Post-production, provenance and archiving

Post-production turns selected raw files into finished masters ready for print, book and screen. In a capstone it also has to do three less visible jobs: keep the series consistent, keep an honest record of how each image was made, and keep every file safe and findable long after the term ends. The editing and retouching tools were taught in PHOT-202, colour management and fine printing in PHOT-301, and content credentials, detection and AI disclosure in PHOT-420 lessons 10 and 11. This lesson is about the workflow that holds them together across a whole body of work.

A post-production workflow for a body of work

Work in a fixed order so that changes upstream do not undo work downstream. A reliable order is: ingest and back up, cull and rate (lesson 6), apply a base adjustment to the whole set, make global corrections per image, make local corrections and retouching, check consistency across the set, proof for each output, export masters, and then make output files from the masters. Keep raw files untouched and work non-destructively in a cataloguing application or with layered files, so that any decision can be revisited.

Consistency across the series

A series looks unfinished when its images disagree in colour, contrast or density for no reason. Choose a reference image for each image type, finish it carefully, and match the others to it. Cataloguing software lets you copy settings across images as a starting point, but images made in different light need individual adjustment to look the same; identical settings rarely produce identical results. Review the whole set at small size side by side, then at print size, and look for outliers in white balance, skin tones, shadow density and highlight brightness.

Amara's portraits were shot under mixed night light, and her decision in lesson 4 was to keep the ambient colour casts while making skin tones look natural. She finished one portrait from each site as a reference, built a base profile and a skin tone adjustment, and then matched the others. Her interiors used a different base with slightly lower contrast to keep shadow detail in long exposures. A final pass with all eighteen wall images on screen at the same size showed that the radio station images were warmer than the rest, which she corrected toward the series average without removing the station's character.

Honest post-production

What counts as acceptable post-production depends on the genre and the claims made about the images. For documentary and journalistic work, the widely accepted standard, discussed in PHOT-320 lesson 9, permits adjustments to tone, colour and contrast that correspond to what was seen, plus removal of sensor dust, but not adding, removing or moving elements in the scene. Fine art and commercial work may allow much more, provided viewers are not misled about the nature of the image. A capstone that presents documentary images must follow documentary standards, and any departure, such as a composite or a staged scene, must be disclosed in captions or the project text.

Generative tools raise the same question with higher stakes. If you use generative fill, AI expansion, AI relighting or AI upscaling on any image, record it, and follow your school's policy and the disclosure practices in PHOT-420 lesson 10. For a documentary capstone the safest practice is not to use generative editing on content at all, and to limit AI tools to tasks such as noise reduction and masking that do not change what the image shows. Write a short post-production statement for the project that says what was and was not done; it will be part of your defence.

Provenance and content credentials

Provenance is the documented history of an image: who made it, when, where, with what equipment and what was done to it afterward. In practice it lives in three places: the raw files and their capture metadata, the edit history in your catalogue, and embedded metadata or content credentials in the files you deliver. PHOT-420 lesson 11 explained how C2PA content credentials record capture and edit history in a signed manifest that later viewers can inspect. If your camera or software supports content credentials, consider enabling them for the capstone; they show viewers and juries that the work is what it claims to be, and they are increasingly expected for documentary submissions. Keep the raw files regardless; they are the strongest evidence of authenticity.

Metadata

Metadata makes files findable and carries rights and caption information with the image. The IPTC Photo Metadata Standard, supported by most cataloguing software, defines fields for the creator, copyright notice, credit line, contact details, caption, headline, keywords, location, date created and, in recent versions, a digital source type that indicates whether an image was captured by a camera, edited, composited or generated. The PLUS fields add model and property release status. Fill the creator and rights fields at ingest with a template; add captions, keywords and release status during the edit using your field notes from lesson 4.

Some subjects or sites may ask that names or exact locations not be published. Respect that in public captions and file metadata exports, while keeping the full information in your private catalogue and field notes. Strip GPS data from public web files when location could put someone at risk, as the security practices in PHOT-410 lesson 10 suggest.

File naming and versions

Consistent file names prevent confusion when hundreds of files move between applications, labs and people. A good scheme includes a date, a project code and a sequence number, and avoids spaces and special characters: for example 2026-10-08_NS_4120.CR3 for a raw file. Masters and output files add a version and purpose: NS_4120_master_v2.tif, NS_4120_print_40x50.tif, NS_4120_web_2000.jpg. Never overwrite a master with an output file, and never rely on "final", "final2" and "final-really" as version names.

Backups: the 3-2-1 rule

Losing raw files during a capstone is a disaster that is almost entirely preventable. The 3-2-1 rule says to keep at least three copies of every file, on at least two different kinds of media or devices, with at least one copy off site. For a student, that usually means the working copy on a laptop or main drive, a second copy on an external drive kept at home, and a third copy in a cloud service or on a drive kept somewhere else. Back up immediately after every shoot, before formatting cards; cards are not backups. Test restoring a few files once a month so you know the backup works.

raw_mb, frames = 45, 6400          # average raw file size and frames shot
masters, master_mb = 60, 180       # layered 16-bit masters
outputs_gb = 12                    # prints, book files, web files
catalog_gb = 8                     # catalog and previews
raw_gb = raw_mb * frames / 1000
total = raw_gb + masters * master_mb / 1000 + outputs_gb + catalog_gb
print(f'raw {raw_gb:.0f} GB, total project {total:.0f} GB')
headroom = 1.5                     # growth and versions
need = total * headroom
for drive in (1, 2, 4):
    print(f'{drive} TB drive: {"enough" if drive * 1000 >= need else "too small"} for {need:.0f} GB with headroom')
copies = 3
print(f'storage across all copies: {need * copies / 1000:.2f} TB')

raw 288 GB, total project 319 GB
1 TB drive: enough for 478 GB with headroom
2 TB drive: enough for 478 GB with headroom
4 TB drive: enough for 478 GB with headroom
storage across all copies: 1.43 TB

The project needs about 320 GB, or about 480 GB with headroom for versions and growth, so a 1 TB drive per copy is enough for this capstone alone. Students working with video, high-resolution or multi-frame techniques from PHOT-420 should rerun the numbers with their own file sizes, which can be several times larger.

Checksums and archiving

At the end of the term, make an archive of the project that will still be usable in ten years. An archive is not just a copy of your working folder. It should contain the raw files, the catalogue or the edit settings, the masters in an open, widely supported format such as TIFF, the final output files, the field notes, releases and agreements (scanned), the process log, the project texts and a short read-me file explaining the folder structure and software versions. Generate checksums, numeric fingerprints of each file, when you create the archive; comparing checksums later proves that files have not been corrupted. Many backup tools do this automatically.

Handing over files

Some capstone outputs leave your control: prints go to a lab and framer, book files to a printer, images to site partners and subjects. Send only what each party needs, in the format they need, with names and metadata in place. Agree in writing what partners may do with images you give them, following the uses in your access agreements, and keep a record of what was sent to whom and when.