X.3.1: Tuberculosis Imaginaries Research Protocol

This was the research protocol used for the Tuberculous Imaginaries installation (3.2.4; 3.2.5; 3.2.6):

1. All resources will be downloaded in full from HathiTrust. (The full corpus can be found here). Both .pdf files and .txt files will be acquired. Upon downloading, the files will be organized in the Airtable document according to their type, with identifying information (author, publisher, year, publication city). Collections of journal articles will have each article identified separately.
   • Articles and books will be flagged as to the implicit audience—scientific, lay people—or their use—advertisement, research, or public knowledge.
2. As these files are organized, all images will be made into separate .pdf files. These images include (but are not limited to) clinical photographs, photographs of sanatoria, clinical and anatomical illustrations, and diagrams for research and teaching. Graphs and data organized as charts will not be included. Further, any image that shows a child (under the age of 18) will be noted in the accompanying .txt file (FILENAME_IMGNotes.txt) and not separated as its own image.
3. These image files will be used for the upcoming installation. They will be examined and selects will be chosen based on their merit to describe clinical visual culture.
   • A secondary concern will be to choose images that do not exploit the subject—choosing framing decisions that do less to humiliate or degrade the sitter.
   • While the patient is always already exploited in these images, the intent of the installation is NOT to maintain an overt exploitation in the shape of the spectacularized medical body.
   • Further, this study will be omitting non-human animal testing, so as to reduce the potential scope.
4. Selected images will be made into video for projectors A & B of the installation (see Installation Protocol below).
5. Once selects are flagged, excerpts from the books (provided through the .txt files) will be made to accompany these images through the use of projector C (again, see Installation Protocol below).

Installation Protocol

Projector A — Clinical Photographs & Illustrations

1. Images will be chosen for this projector based on their use for clinical research.
   • A broad collection of images will be made, to help describe the various kinds of clinical images may be levied by medical researchers. This includes anatomical illustrations, pathological illustrations, photographs of pathological samples, illustrations of microscopic organisms, and the photography and illustrations of symptoms experienced by an ailing, human sitter.
2. The images will be cropped into a 1920x1080p image. While the aspect ratio may not be amenable to this, excess material will be deleted (made transparent) while the final image will remain 1080p. Once framed, cropped, and edited, the final image will be output as a .png file into the Projector A folder. (SOURCE_PAGECITATION_Draft1.png)
   • The intent here is to create a generalizable image that will be more easily managed in the creation of the final video.
3. These images will be imported into a video editing platform. Once uploaded, each image will be used in a slow-moving zoom. The image will be made to fit an 800x600p resolution (the projector’s native resolution) and then made to do a slow zoom and fade. (Parameters: Zoom rate 100% closer over 90 seconds, fade in at the first 20 seconds from 0 opacity to 100 opacity, fade out in the last 30 seconds from 100 opacity to 0 opacity. The next image begins at the moment the previous image begins to fade.) Images will be on screen for 90 seconds each.
4. After the media is made, an export will occur and the exported file will be reimported for color correction (some minor brightness and contrast adjustments to make things clearer), before being exported a second, final time.

Projector B — Imagining Medical Work

1. Images will be chosen for this projector based on their connection to tuberculosis’ medical culture. They will show the relationship between the medical class (doctors and nurses) and the patient. These images will be copied from the source into the project folder (_ClinicalImages).
   • My interest in this feed is to show the various ways doctors and nurses are imaged intentionally (as with posed or candid shots in laboratories) and unintentionally (as they hold patients in place for exposure).
2. This projector will be placed at a 90 degree angle and have a black space in the middle of the frame, creating space for projector C’s video feed. Images for this feed will be developed in photoshop, using a template to standardize this use of space in the image.
3. Images will be selected on an ad hoc basis from the corpus, with care to show the medical doctors and nurses more so than the patients. When an image is selected, it is flagged in the folder (for ease of citation later) and fitted as best into the projector template. (This is done by a case by basis). The image, once framed in this way, is exported into a .png file labeled (SOURCE_PAGECITATION_Draft1.png).
4. Once the first image is exported, a second image using the same material is made. Using the polygonal lasso tool in photoshop any material not related to the embodied presence of the doctors is cut away and deleted, leaving a transparent version of the first image, save the embodied presence of doctors. The image is exported into a .png file labeled (SOURCE_PAGECITATION_Draft2.png). For the video feed, video material is layered as such (from the top down)
   • Video Layer 8: A template for the black box that projector c will project through.
   • Video Layer 7: The edited (draft2) version of the photographs.
   • Video Layer 6: A layer for the animated block (a white rectangle measuring 800px by 20 px)
   • Video Layer 5: A layer for the animated block (a white rectangle measuring 800px by 20 px)
   • Video Layer 4: A layer for the animated block (a white rectangle measuring 800px by 20 px)
   • Video Layer 3: A layer for the animated block (a white rectangle measuring 800px by 20 px)
   • Video Layer 2: The original (draft1) framed version of the photograph.
   • Video Layer 1: A white matte layer.
5. Each image used will be edited using the same parameters (with some variation based on the images sourced and aesthetic choices). The unedited (draft1) and edited (draft2) will run for up to 80 seconds, with other effects applied to the images during this time. After 10 seconds the first of up to 4 animated blocks will move from the left of the final frame until it blocks out a part of the image (the effect will make the edited draft in Video Layer 7 pop out, framing elements of the image). Each block will be placed based on aesthetic consideration, and in the case that only 2 or 3 are warranted, the image duration will be reduced by 10 seconds. After all of the blocks are animated, a final crop effect is applied to the original (draft1) image which will crop out the images over 10 seconds (leaving the white matte and the top edited (draft2) visible). After this is cropped out, the image stays static for 20 seconds before fading out. The white matte is left visible for 5 seconds before the next image is cut into the feed.

Projector C — Rhetoric and Response

1. Using images flagged and selected for preliminary use in both the projector A and B video feeds, a collection of .txt file excerpts will be made from the optical character recognition (OCR; provided by HathiTrust) processed versions of the sources.
2. From here, the text will be cleaned up (removing typos, documentation that shows pagination and so on).
3. Once cleaned, the excerpted texts will be read and commented upon.
   • The original text will remain the same, but occasional breaks in the spacing using the brackets of “// [comment] //” will delineate comments and interventions in the text.
   • Comments are broad, but should be quippy, commenting on granular aspects of the text, assumptions, common metaphors, and noting recurring themes. Comments will also be all in lower caps (for purely aesthetic differentiation).
   • Once an excerpt is commented, the text is copied into a preformatted photoshop file—(800px x 20000px ). The text is formatted as follows:
     • Original Primary Text: Baskerville 16 font, left justified
     • Comments: PT Sans Narrow 16 font, right align
     • In addition to this formatting, the commented lines will be matched with a rectangular block, to signify disruption in the original source.
   • Once finished, the file will be exported as a .png file (SOURCETITLE_ExerptNOTEONCONTENT.png)
4. The final video will use the photoshopped image by placing it under a preformatted template (to fit within the requirements of projector B), this layer will then be placed over an image used to cite the text source (the cover page, the paper title, or other identifying material in the original .pdf file).
5. The photoshop file will be aligned with the left part of the frame and adjusted to better frame the text (a .7% crop on the top part of the frame). The image will move 18-20 frames up every 4 seconds (depending on variation in the source), imitating a crawl.
6. Once a text file is finished, the next file will be layered over the last to make an illusion of infinite crawl.