and is instantly recognizable, for example, in cinema. The experience of
watching Star Wars in a theater in 1977 is not something that can be recaptured by a modern viewer. Time Magazine’s
original review was impressed by “some of the most ingenious special effects ever contrived for film” but to a modern
viewer these same effects are distractingly “dated” at times. The acting in the movie Paisan (1946) may have struck
contemporary audiences as naturalistic, but to a modern viewer the same acting feels stilted and self-conscious.

However, photographic material can be mistaken for an accurate capture of reality, rather than an inherently creative and
editorial creation of a new reality, so it might otherwise be overlooked that shifting aesthetic context alters our visual
understanding of film images in the same way it alters our perception of other art.

For example, Irving Penn’s 1947 color still-life series has a distinctly muted color palette bordering on pastel. It’s
impossible to say how much this color palette was influenced by the practical question of which film and print processes
were available to him, and how much was influenced by Penn’s aesthetic choices. It’s not even possible to separate
those two from each other. While an artist chooses their tools deliberately, they must select from the tools that are
available. As an obvious example, a photographer living in 1830 had to execute their vision within a monochromatic
palette, as color photography was yet to be invented.

A viewer in 2018 will likely perceive this work as muted in tone and color, because they, the 2018 viewer, are surrounded
by images with significantly higher contrast and saturation. Unlike the 2018 viewer, a viewer in 1947 would have been
most accustomed to color photography that was broadly similar in contrast and saturation to this series. This differing
context may lead the 2018 viewer to imbue meaning to the selection of this color palette that the 1947 viewer would not.

To illustrate this see the images on the opposing page. The three images are all part of Irving Penn’s still life series
completed in New York. In the top row the color/tone is a faithful capture of Penn’s original prints digitized by the Irving
Penn Foundation. In the middle row the contrast and saturation have been increased to simulate an out-of-camera
image from an iPhone. In the third row the Instagram filter “Clarendon” has been applied. As a thought experiment,
imagine that you lived in a world where every color photograph you saw had the tones and color exhibited in the first
row. Living in that world the other two rows might seem aggressively, surrealistically contrasty and colorful. Of course,
for most readers the opposite will be true; immersed in a world full of high-contrast high-saturation images the first row
seems somewhat drab and lackluster.

In the unknown future it may be that two-dimensional imagery is usurped by omni-present augmented-reality three-
dimensional images. In a world where the prevailing use of imagery is three-dimensional the two-dimensional context
of today’s photographic work may itself become obscured to the viewer. Black and white family portraits strike many
viewers of today (especially younger viewers) as immediately “historical” or “antiquated”; a similar effect may cause
family portraits captured in “just” two dimensions to be immediately perceived as “old” and “distant”.

15
 3.3 Lost Technical Context
Niels Bohr, the Danish physicist, once wrote “Prediction is very difficult, especially about the future” and this adage
is useful when discussing digitization of transmissive collections. It is very hard to know whether a future viewer
will be familiar with a particular photographic process. This is obvious for niche knowledge such as what should be
expected when cross processing E-6 color slide film in C-41 chemistry. But this is also true for more foundational/basic
information such as what sprocket holes are, or how to tell the difference between a scan of one frame from a roll versus
a scan of a stand-alone large format frame.

Consider that film sales peaked in 1999 which means there are college students in 2018 who were born into a world that
already considered film photography to be “before my time.” Based on US Census projections (https://www.census.gov/
newsroom/cspan/pop_proj/20121214_cspan_popproj.pdf) by the year 2060 only 18% of the country will have been born
before film was on it’s way out of popular use, and many of them will be retired. Cultural Heritage Preservation is meant
for timescales of centuries, so the year 2060 is short-term planning.

Film photography, as a broadly-popular widely-practiced and commonly-experienced media, is a blip on the grand
timescale of art and cultural history.

16
3.4 Big Challenges Going Small
Transmissive materials usually have very physically-small details. For example, the film emulsion on Provia slide film
had a grain size of roughly 10 microns. As a point of comparison a single grain of table salt is approximately 300 microns
across. The technical and physical challenges of imaging subject matter this small are significant.

Lens Design and Quality

Lens designs are typically optimized for a certain range of magnifications, so lenses that are well-suited for
photographing larger material like A4 books or A2 maps may not be well-suited for digitizing smaller transmissive
material. In general, when digitizing small material, a lens labeled “macro” is preferable to one that is not, but neither this
label, nor a particular brand name, nor a particular price point will guarantee proper performance for the task of high-
quality transmissive digitization. Only careful testing of a lens can determine its objective quality.

DT systems use four lenses: the Schneider 72mm digitar, Schneider 120mm ASPH, Schneider 120mm LS BR, and the
Rodenstock 105HR. We selected these lenses after extensive testing of a variety of candidates and determined that they
exceed the quality required to surpass FADGI 4-star image quality. Note that the Schneider 72mm Digitar performance is
best for transmissive material 4x5 inches and larger.

Lens Performance
All lenses have imperfections that are masked by stopping the lens down to a smaller aperture. One attribute that defines
a “high-quality lens” is that it does not need to be stopped down as far to achieve good results.

Using a smaller aperture opening (a higher aperture number) generally improves the performance of the lens, until
diffraction becomes significant; this is discussed below.

Depth of Field
In all standard photographic systems there is a plane of perfect sharpness, in front of which and behind which there is
a range called the “depth of field” that is not detectably or meaningfully different in sharpness, beyond which subject
matter goes out of focus. It’s important to note that subtle details may be lost well before the subject matter becomes
obviously out of focus to an uncritical eye. When imaging general collections items like an A4 book the depth of field
is often in the range of an inch or two. This means if the subject is not completely flat, or the plane of focus is off by a
fraction of an inch, the entire subject matter can remain in crisp focus.

When imaging small transmissive material, the depth of field is often much less than 1 millimeter. That means the
material must be very flat for it to all be equally in focus; any curvature, channeling, or other physical distortion risks
falling outside the depth of field and losing subtle detail, or going entirely out of focus.

Using a smaller aperture opening (a higher aperture number) will increase depth of field.

Diffraction
It’s widely understood among photographers that a given lens+camera combination will have a “sharpest aperture” and
that if the aperture is stopped down too far (e.g. f/22) the image loses fine detail. But many photographers do not know
that the aperture at which this effect occurs changes based on magnification (i.e. how close the camera is focused).
When digitizing small objects that fill the frame, diffraction occurs earlier (at aperture values with lower numbers).

For example, for a Phase One iXG 100mp (which uses a full-frame 645 sensor with 4.6 micron pixels) loss of sharpness
due to diffraction is acceptable up to around f/11 when digitizing large materials (e.g. an A3 poster) but when digitizing a
frame of 645 film diffraction occurs earlier, such that apertures smaller in size (higher in number) than f/8 should not be
used (i.e. an aperture of f/7 or f/8 is acceptable for digitizing 645 film but not f/9 or f/11).

For the technically minded, the formula underlying the above paragraph is “EffectiveAperture = MarkedAperture * (1 +
Magnification)”. Filling the frame of a 645-sized sensor with a piece of 645 film implies a magnification of 1.0x (since the
size of the subject is the same as the sensor). Using the formula a marked aperture of f/8 at 1.0x magnification produces
an effective aperture of ~f/11. Working backwards from the earlier statement that f/11 is acceptable for digitizing large
materials (low magnification) with an iXG 100mp, that means f/8 will be acceptable for 1.0x magnification.

Using a smaller aperture opening (a higher aperture number) will increase diffraction which will decrease sharpness.

17
 Planarity
In nearly all imaging of flat materials, the camera must be sufficiently parallel to the subject; that is, you must have
sufficient planarity. In general collections imaging (e.g. capturing an A4 document) half a degree of misalignment
may not prevent compliance with FADGI 4-star image quality. Therefore a variety of methods for of asserting camera
alignment are often enough (e.g. “just eying it”) and a camera mount that has a bit of rotational play (aka “wiggle”) is not
especially problematic.

When imaging small transmissive objects a small amount of non-planarity can cause significant loss or unevenness of
sharpness. We recommend using a Parallel laser alignment tool to ensure the camera is completely perpendicular to
the table. The DT AutoColumn provides four set-screws that allow the operator to perfectly level the camera to the laser
alignment tool.

Using a smaller aperture opening (a higher aperture number) will decrease sensitivity to non-planarity.

Rigidity
Once planarity has been achieved, it must be rigidly held by the system and the focus must not drift. For workflow and
productivity it is essential that the operator be able to trust that, once set, the focus and alignment of the system will
remain correct. This requires that the camera focus mechanism, camera-arm, column, and working surface are all
sufficiently rigid.

The focus mechanism should not drift or settle over time. With general-purpose cameras the focus mechanism is
designed with speed of focus-changes as a high priority, often resulting in gearing that is especially “loose” such that the
lens, when pointed straight down, can drift or settle. “Focus Drift” is the phenomenon of the lens continually changing
(typically very slowly) after it is set, whereas “Focus Settling” refers to the focus settling into a focus position other than
it was set to, but in a way that is not continuous. Tapping the lens focus mechanism with gaffers tape or painting tape,
or using a large rubber band, can reduce or eliminate focus drift, but does not eliminate the possibility of Focus Settling
since the focus mechanism may still settle into the elastic limit of the tape or rubber band. The DT RCam and Phase One
iXG use focus mechanisms that do not drift or settle while pointing down.

The camera arm should be rigid and lockable. Some copy stands use lightweight arms that allow the camera to droop
(often so much that it is easily visible) when extended or retracted, or which droop differently depending on whether their
position is locked or not. This is highly detrimental to the digitization of transmissive materials. The DT AutoColumn and
AutoColumn XL use a fully-sleeved design that does not droop when extended, retracted, or locked.

The column and working surface should also be rigid. Some budget copy stands have a column that is attached at
one point and a base which is flat. Those design choices limit the rigidity of the system and introduce play that limits
precision and repeatability. DT Stations always restrain the column at two points (one under the table surface) to
increase rigidness. The base of a DT Station is a cube rather than a single surface, to maximize the rigidity and alignment
of the working surface (the table top).

Using a smaller aperture opening (a higher aperture number) will decrease sensitivity to drifting alignment and focus
caused by insufficient rigidity.

Vibration
If either the camera or the material vibrate excessively it can cause loss of sharpness. Even a very small amount of
vibration can be problematic when digitizing transmissive material at high resolution. For example when capturing film
at 5000ppi each pixel represents 1/5000th of an inch in the real world, so if a camera is vibrating such that it moves back
and forth a mere 1/5000th of an inch the resulting image will be slightly blurry when viewed at 100% because details
that should have been projected onto a single pixel will have oscillated between two pixels. Consider where to install the
system. Ideal locations are those that have the fewest and furthest sources of external vibration such as air conditioning
units, foot traffic, and vehicular traffic on nearby roads. It is equally important to take into account the material the floor
is made of. When the floor supports are something other than concrete (e.g. a wood floor), using the corner of a room
can be preferable to the center of a room where the floor has more give/bounce. Secondly, consider what type of table
to use the system on; in general, the more sturdy the table and the stronger the support directly under the system, the
better. Finally, consider using vibration-isolation material such as rubber or rubber+cork pads between the floor and
table, and table and system. DT recommends PneumaticPlus-brand 4x4” pads.

18
The system also needs to create as little internal vibration as possible. The system should use rigid components that
have minimal play. When the system has movements they should be smooth, but when the movement is meant to
stop they should hold very firmly still by means of locks or detentes. DT uses detentes for the smooth lateral and fore-
aft stitching movement of the DT XY Film Scanner. The DT Film Scanning Stage provides smooth but self-arresting
movement of a strip or slide carrier.

If you’re using a CMOS based camera with good live view, such as the iXG or DT RCam with IQ3 50mp or 100mp back, an
easy check how well your system is handling vibration is to open live view and zoom to 100% on a sharply focused piece
of film at the highest PPI you plan to use. With that live view running practice your normal digitization movements (e.g.
move a loaded slide carrier across the stage, stopping at each slide detent). If the live view is visibly shaky then vibration
is very likely to cause issues with sharpness.

Note that shutter speed is an important factor in vibration. The longer the exposure the longer it has for vibration to
accumulate to a problematic level. As the aperture size decreases the exposure time lengthens, exposing the camera to
vibration longer. Therefore using a smaller aperture opening (a higher aperture number) will indirectly increase sensitivity
to vibration.

Aperture Selection Summary

The proper selection of aperture is critical to the digitization of cultural heritage materials. Complying with FADGI 4-star
image-quality standards requires extremely good sharpness throughout the object, including in the edges and corners.
Transmissive materials usually have a consistent grain throughout the image, so if the Preservation Digital Object is not
homogeneously sharp, it will be clearly visible, as the grain will not be equally sharp throughout.

As you’ve read throughout this section many factors are effected by aperture, some of which improve with the size of the
aperture opening and some of which do the opposite.

Factors that improve as aperture size decreases (i.e. as the aperture number increases):
• Depth of field: as the aperture size decreases the depth of field increases, which makes it easier to accommodate
subjects that are not perfectly flat, larger errors in planarity, and more slop in rigidity.
• Lens Performance: as the aperture size decreases the performance of the lens improves.

Factors that degrade as aperture size decreases (i.e. as the aperture number increases):
• Diffraction: as the aperture size decreases diffraction will increasingly soften the image
• Vibration: as the aperture size decreases the exposure time lengthens, exposing the camera to more vibration

Accordingly, the ideal aperture is the one that best balances the above factors. It’s hard to provide a single number that
correctly achieves this balance for all DT systems and potential use cases and environments. Where possible, it’s best to
do your own testing wherein you capture representative materials and test charts at a series of candidate apertures (e.g.
f/8, f/9, f/10) and evaluate them based on visual inspection on screen at 100% and numerical analysis of transmissive
targets in Golden Thread.

However, to provide a starting point and a sanity check, our finding is that, for DT systems, f/8 is a good rule of thumb.
Larger aperture numbers (e.g. f/12) may be warranted when you must prioritize depth of field, such as imaging
channeled negatives, over a slight loss of sharpness. Smaller aperture numbers such as f/5.6 may be warranted when
imaging at especially high magnification (e.g. using an RCam with a Rodenstock 105HR and two 60mm extension tubes
to image 35mm slides at 8000ppi).

In summary, aperture selection is a complex task best achieved by carefully testing your specific set up, but if you’re
looking for the “short answer” we recommend DT customers use f/8 unless they have a clear reason not to.

19
 3.5 Condition Issues
There are many causes and many symptoms of the physical deterioration of transmissive collections. This document
is a practical guide to digitization, not a comprehensive manual for film conservation, so we will skip the very interesting
chemistry and physics that explain these degradations. Instead, we will focus on the considerations important for those
charged with the digitization of transmissive collections.

All types of transmissive materials are vulnerable to degradation. It’s true that certain kinds of film are especially prone
to certain kinds of degradation, but all major types of film are susceptible to at least one form of degradation, even when
properly stored.

Degradation can be slowed by proper storage, but it cannot be completely stopped. The speed of degradation increases
when film is stored at too high a temperature or outside the recommended relative humidity range. Freezing film radically
slows degradation, but storing film frozen costs more money, reduces access, and makes digitization more difficult.

One of the most common symptoms of degradation is “vinegar syndrome” which, as its name implies, causes a distinct
vinegar smell. But not all causes of deterioration carry such obvious symptoms, especially at early stages, and even
obvious symptoms often occur only after the window for practical remediation has already expired. In other words
symptoms like the smell of vinegar are often more of a hot-doorknob than a fire-alarm; they indicate a situation that is
already very serious rather than constituting an early-warning-sign.

Below are some notable visual symptoms of various types of degradation.

Delamination or Channeling Embrittlement Plasticizer Exudation Distortion

The base of acetate films can The base of acetate and nitrate Plasticizers were chemicals added When film shrinks it can do so,
shrink over time, causing the films can lose elasticity over time, to film. As acetate film decays causing warping or curling. This is
emulsion to bunch up in a way that making the film susceptible to rips, these plasticizers can create especially problematic when the
creates visible channels. tears, or breaks when handled. bubbles. film is too brittle to mount in a flat
way for scanning.

All images on this page provided

as a courtesy by the Image
Permanence Institute at the
Rochester Institute of Technology.
Visit filmcare.org for more
information.

Silver Fading Dye Fade Mold Growth

Black and white films can lose The natural breakdown of the dyes Mold or fungus can attack the
silver to corrosion, causing in color films. This can manifest as binder of a variety of film types.
an overall fading that is often an overall loss of contrast or a shift
accompanied by discoloration. or change in color content.

20
3.6 Handling Issues
General Handling Thesis
The method of digitization should adhere to the same conservation handling procedures as those required for in-
person viewing. The film should never be fed through an automatic feeding mechanism. The working surfaces of the
digitization system which contact the film should be cleaned regularly to prevent the spread of any contamination (such
as unwashed fixer or mold) from one batch of film to the next. The operator should wear conservator-approved gloves
while handling the material, and should be careful of touching problematic surfaces with those gloves such as their face
or food.

Carriers: Magnetic vs Glass

Most transmissive materials consist of a base layer and an emulsion layer (or layers). The emulsion bears or contains
the actual image content. Moreover, the emulsion is typically less physically sturdy and easier to mar, scratch, or
otherwise harm. For this reason, while the object of the entire safety must be looked after, the emulsion requires special
attention.

Where possible it is best to use carriers that do not contact the emulsion at all, or only contact the emulsion outside of
the primary image area. For example, the DT Magnetic Carriers hold the film at four edges and have an empty window
for the image area, such that the carrier does not touch the emulsion within the image area. Such carriers must be
specifically sized, so for unusual film sizes DT can produce custom-sized magnetic carriers as special-order items.

Glass carriers, such as the DT ANR Glass Carrier can be used when magnetic carriers are not well-suited. For example:
• Physically-Distorted Materials: e.g. film that is too curled, warbled, or channeled to fit in a magnetic carrier
• One-Off Sized Materials: e.g. hand-cut film
• Thick Materials: e.g. glass plates and lantern slides

Ideally only one layer of glass should be used, with the film laying on top. This is sometimes workable for glass plates
and mounted film, but sheet film and roll film are rarely sufficiently flat for this method, and will require sandwiching
between two sheets of glass. When transmissive material must be sandwiched several precautions should be taken:
• Glass Weight: The top layer of glass should be as light as possible. Any handles or attachments to the top glass
should be of minimal weight. The DT ANR Glass Carrier is made in 5x7, 8x10, and 11x14 sizes; it’s best to select the
carrier that is the smallest size that accommodates the material to avoid excess weight.
• Parallel Decent: If the material has meaningful deformation that is keeping it from being flat, but is still stable
enough to be imaged under glass, then it’s best to lower the top glass down parallel to the bottom glass, rather than
pivoting it downward from a joint. This ensures the glass comes into even contact with the material rather than
producing a pinch point, and reduces shearing force. The DT ANR Glass Carrier provides a pseudo-hinge in the form
of a rubberized rear rail; the operator can use the pseudo-hinge to hold the top glass with one hand or remove the
top-glass, place the object, and then lower the glass parallel using two hands.
• Buffer Space: In some cases it may be prudent to place a set of “buffers” on the bottom glass so that the top glass
doesn’t come to a rest on the bottom glass, but comes to a rest shy of it, reducing the pressure it places on the
subject. Note that this inherently increases the maximum non-flatness of the film and can introduce issues with
depth of field which would require either reduced capture resolution or focus-stacking to overcome.

Cleaning and Dust

Dust is a significant challenge when digitizing film. In the era of film some scanners used a technology called ICE to
identify the location of dust spots and apply a healing algorithm to fill in the missing data. These technologies meant that
parts of the image were guessed at, which is against the ethos of cultural heritage digitization. It would be analogous
to finding a book where letters were physically covered up by dirt, and making a scan where the dirt was removed and a
best-guess was made as to what letter was covered.

Dust can either be embedded in the film, on top of the film, or on the equipment (e.g. on the glass of a glass carrier). A
film conservator may be able to treat a piece of film to remove dust embedded in the emulsion, but this is rarely time or
cost effective and carries risk. Dust on top of the film or on the equipment is much easier to deal with. Firstly, make every
attempt reduce the amount of dust in the air by careful selection of the workspace in which film digitization is done and
by the use of air filters. Secondly, remove as much dust from the film as possible by generous use of a hand-pumped air
“rocket” or a house-air system with a filter and moisture trap that is approved by your conservation team. Digital removal
of dust (automatic or manual) should not be performed on Object Reproduction masters or Content Reproduction
derivatives, but may be performed on Speculative Artist Renderings.
21
 Newton Rings
When two smooth surfaces come into contact they can form a colorful semi-concentric circular pattern called “Newton
Rings.” The visual symptom bears a passing resemblance to moire, but has a completely different cause. Newton Rings
form as a result of the way light reflects and refracts between the two surfaces, and the air gaps between them. The
rings are not part of the object or the content, and are highly undesirable both because they obfuscate the true content of
the object and because they may be misunderstood by the viewer as part of the physical object.

The chance of Newton Rings can be eliminated or reduced by several methods:

• Non-Contact: Using magnetic carriers (see chapter 3.5) wholly eliminates the chance of Newton Rings.
• Anti Newton-Ring (ANR) Glass: As the name implies this is a type of glass where the surface has been etched to
reduce its smoothness. Since Newton Rings form when two smooth surfaces touch, the rougher the surface of the
ANR Glass the less likely they are to produce Newton Rings. However, the rougher the surface, the more the texture
of the surface will interfere with imaging, causing the texture of the glass to appear in the image content, which
is undesirable. The DT ANR Glass carrier has been tested at up to 10,000ppi with minimal loss of sharpness or
introduction of texture.
• Anti-Reflective (AR) Glass: This glass is typically manufactured with the intention of reducing reflections. For
example, they may be used in a framed art work to decrease the “mirror-like” reflection of lights or the person viewing
the art. The anti reflective coating of the glass functions in a way that is similar to ANR glass, but with a typically finer
(less rough) surface topography that is less prone to reducing sharpness. Some testing has shown that AR glass
also reduces reflections between layers of glass when sandwiching transmissive material, increasing contrast and
effective dynamic range of captures.
• Wet Mounting: By filling the space between the glass with a fluid the glass and film become “one” and Newton Rings
are eliminated. This technique was used in the era of flat-bed scanners and drum-scanners. However, fluid contact
of any kind is strongly discouraged in modern Cultural Heritage Digitization.

Temperature
Film is typically stored at room temperature or colder (up to and including a deep freeze). As film changes temperature
the various layers may expand or contract at different speeds, resulting in deformation such as curling. Ideally, the
digitization process should change the temperature as little as is practically possible. For example, it’s typically not
practical to digitize film within a deep-freeze storage unit without removing it from that storage unit. However, it is
possible to limit the exposure to additional and unnecessary heat such as that of a physically hot light source.

Traditional light boxes used legacy light sources that generated a lot of heat. Placement on or near such a light will
expose the material to problematic amounts of heat. While this is unlikely to damage film unless the film is already
degraded (or the light is especially hot), the heat can cause temporary curling which will make it harder to hold the film
flat enough for standard digitization methods.

The DT Photon series uses cool-running LEDs and a fan that remains on 100% of the time. Furthermore the DT Film
Scanning Stage elevates the material being digitized several inches above the light surface, meaning film material will
experience no meaningful change in temperature. Note that additional DT Photon units (with or without the DT Film
Scanning Stage) can be used as stand-alone film-inspection stations so that such non-digitization functions also avoid
introducing the material to unnecessary heat.

22
23
 4
Recommended
Strategy

24
As outlined so far, the digitization of transmissive materials presents many goals, some are in direct conflict with
each other. For example, if a photographic negative is fading in density over time (due to physical deterioration) then
presenting the image as-is will properly serve a conservator, but may mislead a researcher into falsely thinking the artist
intended a dark moody image. Therefore the best strategy, presented in this chapter, is to create multiple derivatives
alongside the Preservation Digital Object, and to leverage metadata and presentation cues.

4.1 Multiple Derivatives

The digitization of transmissive collections cannot be said to fully fulfill the goals of digitization (surrogacy of the original
for most in-person uses) without the generation of multiple masters.

Here we use the words “multiple masters” to distinguish this set of files from traditional derivatives such as “web
friendly” or “watermarked” derivatives which retain the same image content but with different technical specifications
(e.g. different bit depth or resolution).

The authors propose three Master Files for each photographic original: an Object Reproduction Master File, a Content
Reproduction Master File, and a Speculative Artist’s Intention Master File. In most cases the Object Reproduction Master
and Preservation Digital Object are synonymous.

Object Reproduction Content Reproduction Speculative Artist’s Intended

(Preservation Digital Object) Rendering
White Balance Based on Light Source Neutralize based on content Follow the artist’s intentions as
best as they are understood.
Exposure Based on Light Source Adjusted if necessary

Crop Four-Edge Inclusion Crop to content This is best done by, or in

consultation with, someone
Tone Direction If negative, leave negative If negative, invert to positive familiar with the artist and
If positive, leave as positive If positive, leave as positive photographic processes.
Tone Range No adjustment Adjusted if necessary

Color No adjustment No adjustment

Local Adjustments None None

B+W Images Retain Color Force Monochrome

Dust* No adjustment No adjustment Remove where practical,

unless intended by artist
Scratches* No adjustment No adjustment

Degradation* No adjustment No adjustment

*Steps can be taken to reduce the impact of these physical issues before the capture is made. “No adjustment” here
refers to the fact that, once captured, no digital processing or retouching should be done to the captured file.

The Value of Doing Something over Nothing

Most guides written about film digitization do not include any recommendation similar to the Speculative Artist’s
intention. It’s impossible to know exactly how an artist intended their image to be rendered. Parts of the Film Ecosystem
are gone. Reference points such as prints and memory are faded. Our aesthetic context has shifted. Our tools are
fundamentally different than the artist would have used.

Given the impossibility of perfectly recreating the artist’s intended rendering some might decide that it’s not worth trying.
However, every factor just listed will worsen with time. Our ability to recreate the artist’s intentions might be middling
today, but is far better than it will be in the future. We are closer in time, technology, and aesthetic context to these artists
and the film ecosystem they used then our progeny. There are individuals alive today who worked with the specific films,
chemistry, paper, and processes used for much of the film in our institutions’ collections. Therefore, there is great value
in making our best guess now, as it is likely a better guess than future generations will be able to make.

That said, we must take caution to honestly present the results as the informed subjective inherently imperfect
speculation that they are, and we must still capture and preserve a Preservation Digital Object.
25
 4.2 Metadata
The topic of metadata is expansive and complex. Here we only discuss metadata in a narrow context: what information
about a transmissive object can be recorded as metadata to mitigate some of the challenges outlined in this chapter.
(E.g. loss of technical context and loss of aesthetic context).

As discussed in 3.2, a future viewer may not have even a basic understanding of photographic film. There are several
obvious areas (and probably many subtle areas missed here) that can help decrease the chance of confusion.

Object Nature
It’s important to record the basic type/category of object. For example, a capture of only the transmissive area of a
35mm mounted slide and a tightly-gated capture of a 35mm strip can look very similar to the end viewer. Only with a
metadata notation can the viewer be assured an understanding of the true nature of the original object.

Object Condition
Film can have a variety of condition issues that can lead the viewer to misinterpret the content of the image. It’s
important that metadata record these conditions to help the viewer interpret the object and content correctly. For
example, if 120 film rolls were not wound tight enough after capture then a light leak can lead to a bright artifact intruding
into the content of several of the images of that roll. A film-savvy operator of the digitization system can easily identify
such a leak because it is usually rhythmically repeated at the interval of the diameter of the rolled film. But a viewer who
has lost the technical context of film may mistake those light leaks as being either an intentional artistic treatment or,
worse, part of the scene itself (if the artifact fits reasonably with the content of the scene).

4.3 Presentation
The topic of how to provide access to a digital collection is, expansive. Our purpose here is only to bring to the fore the
unique needs of transmissive collections.

When displaying a transmissive object to a viewer we should help them understand both the object and the content.
This can be accomplished by providing immediate and intuitive access to viewing all three derivatives discussed in
4.1, providing the metadata referenced in 4.2 and providing links both to related objects and to relevant background
information about the media, process, and artist in question.

Showing Object Reproduction, Content Reproduction, and Speculative Artist’s Intended Rendering
The end user can benefit from seeing all three versions produced. These should be labeled and described in such a way
that someone who is not familiar with film can access an explanation of the nature of the master they are viewing and
linked to a more detailed explanation, such as this document.

Showing the Metadata

All the metadata described in 4.2 is largely wasted if it is not made easily accessible to the viewer.

Linking to Related Objects

Transmissive material often has built-in organization of associated images. For example an image on a roll of film is
often intimately related to the images that precede and follow it on that roll, and may often be equally related to images
on other rolls captured during the same shoot, day, or period. It is bad practice to present a single page from a book as a
stand alone island, without an easy ability for the viewer to understand where that page fits in the context of the book it
is from, and without an easy way for the viewer to navigate through the structure of that original object. In a similar vein,
film images should be presented in their physical context and the viewer accorded the opportunity to easily navigate that
structure.

Linking to Relevant Background Information

Keeping in mind that the future viewer may not have familiarity (discussed in 3.2 and 3.3), providing links to general
information relevant to the object may help inform their understanding of the object. While this is true for nearly any
museum, library, or archival collection it is especially true for transmissive materials. For example, when viewing a
Cibachrome a user should be directed to information about the Cibachrome process; this process rendered an image
made up of individual dots and this technical attribute of Cibachrome, if not described to the viewer, might lead a viewer
to make a mistaken assumption that the original scene was made up of points. Or if the image is known to be cross
26
processed information should be presented to the user to help them understand what cross processing was and how it
affected the content.

Note that IIIF (International Image Interoperability Framework) has taken on a myriad of image-presentation topics such
as those above, along with the broader topics of providing controlled and intelligent access to your collection assets. We
strongly suggest keeping up with this evolving specification.

27
 5
Method:
Object Reproduction

28
5.1 Select Lens & Extension Tubes
Different sizes of film will require different combinations of lens and extension tubes.
For the Phase One iXG and DT RCam: consult the charts below for the appropriate combination.
For the Phase One XF: use the Schneider 120LS Blue Ring lens with no extension tubes for all film formats.

To include all four edges of the physical film, a small margin or padding must be maintained. This chart assumes a very
small amount of padding around the subject; a larger margin/padding may sometimes requires less extension.

This chart has been simplified and rounded for easier reading. For a more detailed and precise chart email info@dtdch.
com and ask for the “DCH Distances and Measurements” document.

Phase One iXG 50 MP Phase One iXG 100 MP Phase One iXG 150 MP
Film Film Tubes Film Tubes
Format Lens Tubes Resolution Format Lens Extension Resolution Format Lens Extension Resolution

8 x 10” 72 mm None ~800 ppi 8 x 10” 72 mm None ~1100 ppi 8 x 10” 72 mm None ~1300 ppi
120 mm None ~800 ppi 120 mm None ~1100 ppi 120 mm None ~1300 ppi

5 x 7” 72 mm None ~1200 ppi 5 x 7” 72 mm None ~1700 ppi 5 x 7” 72 mm None ~2100 ppi

120 mm None ~1200 ppi 120 mm None ~1700 ppi 120 mm None ~2100 ppi

4 x 5” 72 mm 21 mm ~1600 ppi 4 x 5” 72 mm 21 mm ~2200 ppi 4 x 5” 72 mm 21 mm ~2600 ppi

120 mm None ~1600 ppi 120 mm None ~2200 ppi 120 mm None ~2600 ppi

6 x 7 cm 120 mm 42 mm ~2600 ppi 6 x 7 cm 120 mm 42 mm ~3700 ppi 6 x 7 cm 120 mm 42 mm ~4500 ppi

6 x 6 cm 120 mm 42 mm ~2600 ppi 6 x 6 cm 120 mm 42 mm ~3700 ppi 6 x 6 cm 120 mm 42 mm ~4500 ppi

645 120 mm 63 mm ~3500 ppi 645 120 mm 84 mm ~5400 ppi 645 120 mm 84 mm ~6000 ppi

35mm 120 mm 84 mm ~4700 ppi 35mm 120 mm 84 mm ~5400 ppi 35mm 120 mm 84 mm ~6000 ppi

Notes:
- The iXG 72mm can technically be used for film formats smaller than 4x5” but not with FADGI-4 Star performance.
- 63mm extension is achieved by stacking a 42mm and 21mm tube.
- 84mm extension is achieved by stacking a 21mm, 42mm, and 21mm tube, in that order. Capture One reports the wrong PPI with this much extension.

DT RCam 50 MP DT RCam 100 MP DT RCam 150 MP

Film Film Tubes Film Tubes
Format Lens Extension Resolution Format Lens Extension Resolution Format Lens Extension Resolution

8x10” 72 mm None ~800 ppi 8 x 10” 72 mm None ~1100 ppi 8 x 10” 72 mm None ~1300 ppi

72 mm None ~1200 ppi 105 mm None ~1100 ppi 105 mm None ~1300 ppi
5 x 7”
105 mm None ~1200 ppi 5 x 7” 72 mm None ~1700 ppi 5 x 7” 72 mm None ~2100 ppi

72 mm 20 mm ~1600 ppi 105 mm None ~1700 ppi 105 mm None ~2100 ppi
4 x 5”
105 mm None ~1600 ppi 4 x 5” 72 mm 20 mm ~2200 ppi 4 x 5” 72 mm 20 mm ~2600 ppi

120 mm 20 mm ~2600 ppi 120 mm None ~2200 ppi 120 mm None ~2600 ppi
6 x 7 cm
105 mm 20 mm ~2600 ppi 105 mm None ~2200 ppi 105 mm None ~2600 ppi

120 mm 20 mm ~2600 ppi 6 x 7 cm 120 mm 20 mm ~3700 ppi 6 x 7 cm 120 mm 20 mm ~4500 ppi

6 x 6 cm
105 mm 20 mm ~2600 ppi 105 mm 20 mm ~3700 ppi 105 mm 20 mm ~4500 ppi

120 mm 40 mm ~3500 ppi 6 x 6 cm 120 mm 20 mm ~3700 ppi 6 x 6 cm 120 mm 20 mm ~4500 ppi

645
105 mm 40 mm ~3500 ppi 105 mm 20 mm ~3700 ppi 105 mm 20 mm ~4500 ppi

120 mm 80 mm ~5800 ppi 645 120 mm 60 mm ~5400 ppi 645 120 mm 60 mm ~6000 ppi
35mm
105 mm 80 mm ~5800 ppi 105 mm 60 mm ~5400 ppi 105 mm 60 mm ~6000 ppi

35mm 120 mm 120 mm ~8200 ppi 35mm 120 mm 120 mm ~9,900 ppi
105 mm 120 mm ~8200 ppi 105 mm 120 mm ~9,900 ppi

Note: 20, 40, and 60mm RCam Extension Tubes can be stacked in any order to achieve the desired total.
29
 5.2 Set Camera Position & Focus
Load the film into the carrier and place the carrier in the DT Film Scanning Stage. The scan resolution (aka PPI) and scan
area (aka Field Size) are determined by the camera position (aka camera-to-subject distance). In the previous step you
selected the appropriate lens and extension tube. Now you must position the camera at the correct distance from the
film, and focus the lens. The steps to do so depend on your equipment.

With a DT AutoColumn + Phase One iXG

Simply enter the desired Frame Size and push “Start”. For more details see the document “DT AutoColumn with iXG
Setup”.

With a DT AutoColumn + Phase One XF or DT RCam

The first time you set up for a given type of film you will need to guess-and-check the appropriate camera-to-subject
distance. Technically this distance can be mathematically calculated, but that’s typically slower in practice. To Guess-
and-Check:

1) Open Live View using the icon in the Camera tool, or by the dropdown menu [Window > Live View]

2) Using the Copy Stand Tool, move the camera up/down until the subject is in focus.

3) If the subject has too much margin: Move the camera closer. If it does not have enough margin or is partially cropped.

4) Zoom to 100% somewhere in the center. Click the Focus Meter cursor and then click in a area of the film. Ideally pick
an area of the film with contrasty subject matter or film grain rather than a deep shadow or highlight. Details that are
focused properly have more contrast than details that are mis-focused. The Focus Meter shows you the current level of
contrast as well as the high-water-mark for contrast; at first these will be the same.

5) Slowly focus the camera to-and-then-beyond the perfect point of focus; as you move beyond the perfect point of focus
the current-contrast-line will recede away from the high-water-mark. Finally, slowly return the focus until the current-
contrast-line returns to the high-water-mark. This will be the point of perfect focus.

6) Repeat steps 3-5 until you’ve reached the exact right framing. As you get closer to the desired framing you will find
that using the Column Control tool to move the camera up and down in single steps will become a more useful focusing
method than using the focus of the camera.
30
7) Using the Column Control tool, take note of the resulting camera position (typically shown in centimeters) in order
to easily return to this position in the future. Note that the PPI readout of this tool is only valid for a single combination
of lens and extension; therefore, if you have the PPI calibrated for the 72mm the readout will not be accurate for the
120mm. For more details see the document “DT AutoColumn with DT Rcam or Phase One XF Setup”.

31
 5.3 Set Mode, Profile, and Curve

Mode: Set your mode to Film Positive. This should be used even when imaging tonally-inverted materials (e.g. color
negative film) because an Object Reproduction should show the object as it would appear on a lightbox.

Profile: For transmissive media digitization, choose the specialized ICC-Profile in Capture One CH for your light source.
In most cases this will be the DT Photon.

Curve: Set the curve to Linear Scientific. This provides absolutely linear tonal values, without any contrast added or any
compression of the tones present in the original negative. Do note that using Linear Scientific does not gracefully handle
clipped highlight values. Therefore it’s especially important to avoid any clipping during capture and adjustment. Since
DT systems have a very large dynamic range it’s possible to capture the entire tonal range of nearly any transmissive
material, without clipping the highlights or shadows.

You can optionally use [... > Save As Default] to save the profile, mode, and curve as the defaults.

5.4 Set Sharpening & Noise Reduction

As a starting point, set the sharpening as follows:
Amount: 100
Radius: 1
Threshold: 0
Halo-suppression: 0

An amount of 100 provides modest visual sharpening without over sharpening the image. In the Quality Control Chapter
we discuss sharpening in more detail.

A threshold of 0 means that the sharpening amount will be applied homogeneously across the entire image. Because we
are effectively photographing the structure of the film grain itself, a threshold of 0 also prevents the artificial masking and
sharpening of the image content recorded on the film.

Set your Luminance Noise Reduction as follows:

Luminance: 0 for CMOS backs, 50 for CCD backs
Details: 50
Color: 40 for CMOS backs, 50 for CCD backs
Single Pixel: 0

CMOS sensors (50mp, 100mp, and 150mp digital backs) inherently have a very low noise floor. Reducing the Luminance
noise reduction to 0 prevents the software from artificially smoothing over the structure of the film grain.

You can optionally use [... > Save As Default] to save these settings as the defaults

5.5 Capture LCC Raw File

Even a high-quality light source like the DT Photon will have very slightly variations in brightness across the surface
of the light. In addition, even high-quality lenses used in DT systems are slightly vignetted in the corners. While these
variations are typically very small, they are worth correcting in the context of preservation-grade digitization.

32
Remove the film and the film carrier together from the positioning stage. Replace the carrier with a piece of matte,
translucent Plexiglas seated in the same plane of focus as the film.

Make a capture of the Plexiglas, ensuring it is unobstructed and fills the entire field of view. Adjust the exposure and
recapture as needed, so the resulting raw is around middle gray.

Rename the file so the name reflects the current capture setup (e.g. “35mm at 6000ppi”).

33
 5.6 Analyze LCC
You can now generate an LCC profile from the image of the Plexiglas. There are two possible ways to begin:

Method One

Right-Click on the thumbnail of the gray cardboard and select Create LCC.

Tip: On a Mac without a right mouse button, use “CTRL + mouse click” for entering the context menu.

Method Two

Via the Lens tool tab select the image of the captured gray cardboard in the Browser and choose Create LCC in the LCC
tool.

After pressing Create LCC (either started with the first or second method), a dialog box will open showing the following
options for the LCC profile that will be calculated. DO NOT check any of the options. Press Create to start the calculating
process.

After calculating you should see the selected reference image of the captured gray cardboard marked as a LCC reference
file. The letters LCC will appear above the thumbnail image. In the LCC tool the boxes for Color Cast and Enable Uniform
Light should be checked and available. You have now created a valid LCC Profile for this capture setup.

Note:

An LCC profile is only valid and usable for images taken with the same Digital Back or Camera under identical lighting
conditions. You have to prepare a new LCC Profile to create luminance uniformity if you:
- Change the aperture
- Change the distance to the subject
- Change the angles of the light setup
- Change the lens or digital back

Tip:

You may wish to rename the capture of the gray board before you generate the LCC from it. For example, you could name
it “72mm, f8, 600ppi distance”. This will be the name of the LCC that is subsequently applied to incoming captures and
will be more intuitive.

5.7 Apply LCC

In the Camera tool tab there is a tool called Next Capture Adjustments. When shooting tethered with the camera
attached to the computer, you can choose the settings and adjustments to be applied to the next capture. The factory
default setting is useful in this example to automatically apply the LCC profile.

The first option – ICC Profile is set to Default in the factory default settings so it will choose the ICC Profile you set in the
Base Characteristics tool (See also Chapter 3.2)

The All Other option is set to Copy from Last by default. This will ensure that every setting that is made with one of the
Tools in Capture One is automatically applied to the next captured image. This includes the newly created LCC profile
from the last image. All subsequent images will be precisely corrected, using this LCC, until you make changes to the
lighting.

34
5.8 Set White Balance
When doing Object Reproduction you should white balance to the light source.

Remove the film and film holder from the positioning stage and photograph the light source directly. Adjust your
exposure so that the light source renders roughly middle gray.

The White Balance picker is located in the Cursor Tools. It can also be found in the Color Tool Tab in the White Balance
Tool or in the Capture Tool Tab in the Camera Tool. Using the White Balance picker, click your white balance in the center
of the frame.

Do not place the film back onto the stage yet. Leave it off for step 5.8.

5.9 Set Exposure

For an Object Reproduction (the Preservation Digital Object) the exposure is set to accurately reflect how the object
would appear in in-person use on a light table or other viewing device. In some cases this means the content itself will
be impractical to view, for example a very dense glass plate negative may appear nearly black or a thin negative (one
that was underexposed or under-processed at the time of creation) will appear both thin and low in contrast. However,
the object reproduction should not attempt to correct for these situations because any such correction would obfuscate
the true nature of the object. For example, a researcher may benefit from knowing that a series of negatives was thinner
or denser as it may inform the researcher about the methods the photographer used, or the conditions in which the
photographer was working. The creation of the Content Reproduction (in the next chapter) addresses adjusting exposure
as needed to provide for practical viewing of the content itself.

Select the Add Color Readout cursor by clicking and holding on the white balance cursor. Using this cursor, place a color
readout several places throughout the frame, including near the center. If the readout is in Red/Green/Blue then select
[Lab Readout > Golden Thread > ICC] to switch to LAB readouts. Adjust shutter speed until the brightest luminance value
of the LAB Color Readouts is as close to 98, without going over, as possible.

5.10 About the Crop

In step 5.2 we carefully established a specific frame size for the current material type. Typically no further cropping
should be required for the Object Reproduction (Preservation Digital Object) since most cases transmissive materials
are fairly homogeneous in size within a given type of material. For example, the variation among 4x5 sheet film negatives
is typically quite small. The fact that we’re using the entire frame (with only a small margin reserved to ensure all four
edges are shown and to ensure it is practical to quickly place additional material within the same frame) maximizes
resolution and eliminates the need to set a crop for the Object Reproduction. In chapters 6 and 7 further discussion of
crop is provided in the context of Content Reproduction and Speculative Artist’s Intended Rendering.

5.11 Ready to Capture

The system is now properly set up for scanning any transmissive material that fits within the field size established
in step [5.2] and for which the current PPI is acceptable. Digitizing additional material with these settings is very fast,
whereas changing to another field size or PPI requires repeating steps 5.1 to 5.11. Therefore it is desirable to collate your
digitization such that like-sized materials are captured in batches. More information on this topic is in our Digitization
Program Planning Guide in the chapter titled “Planning Collection Pulls.”

35
 6
Method:
Content Reproduction

36
As discussed in Section 2.3.2, several derivatives be derived from transmissive media to fulfill digital surrogacy.
• Object Reproduction: A faithful reproduction of the entire physical object, as it would appear on a light table.
• Content Reproduction: A human-readable rendering of the image contained by the object.
• Speculative Artist’s Intention: A best-guess at how the artist would have rendered their final work.

In chapter X we created a Object Reproduction Preservation-Digital Object (aka “master file”). In this chapter we look at
how to create Content Reproduction derivatives using the tools in Capture One CH.

This does not require recapturing the physical original so steps 5.1 and 5.2 do not have an equivalent in this chapter. In
addition, there is no change to sharpening, noise reduction (5.6) or LCC (5.7, 5.8, 5.9) so those sections are not repeated
in this chapter.

6.1 Clone Variant

Select the Object Reproduction and select [Image > Clone Variant] or use the keyboard shortcut F3. This creates a virtual
copy of the raw file, so does not increase disk space. Later this virtual copy (variant) can be processed (a.k.a. “Exported”)
to a TIFF or JPG separately from the first variant. This allows you to create an Image Reproduction derivative without
having to capture the film material again.

6.2 Set Mode

For Object Reproduction (reminder)
For the Object Reproduction we used Film Positive no matter the kind of transmissive material we were digitizing.

For Content Reproduction of negative materials

Set your mode to Film Negative to invert the colors and tones of the image so that the user can view the negative as a
positive. This retains the intuitive functionality of the image adjustment tools, in that moving the exposure slider to the
positive direction brightens the image, while sliding it towards the negative direction darkens the image.

37
 For Content Reproduction of positive transmissive materials
Set the mode to Film Positive. While at first glance, Film Positive will not seem all that different from Photography,
this mode is aligning the color channels in a way that is more in tune with the non-linear tone/color distribution of
photographic film.

6.3 Set Curve

For Object Reproduction (reminder)
For the Object Reproduction we used Linear Scientific to provide a 100% tonally faithful version of the physical material.

For Content Reproduction

Set the curve to Linear Response. This provides a very slight roll off in the highlights that maintains the color/hue of
bright subject matter that clip in one channel. While this is technically (very slightly) less scientifically accurate it is more
practical as it’s less likely to lead to technical artifacts such as banding or abrupt tonal clipping.

6.4 White Balance

For Object Reproduction (reminder)
For the Object Reproduction we white balanced based on the light source. That ensured that any color bias in the subject
itself was faithfully reproduced. For example, when digitizing black and white negative film the slight non-neutrality of the
film itself is correctly captured by white balancing to the light source.

For Content Reproduction

Inside the image itself, locate a matte-finish object that is likely to have been roughly neutral in tone at the time of
capture. Consider looking for asphalt, gray fabrics, paper, or other neutral materials with matte-like surfaces. Avoid shiny
or reflective surfaces; for example, a gray car might, at first, seem like a good candidate for white balance, but since the
paint is glossy, it is likely to be reflecting blue the sky, brown from a brick building, etc—all of which would result in an
incorrect white balance for the scene.

The surface you select should be illuminated by the scene’s primary light source (or “key light”). If a scene has two
sources of light with different color temperatures the key light is the one that is strongest on the main subject of the
image. For example, in an early morning sunrise image there are two sources of light: the sun itself that is harsh and
warm and the blue dome of the sky that is diffuse and cool; if the main subject is being illuminated by the sun then the
sun would typically be considered the key light.

By white balancing to neutral subject matter illuminated by the scene’s key light we are (roughly) neutralizing the
overall scene color. In some cases this may result in a dramatically different result than the Object Reproduction image,
especially in cases such as:
• Film whose color has shifted over time due to physical deterioration
• Tungsten-balanced film shot in daylight scenes
• Daylight-balanced film shot in tungsten-scenes
• Scenes with unusual color balance like sunrise, sunset, stage lighting, or night scenes

38
6.5 Set Exposure
For Object Reproduction (reminder)
For the Object Reproduction we set the exposure based on the light source so the object appeared as it would on a light
table.

For Content Reproduction

In some cases the content will require a change to exposure or contrast to serve as a practical Content Reproduction.
For example, if the original physical material is a very thinly exposed negative then the inverted image may be too dark
for the viewer to easily understand the content. However, unlike the Speculative Artist’s Intended Rendering the Content
Reproduction derivative is not intended to interpret or alter the underlying nature of the image content, only to ensure
that it is visually accessible.

Using the exposure slider make the minimum change required for the viewer to visually understand the content.

6.6 Set Crop

For Object Reproduction (reminder)
For the Object Reproduction we did not crop the image, so that all four edges of the physical object were retained.

For Content Reproduction

For content reproduction it is appropriate to crop such that the final image does not include any borders or edges around
the image content. By definition this means a very small amount of the content will be cropped.

39
 7
Method:
Speculative Artist’s Intended
Rendering

40
As discussed in Section 2.3.2, several derivatives be derived from transmissive media to fulfill digital surrogacy.
• Object Reproduction: A faithful reproduction of the entire physical object, as it would appear on a light table.
• Content Reproduction: A human-readable rendering of the image contained by the object.
• Speculative Artist’s Intended Rendering: A best-guess at how the artist would have rendered their final work.

In chapters X and Y we created a Object Reproduction Preservation Digital Object (aka “master file”) and Content
Reproduction derivative. In this chapter we look at making a Speculative Artist’s Intended Rendering derivative using the
tools in Capture One CH.

7.1 Clone Variant

Select the Content Reproduction variant and select [Image > Clone Variant] or use the keyboard shortcut F3. This creates
a virtual copy of the raw file, so does not increase disk space. Later this virtual copy (variant) can be processed (a.k.a.
“Exported”) to a TIFF or JPG separately from the Content Reproduction variant. This allows you to create an Speculative
Artist’s Intended Rendering derivative without having to recapture the physical originals, and without having to repeat
steps taken in chapter 6.

Because the Speculative Artist’s Intended Rendering starts from the Content Reproduction Variant it does not require
recapturing the physical original so steps 5.1 and 5.2 do not have an equivalent in this chapter. There is no change to
sharpening, noise reduction (5.6) or LCC (5.7, 5.8, 5.9) so those sections are not repeated in this chapter. There is also no
change to Mode (6.2) or Curve (6.3) or White Balance (6.4) so these sections are not repeated in this chapter.

7.2 Research the Artist’s Intentions

The Speculative Artist’s Intended Rendering is meant to produce an image similar to the one the artist themselves would
have presented. Before we can pursue the creation of a Speculative Artist’s Intended Rendering we need to research,
to the extent practically possible, what methodological, technical, and aesthetic intents the artist would have likely
employed. Potential sources for this information are listed below, starting with the sources most likely to provide useful
and accurate information:
• Direct conversation with the artist themselves
• Prints or other outputs made by or approved by the artist
• Documentation of prints or exhibitions (e.g. photos of gallery shows)
• Notes or writing of the artist
• Historical interviews of the artist
• Consultation with experts such as curators or researchers
• The work of similar artists
• The work of other artists who used the same film
• The work of artists during the same time period
41
 There are many practical limitations here. The artist may be deceased with no surviving prints, interviews, or notes.
The work may not have a known artist. The film used may have been obscure or hand-made or the process created by
the artist without documentation. Even in the best case, we are still limited to trying to recreate the rendering the artist
would have intended using a tool-set that is fundamentally different than the tool set the artist would have employed.
Nevertheless we cannot allow the perfect to be the enemy of the good. The motivation for creating a Speculative Artist’s
Intended Rendering is that something is better than nothing, especially given the strong likelihood that, in the future, even
less will be understood about the Artist, Film Ecosystem, Aesthetic Context, and Technical Context of Film.

7.3 Color and Tone

For Content Reproduction (reminder)
For the Object Reproduction we used a white balance derived from the subject. This neutralized the scene so that subject
matter that was neutral in the real world would also be neutral in the scan. We also used the minimum adjustment to
exposure that was required to make the content visually understandable to the viewer.

For Speculative Artist’s Intended Rendering

We will leave the white balance set as it was in the Content Reproduction and use exposure, levels, curves, and Color
Editor to adjust the color to be in line with artist’s intended rendering.

There are several common sources of differences between the global color rendering of the Content Reproduction and
the intention of the artist.
• Physical Deterioration: The color and tone range of the physical object may have drifted over time (see chapter 3.4)
• Non-Neutral Intention: The Content Reproduction image was white balanced based on the image content,
producing normal content. But a non-neutral intention was very common. For example, an artist may have intended
an image at sunset to be warm.
• Film Ecosystem: Many films were coupled with a specific printing process to produce a specific look. Without that
printing process the color and tone will not match the intent without intervention.

Each of these categories calls, primarily, for it’s own tool in Capture One.

Physical Deterioration & Film Ecosystem: Auto R,G,B Levels

The Levels tool is often helpful either when the color of film has shifted due to aging or improper storage or as a first
stage in matching a particular film ecosystem. See Chapter 8.3 for more information.

In [Capture One > Preferences > Exposure] select “Channel Mode > Red, Green, Blue Channels” and 0.01% for Auto Level
Clipping Threshold for both Shadows and Highlights.
42
Crop the image such that only image content is visible. Neither the frame of the image, nor the light source under the film,
should be visible in the crop.

If needed, use the Exposure Slider to move the histogram is centered and no content is clipped in the highlights or
shadows.

Use the White Balance cursor to roughly white balance the image using something reasonably neutral inside of the frame.

Click the Auto button in Levels. This will adjust each channel (red, green, and blue) such that the image now has a neutral
black and neutral white. This automatic adjustment works best when the image content includes subject matter that
should be black and subject matter that should be white. Images where the subject matter itself is intentionally low
contrast or has a strong color cast; for example, a photo of a boat in fog should not have a true black or true white, but
this automatic level adjustment will force both. In such cases manual adjustment of the levels may be required.

In the case that many similar frames are captured at the same time it may be more consistent and accurate to identify
the frame that contains the most neutral shadow and highlight subject matter, auto adjust to that frame, and then copy-
apply those adjustments to the other frames.

43
 Non-Neutral Intention: Color Balance
It’s very common that the artist’s intention would be a warm or cool image, or even one that had an unusual color cast
such as magenta or cyan. Rather than try to reproduce these intentions using Levels or White Balance we will do so
using the Color Balance tool. This moves the adjustment later in the process and makes it easier to isolate it from other
color-impacting adjustments.

Use the Master Color Balance tool to imbue the desired color balance on the overall image. If needed, use the 3-Way
Color Balance tool to tweak the color balance of just highlights, mid-tones, and shadows.

Film Ecosystem: RGB Levels, Curves and Color Editor

The nuanced color reproduction of specific films used with specific paper and printing processes are impossible to
recreate with 100% accuracy. However, the broad strokes of such combinations can often be achieved through fairly
straight-forward adjustments using the RGB Levels, Color Editor and Curves.

The levels tool was described earlier and is a common first step in this process.

The Curves tool in Capture One functions much the same way it does in other image editing software. By adding points
to the curve you can brighten or darken specific tonal regions. Note that the steeper the line is in a given tonal region, the
more contrast that tonal region will have. In most cases, because the dynamic range of DT systems is so high, the scan
will be lower in contrast than is visually desired for an Speculative Artist’s Intended Rendering, even after using an Auto
Level command. It is therefore common to add contrast using the Curves tool.

44
Go to the Color Editor Advanced tab and use the Pick Color Correction Cursor to select a color that needs to be
manipulated. Turn on View Selected Color Range to see what range of color is effected by the current selection.

Modify the four edges of the selection wedge and the smoothness slider until the desired color range is selected in
the image. Generally the Smoothness slider should be left as high as possible to reduce the specificity with which the
adjustment is made. This reduces the likelihood of artifacts and makes it more likely that the adjustment can be copy-
pasted to other similar images. However, lower smoothness may be required to limit the selection to a specific color.

Turn off View Selected Color Range and modify the color’s Hue, Saturation, and Lightness to better match the intention
of the artist. Do not be surprised if some of the adjustments are fairly strong; many film era images departed from a
“neutral” rendering quite profoundly.

In the case of modern color negative film with ongoing commercial availability there is also a profiling process that DT
can provide. Please contact DT for more information.

45
 8
Batch & Automatic Processing

46
In chapters 5, 6, and 7 we discussed how to capture a Object Reproduction (Preservation Digital Object), Content
Reproduction, and Speculative Artist’s Intended Rendering. Those chapters focused on working on a single image at
a time. In practice, transmissive collections are often expansive in scale, so their digitization must leverage batch and
automatic processing. In this chapter we discuss the many such tools and workflows in Capture One CH.

8.1 AutoCrop
Capture One CH can automatically crop many transmissive materials, saving the operator the time and tedium of manual
cropping. However, transmissive materials present a complicated task for automatic cropping; such material often
comprises cascading/concentric rectangles and are often missing edges or corners (e.g. an underexposed area in the
corner of a frame may bleed into the unexposed base without presenting a visible corner).

Autocrop in Capture One CH is a tool that applies a crop to the selected raw file variants. Capture One CH is used for all
three workflow steps: applying an automatic crop, QC’ing the resulting crops, and manually correcting any errant crops.
This is far preferable to other automatic cropping tools that run on TIFFs and often require going backwards through the
workflow in the case that an errant crop is uncovered.

Loose Material
This method is the most generic, and looks for the largest contiguously-enclosed shape within the current crop. This can
be useful when using the ANR Glass Carrier to digitize sheet film or Glass Plate Negatives.
1. Select all relevant frames
2. Enter your desired padding. Note that this can be a negative or positive value.
3. Click AutoCrop
4. QC the resulting crops, manually adjusting the crop where needed

Roll Film
This method is specifically designed to address some of the complications presented by roll film. It requires the user to
provide an example crop, which the algorithm them attempts to repeat on additional frames. One unique advantage of
the Roll Film method is the confidence feedback it provides which helps the user more quickly QC the resulting crops and
fine-tune their cropping workflow.

1. Select a frame that is representative of the overall group of images you wish to crop and which has four cleanly
defined edges
2. Very carefully crop this frame to the relevant subject matter. For a Content Reproduction that means the crop edges
should perfectly align to the edge of the exposed image frame. Padding will be added in a later step.
3. Click [Set Master Crop]
4. Enter your desired Padding. Note that this can be a negative or positive value.
5. Select all relevant images
6. Click AutoCrop

Each image will then be tagged according to how confident Capture One CH was in the crop applied.
Red: Low Confidence. The assigned crop is very likely to be wrong.
Yellow: Medium Confidence. The assigned crop may be wrong.
Green: High Confidence. The assigned crop is likely correct.
7. QC the images according to their tag. Using the filter tool filter the current view to red-tagged and yellow-tagged
images and carefully check each crop with the assumption that each is wrong. Then filter the view to green-tagged
images; while the crop of each of these images should be checked, they can be checked at a faster rate, and in some
cases can be checked using the thumbnail-view only.

For roll film in good condition, and with clearly defined frame edges, the Roll Film algorithm should produce mostly
green-tagged and correctly cropped images. If not, please contact DT’s support team (support@digitaltransitions.com)
so we can help you troubleshoot.

47
 AutoCrop Tips: Pre-Crop ROI
AutoCrop (any mode) always begins with the currently assigned crop. Sometimes it is advantageous to apply a generic
crop across a range of images, before engaging autocrop. For example, in the image

1. Crop the first image to the region you know the subject will lie within for all relevant images (e.g. crop out the top
100px and the left 200px)
2. Select-all relevant images
3. Copy-apply the crop using the local copy-and-apply in the crop tool
Run the autocrop tool.
Using the above workflow the AutoCrop will only search within the crop established in step 1; therefore anything outside
that crop will be ignored when searching for the subject.

AutoCrop Tips: Contrast Enhancement

Slide film that is only transilluminated may have dark content at the edge or corner of the frame. However, this dark
content is rarely as dark as the black non illuminated slide casing around it. Therefore you can temporarily increase the
visual contrast between the slide and the slide casing by using the Levels or Curves tool. Likewise, when using the ANR
Glass Carrier to digitize sheet film, the contrast between the edges of the film object and the unobstructed light source
can be temporarily increased.

1. Select all relevant/similar raws

2. Toggle the viewer to Primary View
3. Adjust the curves or levels to increase the subject and background
4. Use the Local Copy-Apply button to sync this adjustment to all other selected raws
5. Run AutoCrop
6. Use the Local Reset button to remove the adjustment from step 4
7. QC the resulting crops and manually adjust any failed crops

The tool used to temporarily increase the contrast can be reset after the automatic cropping by using the reset button.

48
8.2 Modify Crop & DT Crop Control CH
Once AutoCrop or manual cropping has been used to establish an initial crop on a variety of images, the modify crop
and DT Crop Control CH suite can be used to modify those crops. This is preferable to standard copy and apply methods
because those would overwrite the unique current crop location and size of each image rather than starting with the crop
as it already is.

Modify Crop
Modify Crop is a native command of Capture One CH. It provides resizing and relocation, starting with each image’s
current crop. It provides less flexibility and power than the DT Crop Control CH Suite, and cannot be triggered by
keyboard shortcut. However it may be more intuitive than the DT Crop Control CH Suite for first-time users.

Start by selecting all relevant images, and making sure that [Edit > Edit All Selected Variants] is enabled (orange). If this
option is not enabled (grey), then only the Primary selected image will have the crop changed.

In the top right of the Crop or AutoCrop tools, click the [...] button.

• Location: To change the location of the crop, enter a value in the two Location fields, the first handles a horizontal
move and the second a vertical move. Note that positive values will move the crop right and downwards, while
negative values (adding a minus to the value) will move the crop left and upwards.
• Size: To change the size of the crop, enter a value in the two Size fields. The first changes the width and the second
the height of the crop. Positive values make the crop larger, while negative values will make it smaller.
• Anchor: Choose where the current crop should be anchored when changing the size of the crop by clicking on one
of the nine points in the Anchor box. By default, the changes will be done from the center out. By choosing the top-
right corner, for example, it is possibly to make sure that all changes to the size happen in the opposite direction,
e.g. to left side and bottom of the crop.

Click on Modify to apply the changes.

49
 DT Crop Control CH Suite
This suite of cropping tools can downloaded from dtdch.com/dt-crop-control-ch/. The below commands are available
out-of-the-box and can be activated by keyboard shortcut.

• Resize: Expand or contract the height or width of the crop on the selected variants without moving the center.
Useful for modifying the AutoCrop result to be asymmetrical (e.g. more padding on the left/right than on the top/
bottom).
• Trim: Expand or contract one side (e.g. top, bottom, left, or right) of the crop on the selected variants without
moving the other three sides. Useful for including Object Level Targets which are below the AutoCrop’d subject.
• Set Size: Set the width or height of all selected crops to a specific value without moving the center of the crop.
• Standardize: Adjust the width or height of all selected images to match the largest width or height of those images.
For example if you select four variants which are currently cropped with widths of 400, 350, 550, and 500, all four
images can be set to a width of 500, without changing the center of their crops.
• Duplicate: Duplicates the selected variants. When this command is used by itself it is no different than using Clone
Variants in the standard Capture One interface. However, this command can be part of a chained sequence of
commands to facilitate automatic page splitting and gutter sweeping.
• (Page) Split: When using the DT RGC180 to image two-up facing pages of bound material the AutoCrop tool’s
Book Scanning mode does a great job of cropping the frame to the page-spread. For those clients that would like
a crop for each of the two pages this tool splits the crop to it’s left and right halves. That creates a 0 pixel gutter
in the dead center throughout the book; the user can then control the size, location, and sweep of the gutter with
subsequent tools. Note that this tool assumes the selected range of images already have two variants each (i.e.
that the Duplicate command has already been used).
• Gutter Overlap: Increases or decreases the amount of overlap between the left and right crops. This assumes the
selected range has already been Duplicated and Page Split.
• Gutter Shift: Shifts the gutter left or right by a fixed amount without changing the size of the gutter. This assumes
the selected range has already been Duplicated and Page Split.
• Gutter Sweep: Progressively sweep the gutter to the left or right. When scanning thick bound material it’s common
that the physical gutter slowly moves from the left-center to right-center rather than staying in the absolute center
of the book.

This suite is open source and the user is encouraged to modify the code to fit their specific needs. More information,
including a tutorial on modifying the code, can be found at dtdch.com/dt-crop-control-ch/.

50
8.3 Auto Levels
Note that the Auto Levels tool used in steps 7.3 can be used for batch automatic adjustment. When selecting more
than one image. Each image will be assessed on its own and receive it’s own levels adjustments as if it was selected
individually. Since the histogram of each image must be individually analyzed running Auto Levels on large batches of
images may take several seconds or even minutes.

In [Capture One > Preferences > Exposure] Auto Levels can be set to run in one of two modes:

• Red, Green, Blue Channels: Each color channel is treated independently. If the image has a color cast in the
highlights or shadows, this mode will often successfully reduce or eliminate that cast. This is very useful for
adjusting color negatives. Note that if the subject matter itself does not have neutral subject matter in the
highlights and shadows using this mode can introduce a color cast. For example, in a photo of tightly-framed
green grass (lacking neutral content like a cement path) this mode is likely to introduce a slight magenta cast to
counteract what it sees as a green-cast.
• RGB Channels: This mode operates on the RGB composite channel. It therefore does not modify color, only
contrast. This is especially useful for monochromatic subject matter (e.g. black and white negatives) or subject
matter where Red, Green, Blue proves problematic (e.g. the green grass image described above).

In [Capture One > Preferences > Exposure] Auto Levels can be set to use a clipping threshold between 0.01 and 10% for
the highlights and shadows. This much of the subject matter is allowed to be clipped. Generally a lower value is preferred
so that very little of the image is clipped to black or white, since clipping causes a loss of detail. However, a higher value
can be used to provide a greater chance that a very small image area will throw off the automatic removal of color cast.

51
 8.4 Copy and Apply
The Copy and Apply tool is a basic tool for batch workflows. In other programs this functionality is called “synchronize
adjustments” or “Lift and Stamp” or “Copy and Paste Adjustments”. This allows you make adjustments such as changes
to color or exposure, on one image and quickly apply those same adjustments to other images. There are two main ways
to copy and apply adjustments in Capture One

Global Copy and Apply

When using the copy and apply icons in the top of the user interface all adjustments will be copied from the primary
selected image and applied to the other selected images.

Local Copy and Apply

The copy-apply icons located in each individual tool allow you to quickly copy-apply just that tools settings to the other
selected images.

For both Local and Global Copy and Apply image parameters that have not be adjusted from their default values will not
be copied; for example, if the primary selected image has not been cropped then a global copy-apply or local copy-apply
of the Crop Tool will not change the crop of the other selected images.

8.5 Naming and Folder Structures

Capture One provides a variety of tools that support advanced naming and organization workflows. These use “tokens”
which are placeholders for metadata that will be completed at the time of capture or at the time of export/processing.
These tokens are useful for setting Next Capture Naming and Process Recipe Output Folder. Images can be captured
with tokens used to form the resulting file names, and exported with tokens used to modify the names or folder-structure
of the exported files.

Useful Tokens
• [Document Name]: This is the name of the current session. This token is well suited for information representative
of the overall set of objects.
• [Destination Folder Name]: This is the name of the current Capture Folder at the time of capture. This is most
useful if the operator creates one or more sub-folders for captures and can be used to designate an intermediate
hierarchy level.
• [X Digit Counter]: is a continuous counter with X digits forced. In the [...] menu of Next Capture Naming it can be
set to increment by any positive or negative value, and can be set to any value. Note that keyboard shortcuts can
be set for decrementing the counter, useful when deleting the most recent capture.
• [Current Date]: Date tokens such as YYMMDD can help collate projects spread across multiple stations and days.
In contrast, Current Time is probably not desirable since, unlike [X Digit Counter], it will not properly handle the order
of inserted or retaken images.
• Fixed Text: Fixed text is also allowed anywhere in the naming. For example, the Library of Congress could add a
“LOC-” name prefix to all files. This is also how the user designates what character or characters to use to separate
tokens (e.g. dashes, underscores, spaces etc).

52
Example Advanced Workflow: Box, Folder, Item naming
Some institutions follow a “Box ID - Folder ID - Item ID” naming scheme. In this scheme the filename “01458 - 05 -
08” would indicate the 8th item in the 5th folder of box #01458. Capture One can accommodate this naming with the
following workflow:
1. Set Next Capture Naming to [Document Name] - [Destination Folder Name] - [X Digit Counter]
2. Create a New Session for each box, and give the session the name of the ID for that box.
3. Each time a new physical folder is reached, create a new Capture Folder named for the ID of that folder, and reset the
counter.

Alternatively the box ID and folder IDs can be accommodated by naming each new Capture Folder with both the Box ID
and Folder ID. This would be preferred when the number of items and folder per box is relatively small.

Note that AppleScripting can be used to automate the creation of new sessions and capture folders, either with manually
names, or by drawing against a reference file or database.

8.6 AppleScript
Effective AppleScripting can make Capture One CH even more powerful, automating repetitive tasks, and removing
drudgery. This Mac-only programming language integrates deeply with Capture One CH allowing to natively call many of
its features and functions. To help our clients get started with AppleScripting we’ve developed a toolkit called DT Building
Blocks: AppleScripts.

DT Building Blocks: AppleScripts is a powerful teaching aid designed to use as a reference when building their own
scripts for Capture One. DT Building Blocks provide inspiration and reference code to improve the workflows of cultural
heritage digitization. The kit includes 30 scripts. Some of the scripts especially useful for cultural heritage digitization
include:

• Create Capture Folders from Clipboard

• Make Nested Capture Folders
• Toggle Capture Ready Alert
• DT CropControl CH
• Select every other variant
• Select every tenth variant
• Select next green tagged
• Select variants with blank copyright notice
• Select variants with vertical crops
• Selection List to Text Document
• Email Processed File
• Process Droplet – saved as application without start-up screen
• Process On Capture
• Pull Metadata from Numbers Spreadsheet

More info about the DT Building Blocks: Applescript can be found at https://digitaltransitions.com/building-blocks/.

53
 9
Advanced Techniques

54
In the interest of keeping this guide as approachable as possible, chapters 5-7 omitted potentially useful advanced
techniques, holding them back for discussion in this chapter. The fact that these techniques are in their own separate
chapters should not be construed as an indication that they are unusual or esoteric. Indeed many of our clients use these
techniques on a frequent basis.

9.1 Stitching
In some cases the amount of resolution available in a single capture is not enough for a desired purpose. For example,
to scan an 8x10 piece of film at 3000ppi requires ~720 megapixels, which is far more than any camera can produce
in a single capture. In such cases, stitching multiple captures together can be an effective means of increasing total
resolution.

Future versions of this document may include a full step by step procedure for stitching. In this version we simply point
out some of the factors that anyone undertaking film stitching should control as tightly as possible.

Factors to Control
The software used to stitch captures can adapt for a variety of unintended capture variation. However, the less such
variation the better the stitch will be successful.
• Illumination uniformity: If the lens and camera move, but the light source does not, then each capture requires its
own LCC. Less variation can be achieved by moving the film, while holding the light and camera/lens static. The DT
XY Film Scanner takes this later approach.
• Vibration: Stitching algorithms work off of matching up detail. If vibration is degrading the sharpness of the
captures then there will be less detail for the algorithm to match, and also the detail may be rendered differently in
adjacent captured depending on the nature of the vibration at the time each is captured. The DT XY Film Scanner
uses a heavy-duty column, heavy duty bench, carefully designed detents, and a wall-anchor to reduce the chance
of vibration.
• Non-Intended movement: Stitching is achieved by X or XY movement. The system used to make this movement
should reduce unintended movements such as rotation, displacement on the Y axis (toward/away from the
camera) or roll/pitch (swinging, even slightly, off of parallel). The DT XY uses linear slides and rigid high-quality
components to reduce non-intended movement to a minimum.
• Light source variation: The light used should be highly repeatable. Even high-end strobe systems vary in output
from one frame to another, and repeated strobe pops can create ergonomic wear on the operator if high volumes
of material are stitched. Therefore we suggest a continuous light source. The DT XY uses a DT Photon continuous
LED light which produces an incredibly stable output.
• Aperture variation: Mechanical apertures vary from one frame to the next. We recommend using a camera that
will stop the lens down once, and remain stopped down throughout the stitch, rather than returning the aperture to
the maximum opening size each capture. The DT XY can use either the DT RCam or Phase One iXG, both of which
keep the lens stopped down unless the user explicitly commands them to open.
• Mechanical shutter variation: Mechanical shutters will vary in their exact length of exposure. Use of an sensor-
based electronic shutter can ameliorate shutter speed variation, but can introduce internal geometry distortion.
• Internal geometry distortion: The use of a sweeping sensor-based electronic shutter, or a focal plane shutter, can
cause the image to slightly compress in parts of the image compared to the rest of the image. This occurs when
vibration in the system moves the subject in a different direction during part of the exposure compared to the rest.
While the impact on single-frame captures is minimal, this can cause micro detail to misalign between frames in
a stitch. The DT XY allows the use of a sweeping sensor-based electronic shutter and a heavy-duty column, heavy
duty bench, carefully designed detents, and a wall-anchor to reduce the chance of vibration. In the case ambient
vibration makes the sensor shutter inoperable the system also provides an electronically-controlled shutter.
• Lens distortion: The lens should have as little distortion as possible, as this will cause image content not to line
up between captures, unless the stitching software specifically addresses lens distortion. Even in the case that
the stitching software addresses lens distortion the correction of distortion will decrease image detail in the areas
affected. The DT XY uses three lenses, all of which are natively extremely low in distortion.

Tools
There are several tools that stitch tiled captures into a composite image. During the development of the DT XY we
evaluated AutoPano Giga, PTGui, and a few others. Our testing showed that AutoPano Giga offered the best workflow
but, regardless of settings, often produced slight “echo” artifacts along the stitching seams. In contrast to this, Adobe
Photoshop’s PhotoMerge tool produced the best quality but had no provisions for batch workflows. For this reason we
developed DT BatchStitch, an open source script that uses Photoshop’s PhotoMerge math with a custom interface that
supports stitching in large batches. More information at dtdch.com/dt-batchscript-bulk-stitching-photoshop/

55
 9.2 Reflective Content
Some transmissive collections contain both reflective and transmissive content. For example, mounted slides contain
transmissive image content in the film area and also contain reflective content such as a stamped date, slide number, or
handwritten annotation. Both reflective and transmissive content can be captured. Below we outline the suggested steps
for doing so with a DT Atom, but the process is similar with our other benches.
1. Leave one DT Photon in the standard location for transillumination of the transmissive content.
2. Mount and turn on a second DT Photon on one of the arms of the DT Atom
3. Follow the steps outlined in Chapter 5 to set up for a Object Reproduction
4. Depending on the reflective content it may make sense to add a sheet of ND Filter to the DT Photon to reduce the
ratio of the reflective light to the transmissive light. To do so, remove the four thumbscrews that hold the frame of the
DT Photon, remove the frame, cut the sheet to fit the inside dimension of the frame, and place the ND filter between
the clear Plexiglas and lensed Plexiglas, and replace the frame and frame thumbscrews.

The DT Photon on the arm of the DT Atom provides a low-angle of illumination across the film. That decreases the
chance of glare in the transmissive content area. However, the introduction of any light above the plane of the film can
cause a loss of contrast, and loss of shadow detail. Check the image quality with and without the reflective illumination
and if there is a meaningful difference you may need to capture two versions: reflective+transmissive and transmissive-
only.

56
9.3 Capture Multiple Variants
Capture One CH has the ability to create multiple variants immediately after creation. This has a variety of applications.

General workflow
1. Make a capture
2. Make one or more additional variants of that capture, and adjust them as desired
3. Select Copy Variants from Last
4. Capture a new image. The new image will come in with the same variants as the last image.

Positive/Negative Use-Case
When digitizing negative materials it can be useful to see both the Object Reproduction and Content Reproduction
versions at the time of capture. To do so, simply use the workflow above, making sure these two variants are created in
step 2.

Multiple Crop Regions Use-Case

DT systems feature up to 150mp resolution. This provides three main advantages: capturing a given material size at
higher PPI, capturing larger subjects at a given PPI, and the ability to capture (with good resolution) more than one object
at a time. For example, six hundred envelopes can be captured in fifty batches of 12 slides arranged on the DT ANR Glass
Carrier in a grid of 4 columns of 3 rows. In bulk, every slide can be saved as a separate, perfectly cropped and deskewed
file, using Copy Variants from Last and AutoCrop in the following workflow:
1. Setup Physical Grid: Create a grid on your working surface to help place each item within a consistent area of
the frame. We will call these regions of interest or ROIs. For example, you could place a mat board over the entire
working surface and use a ruler and conservation-friendly drawing implement to draw lines. Each ROI should be
large enough that an item can be quickly placed entirely inside the lines with padding.
2. Setup Digital Grid: Capture an image of the table-with-physical-grid. Crop it to the first ROI. Use F3 to Clone this
variant and in the clone, shift the crop to the second ROI. Repeat this for each ROI. Set Next Capture Adjustments to
Copy Variants from Last.
3. Capture All Material: Replace the material with the next set, and capture. The new capture will come in with as many
variants as there are ROIs. Ensure that all material fits entirely within its respective ROI. Repeat until all material has
been captured.
4. Run AutoCrop: Select all variants of all captures, and run AutoCrop. Since AutoCrop always begins with the current
crop, each variant will start looking for the subject in the ROI it represents. Accordingly each item digitized will now
have its own properly cropped and deskewed variant.
5. Handling Naming: Capture One does not allow you to name a variant separately from the name of the raw file it is a
variant of. However, you can use a Counter or Variant Position token in Output Naming to produce cohesive, ordered
naming upon output.

57
 APPENDIX A
What PPI is Best?

58
A.1 Methods to Pick a PPI
In a camera-based digitization systems the camera-to-subject distance and lens determine the size of the scan area
(aka“field size”) and the scan resolution (ppi). There are two approaches to making this selection:
• Specific PPI: Such a PPI can be selected based on a guideline like FADGI, or based on an established institutional
practice a specific PPI is selected based on the material type. Regardless of how the specific number is selected,
this method means the PPI is set without regard to how much or how little of the object fills the frame.
• Fill the Frame: Alternatively, the operator can fill the frame (with a small margin/padding outside the edges) with a
given material type ensures the highest PPI the system is capable of for that film in a single capture.

Intuitively it might make sense that filling the frame is always preferable, as it maximizes resolution for the camera
system being used. However, this path is equates the question of “What PPI is Best?” with “what can we do?” rather than
“what should we do?” which is not well-aligned with the goal of preservation-grade digitization. If the goal is the creation
of a PDO that can serve as a surrogate to the original object, replacing most needs for physical access to that original
object then the selection of PPI must be based on the content of the original transmissive material.

A.2 What do the Standards Say?

The FADGI, Metamorfoze and ISO imaging guidelines outline those elements of image quality that are required to
achieve such surrogacy. Currently FADGI is the only one of those three that provides detailed resolution requirements
for digitization of transmissive materials. FADGI first formally addressed transmissive digitization in the 2010 version of
their document Technical Guidelines for Digitizing Cultural Heritage Materials.

FADGI 2016 vs FADGI 2010

The more recent 2016 revision of this document significantly increased the suggested resolution at which to digitize
transmissive materials. Some institutions, especially lower-end service bureaus still refer to the legacy 2010 version
because its lower resolution requirements make it easier to claim FADGI conformity. We strongly advice institutions
to specifically use “FADGI-2016” rather than simply “FADGI” when specifying a number of stars for any in-house, in-
sourced, or outsourced digitization.

Summary of Standards
Below is a quick-reference chart not meant to substitute for a thorough reading of the Technical Guidelines for Digitizing
Cultural Heritage Materials document. Also note that resolution and sampling efficiency are not the only metrics by
which FADGI measure image quality.

FADGI 2016 4-Star FADGI 2010 4-Star

35mm Film 4,000 ppi at 90% Sampling Efficiency 2,800 ppi at 95% Sampling Efficiency

645 Film 1,800 ppi at 95% Sampling Efficiency

4x5 Film 1,200 ppi at 95% Sampling Efficiency

8x10 Film 2,000 ppi at 90% Sampling Efficiency 600 ppi at 95% Sampling Efficiency

Variation in Materials
The purpose of FADGI is to provide guidelines, not hard-and-fast rules. The resolution cited by FADGI is based on a
survey of common transmissive materials with the goal of picking a resolution that was sufficient for nearly all. However,
some materials will contain denser content than would be covered by the FADGI guidelines. For example, the super-fine-
grain film Kodak TechPan may benefit from higher PPI. As a more extreme example the world of transmissive materials
includes physical specimen slides that may benefit from digitization at 10,000 ppi or higher.

When possible, we suggest surveying your collection for representative materials and digitizing them at FADGI-2016
recommended resolution as well as higher resolutions, to see if your specific materials benefit from higher resolution.

59
 A.3 Practical Considerations
File Size
File storage strategy is out of scope for this document. But for completeness we mention the obvious trade-off of
scanning at a resolution higher than is needed for preservation goals: file size. Below are

PPI for FADGI-2016 4-Star Pixel Dimensions 16 Bit TIFF File Size Space required for 100,000 scans

35mm Film 4000 5,700 x 3,800 123 MB ~12 TB

645 Film 4000 9,400 x 7,100 380 MB ~37 TB

4x5 Film 4000 20,000 x 16,000 1,800 MB ~174 TB

8x10 Film 2000 20,000 x 16,000 1,800 MB ~174 TB

11x14 Film 2000 28,000 x 22,000 3,500 MB ~336 TB

Numbers rounded to the nearest 100 pixels, 100 MB, and 1 TB. They do not include any margin/padding, and do not account for use of optional compression.

File Format Size/Pixel Limitations

Note that the TIFF specification is limited to 4 GB which becomes problematic

The BigTIFF format and PSB (Photoshop Large Document Format) support file sizes and pixel dimensions far larger than
any foreseeable use in Cultural Heritage imaging would require. The Library of Congress lists BigTIFF as having favorable
sustainability factors, but BigTIFF is not broadly supported by commercial software (e.g. neither Adobe Photoshop nor
Capture One CH support this format). PSB is supported by Adobe Photoshop but is not listed by the Library of Congress
as having favorable sustainability factors.

If a cultural heritage digitization project runs up against the limitations of TIFF, it is likely to do so in the context of
stitching. In such cases we’d recommend saving the individual tiles alongside the composite format chosen.

Collation: PPI Changes vs Additional Captures At Current PPI

Earlier in this chapter we discussed using a Specific PPI rather than Fill the Frame method of selecting a capture PPI; one
benefit of this is the reduced number of times the PPI will need to be changed. With any camera-based capture system it
is faster to capture the next item at the same PPI than to change PPI. Therefore, selecting a fixed PPI is a faster workflow
than changing the PPI for each new object. For example with an IQ 150mp at 4000ppi the capture window is 3.5” by 2.6”
which means 35mm, 645, 6x6, 6x7, and 6x8 film can all be captured without changing PPI or stitching.

Where possible, materials should be collated prior to arriving at the digitization stage, so that like-sized materials can be
digitized in batches. More information on a collated workflow can be found in our Digitization Program Planning Guide.

Not Enough Resolution?

When a required PPI cannot be captured with a given capture system there are four options.
• Compromise: Use the highest single-shot PPI possible
• Stitch: Capture the object in multiple tiles, and stitch these tiles to a composite output. For more information see
chapter 9.1.
• Upgrade Capture Systems: Invest in a capture system with a higher resolution. For example, if your capture
system is 50mp, upgrade to 100mp or 150mp.
• Wait: Delay digitization of the material until capture systems with higher resolution are available.

The Wait option is typically a matter of intra-project prioritization rather than absolute delay. That is, digitization of a
large collection is often a multi-year project, and scheduling the largest material for the end of the project may pay
dividends in efficiency as higher-resolution capture systems become available to the project. Of course, this is just
one of many considerations when determining the priority and scheduling of digitization. For more information about
Prioritization of materials see our Digitization Program Planning Guide.

60
61
 APPENDIX B
Additional Resources

62
B.1 Written Resources
FilmCare.Org, Image Permanence Institute
FilmCare.org is a web-based film resource created by the Image Permanence Institute, a non-profit preservation
research lab devoted to the development and deployment of sustainable practices for the preservation of images and
cultural property since 1985. The Image Permanence Institute is a department of the College of Imaging Arts and
Sciences at Rochester Institute of Technology (RIT) in Rochester, NY, USA.

FADGI Guidelines
Federal Agencies Digitization Guidelines Initiative (FADGI), Technical Guidelines for the Still Image Digitization of Cultural
Heritage Materials
www.digitizationguidelines.gov

International Standards Organization

ISO 19262: Photography - Archiving Systems - Vocabulary
ISO 19263: Photography - Archiving Systems - Best Practices for Digital Image Capture of (reflective) CH Materials
ISO 19264: Photography - Archiving Systems - Image Quality Analysis

Other Documents by Digital Transitions

Digitization Program Planning, a Comprehensive & Practical Overview of Cultural Heritage Digitization
www.dtdch.com/digitization-program-planning/

Color Guide for Reflective Digitization by Digital Transitions

www.dtdch.com

B.2 Expert Individuals

Tom Rieger Eric Shows
Library of Congress New York Public Library
Digital Collections Management and Services Division ericshows@nypl.org
trie@loc.gov Digitization Services Manager

Don Williams Peter Riesett

Owner and Lead Image Scientist New York Public Library
Image Science Associates Head Photographer, Digital Imaging Unit
don-williams@rochester.rr.com peteriesett@nypl.org

Joseph Rheaume Doug Peterson

Program Coordinator, Digital Projects Digital Transitions
Center for Creative Photography Head of R+D and Products
RheaumeJ@ccp.arizona.edu dep@digitaltransitions.com

Eric Philcox Spencer Zidarich

Owner Digital Transitions
Pixel Acuity Technical Specialist
eric@pixelacuity.com srz@digitaltransitions.com

Michael Smith
Manager, Imaging Services
The Getty Trust
MSmith@getty.edu

63
 APPENDIX C
Workflow Summaries

64
C.1 iXG, AutoColumn, & DT Photon
Object Reproduction
Section Page Step Summary

5.1 29 Select Lens & Extension Consult table for lens and extension tube options.

5.2 30 Set Camera Position & Focus Enter desired Frame Size in Camera Focus and push “Start”
5.3 32 Set Mode [Base Characteristics > Mode > Film Positive]

5.3 32 Set Profile [Base Characteristics > ICC Profiles > Cultural Heritage > DT Photon]

5.3 32 Set Curve [Base Characteristics > Curve > Linear Scientific]

5.4 32 Set Sharpening Sharpening: Amount=100, Radius=1, Threshold=0, Halo=0

& Noise Reduction Noise Reduction : Luminance = 0, Color = 40, Single Pixel =0

5.5 32 Capture LCC Raw File Remove film. Insert Plexiglas. Expose for mid-gray.

5.6 34 Analyze LCC Right Click thumbnail and select Create LCC.

5.7 34 Apply LCC [Next Capture Adjustments > All Other > Copy From Last]

5.8 35 Set White Balance With film removed, white balance to the light source

5.9 35 Set Exposure With film removed, change shutter speed until exposure warning of 250 barely shows

5.10 35 About the Crop Step 5.2 already set the crop. No need for further crop.

5.11 35 Ready to Capture Capture all material of this type

Content Reproduction
Section Page Step Summary

6.1 37 Clone Variant Select Object Reproduction and use keyboard shortcut F3

6.2 37 Set Mode [Base Characteristics > Mode > Film Positive]

6.3 38 Set Curve [Base Characteristics > Curve > Linear Scientific]

6.4 38 Set White Balance With film removed, white balance to the light source

6.5 39 Set Exposure With film removed, change shutter speed until exposure warning of 250 barely shows

6.6 39 Set Crop

Speculative Artist’s Intended Rendering

Section Page Step Summary

7.1 41 Clone Variant Select Content Reproduction and use keyboard shortcut F3

7.2 41 Research Artist’s Intentions Consult primary sources (artist, print, writing), secondary sources (curators, researchers),
or similar artists to the extent possible and practical
7.3 42 Adjust Color and Tone Physical Deterioration: Auto R,G,B Levels
Non-Neutral Intention: Color Balance
Film Ecosystem: R,G,B Levels, Curves and Color Editor

65
C.2 DT RCam, CMOS, DT AutoColumn,
& DT Photons
Object Reproduction
Section Page Step Summary

5.1 29 Select Lens & Extension Use 120mm lens (if available) with no extension tubes for 8x10, 5x7, or 4x5.
For smaller formats consult table for number of extension tubes.

5.2 30 Set Camera Position & Focus Move camera up/down using Copy Stand tool, roughly refocusing each time, until framing
is close. Then use Copy Stand tool with Focus Meter to fine tune focus.

5.3 32 Set Mode [Base Characteristics > Mode > Film Positive]

5.3 32 Set Profile [Base Characteristics > ICC Profiles > Cultural Heritage > DT Photon]

5.3 32 Set Curve [Base Characteristics > Curve > Linear Scientific]

5.4 32 Set Sharpening Sharpening: Amount=100, Radius=1, Threshold=0, Halo=0

& Noise Reduction Noise Reduction : Luminance = 0, Color = 40, Single Pixel =0

5.5 32 Capture LCC Raw File Remove film. Insert Plexiglas. Expose for mid-gray.

5.6 34 Analyze LCC Right Click thumbnail and select Create LCC.

5.7 34 Apply LCC [Next Capture Adjustments > All Other > Copy From Last]

5.8 35 Set White Balance With film removed, white balance to the light source

5.9 35 Set Exposure With film removed, change shutter speed until exposure warning of 250 barely shows

5.10 35 About the Crop Step 5.2 already set the crop. No need for further crop.

5.11 35 Ready to Capture Capture all material of this type

Content Reproduction
Section Page Step Summary

6.1 37 Clone Variant Select Object Reproduction and use keyboard shortcut F3

6.2 37 Set Mode [Base Characteristics > Mode > Film Positive]

6.3 38 Set Curve [Base Characteristics > Curve > Linear Scientific]

6.4 38 Set White Balance With film removed, white balance to the light source

6.5 39 Set Exposure With film removed, change shutter speed until exposure warning of 250 barely shows

6.6 39 Set Crop

Speculative Artist’s Intended Rendering

Section Page Step Summary

7.1 41 Clone Variant Select Content Reproduction and use keyboard shortcut F3

7.2 41 Research Artist’s Intentions Consult primary sources (artist, print, writing), secondary sources (curators, researchers),
or similar artists to the extent possible and practical

7.3 42 Adjust Color and Tone Physical Deterioration: Auto R,G,B Levels
Non-Neutral Intention: Color Balance
Film Ecosystem: R,G,B Levels, Curves and Color Editor

66