Personal illustration practice often produces dozens of sketches, studies and half-finished ideas that become hard to locate or compare once they accumulate. An artwork index paired with brief scene notes gives each piece a stable reference point and preserves the reasoning behind compositional choices while they are still fresh.

A Hypothetical Cloud-and-Rooftop Sketch as Starting Point

This example begins with a single imagined project: a small series of cloud-and-rooftop sketches drawn over several weekends purely for compositional practice. The artist in this scenario is not depicting any real building, city or weather event. The subject is chosen because it offers two distinct visual layers, sky and structure, that invite repeated variation without requiring reference to a specific place. Each sketch explores a different balance between cloud mass and rooftop geometry, and the goal of the index is to make those variations retrievable and comparable later.

In this example the first sketch places a dense cluster of cumulus forms across the upper two-thirds of the frame, leaving only a narrow band of angular rooftops below. The second sketch inverts the ratio, giving rooftops more visual weight and pushing clouds into a thin strip near the top edge. The third introduces a diagonal division where cloud and roof shapes interlock along a slanting line. These three pieces already demonstrate why an index helps: without written notes, the artist would remember only that they drew clouds and roofs several times, not which arrangement solved which compositional question.

The working assumption here is that the index lives in a plain text file or a simple table, not in specialized portfolio software. This keeps the method accessible to anyone with a text editor and avoids depending on tools whose features may change. The fields described below are suggestions shaped by this particular scenario; a different practice might need more or fewer of them.

Choosing Index Fields and Entry Structure

Record a short set of consistent fields for every new entry before drawing begins or immediately after finishing a sketch. In this example the fields are: a sequential identifier, a working title, the date of the session, a one-line compositional note and a status label. The identifier uses a project prefix followed by a three-digit number, such as CRT-001 for the first cloud-rooftop piece. The working title is deliberately brief and descriptive rather than poetic, for instance Cumulus Over Narrow Rooftops, because its purpose is retrieval, not exhibition. The status label distinguishes between sketch, study and resolved, acknowledging that these categories are the artist's own working definitions rather than any external standard.

The compositional note is the field that most directly supports future decision-making. Here I suggest keeping it to a single sentence that names the primary spatial relationship being tested. For the first sketch the note reads: dense cloud mass occupies upper two-thirds, rooftops compressed into lower band. For the second: rooftop geometry dominates, clouds reduced to thin upper strip. This constraint forces the artist to identify the single most important variable in each piece rather than writing a general impression. When the index grows beyond twenty or thirty entries, these one-line notes become the fastest way to scan for a specific arrangement without reopening image files.

Use a consistent file-naming convention that mirrors the index identifier so that the text record and the image file can be matched without opening either one. In this example the image files follow the pattern CRT-001_v1.png, where the version suffix is reserved for cases where the same composition is redrawn with deliberate changes. This pairing between index entry and file name is a design choice specific to this scenario; other practices might prefer date-based names or folder hierarchies. The key principle is that the index and the file system should agree on at least one shared token, so that neither depends entirely on memory.

Sharing Your Index and Preparing a Public Address

Once your cloud-and-rooftop index has consistent fields, clear file names and a reliable structure, you may want to share it with other illustrators or make it accessible as a reference page. If that is your intention, a custom web address is one practical preparation to consider. Registering a domain is not required to complete the indexing work described above; it simply secures a memorable name you can use if you later decide to publish your work online. The table below shows the initial lookup step at two registrars, drawn from their respective documentation.

CheckNamesLinkNamecheap
First lookupEnter your desired domain name in the search box on the homepage to see availability across suffixesEnter your desired domain name in the Domain Name Search field and click Search to view availability status

12

Writing Scene Notes That Preserve Intent

Write scene notes as a short paragraph attached to each index entry, separate from the one-line compositional note described above. The scene note captures the reasoning behind choices that are not visible in the finished sketch: why a particular cloud shape was selected, what the artist was trying to avoid, or what problem the next sketch should address. In this example the scene note for CRT-001 explains that the dense upper cloud mass was an attempt to create a sense of atmospheric pressure pressing down on the architecture below, and that the next piece should test whether inverting the ratio produces the opposite feeling of openness.

Check that each scene note answers at least one question of the form: what would I want to know about this piece six months from now? This self-directed question prevents the notes from becoming either too vague to be useful or so detailed that writing them feels like a burden. In this example the artist resists the temptation to describe color temperature or line quality in the scene note, because those qualities are visible in the image itself. The note instead records the invisible layer: the intention, the constraint being tested and the planned direction for the next iteration.

Scene notes also serve as a natural place to record what was deliberately excluded. In this example the artist notes for CRT-003 that human figures, birds and text elements were all considered and rejected because they would introduce narrative reading where the goal is purely spatial tension. Recording exclusions may seem unnecessary at the moment of drawing, but it prevents the artist from re-litigating the same decision weeks later when memory of the original reasoning has faded. This practice is a suggestion for this particular workflow; it does not claim that exclusion notes are universally necessary for all illustration practices.

Tracking Versions Without Multiplying Entries

List the conditions under which a redrawn sketch receives a new version suffix rather than a new index entry. In this example the rule is: a new version is created only when the core compositional arrangement remains the same but specific elements are adjusted for clarity, proportion or line weight. If the spatial relationship between cloud mass and rooftop geometry changes fundamentally, the result receives a new index number and its own compositional note. This distinction keeps the index from inflating with minor corrections while still preserving meaningful iterations.

For CRT-001 in this example, the artist produces a second version after noticing that the original rooftop band is so compressed it reads as a solid dark rectangle rather than as distinct architectural forms. The redrawn version keeps the same two-thirds cloud ratio but introduces visible gaps between roof shapes so that individual structures can be parsed. This becomes CRT-001_v2. The scene note is updated with a single appended sentence explaining the change, not by overwriting the original note but by adding a revision line below it. This preserves the reasoning trail and makes it possible to see how understanding of the composition evolved between versions.

When a version reaches a point where the artist considers it resolved enough to stop iterating, the status label changes from sketch to study or resolved. In this example the threshold is intentionally subjective: the artist decides that a piece is resolved when they can look at it and identify no single change that would improve the primary spatial relationship being tested. This is not a claim about objective quality; it is a personal stopping rule that prevents endless revision. The index records the final status so that future browsing can filter by completion level without re-examining every image.

Maintaining the Index as the Practice Grows

Review the index at natural pause points rather than on a fixed calendar schedule. In this example the artist reviews entries after completing a batch of five to eight sketches, which corresponds roughly to two or three drawing sessions. The review consists of reading the compositional notes in sequence and checking whether the series is still exploring distinct arrangements or has begun repeating itself without new information. If repetition is detected, the note for the most recent piece becomes the place to record that observation and to propose a different variable for the next batch.

As the index grows beyond thirty or forty entries, consider adding a secondary grouping field. In this example the artist introduces a short tag such as inverted-ratio, diagonal-split or layered-depth to cluster pieces that test the same type of spatial relationship. These tags are not a formal taxonomy; they emerge from the notes already written and simply make scanning faster. The tags are added during the batch review, not during the initial entry, because the artist cannot always predict which category a sketch belongs to until several related pieces exist for comparison.

If the practice eventually shifts to a different subject, start a new index file with a new project prefix rather than appending to the existing one. In this example, if the artist moves from cloud-and-rooftop scenes to a series exploring interior light through windows, the new project receives its own prefix, such as ILW-001. This separation keeps each index focused and prevents the need to scroll through unrelated entries. The scene notes and version rules carry over as method, but the content and identifiers remain project-specific. This is a structural suggestion for this workflow; other artists may prefer a single continuous index with subject tags instead of separate files, and both approaches are valid depending on volume and preference.

Frequently asked questions

Do I need specialized software to maintain an artwork index?

No. This article's example uses a plain text file because it is universally accessible and easy to search. A spreadsheet, note-taking app or even a physical notebook can serve the same purpose as long as the fields remain consistent and the file or page can be located quickly.

How detailed should scene notes be if I am short on time?

A single sentence naming the primary spatial decision and one sentence stating what the next piece should test are usually sufficient. The goal is to preserve reasoning that is invisible in the image itself, not to write a full description of what is already visible.

When should I create a new index entry versus a new version of an existing one?

In this article's example, a new version is used when the core compositional arrangement stays the same but specific elements are refined. A new entry is created when the fundamental spatial relationship changes. The exact threshold is a personal judgment call, but having a stated rule prevents the index from filling with near-duplicates.