A 3D model grows through iterations. Texture and sculpt reference is not one artifact—it is dozens of specific searches: "show me every surface breakdown I saved for faces," "what was that leather wrinkle detail I liked last month," "all my best metal corrosion samples rated 4 or higher."
A reference system built for 3D requires that every type of material—fabric folds, skin detail, weathered metal, wood grain—stays across projects. A search should find "all sculpt studies I've rated as teaching examples" or "surfaces tagged as production-ready from any archived job." And when a model evolves, the old reference stays in place, tagged and dated, so you can return to "what did this surface look like two weeks ago" without excavating a backup.
Coii Ref keeps all of it in your folder structure. Tags are metadata you layer on top, not a structure you have to reorganise. Material libraries grow year to year. Searches are one line and run across a decade of saved reference.
This is how a 3D workflow builds around it.
Day one: Set up the material taxonomy
The first model of the quarter is an armoured character. You need: base anatomy, muscular forms under the armour, hard surface construction, and weathering. This is your first time building the specific material setup, so the tag structure matters.
Create a root folder: /Volumes/3D-Work/Projects/armour-character-2026/. Under
it, three reference folders: Anatomy/, HardSurface/, WeatheringDetails/.
Open the workspace and create saved views for the project. These are queries you will run dozens of times:
tag:anatomy in:/Volumes/3D-Work/Projects/armour-character-2026tag:hardsurface:joints in:/Volumes/3D-Work/Projects/armour-character-2026tag:weathering rating:>=3 in:/Volumes/3D-Work/Projects/armour-character-2026Save each one. They become the queries you open every morning. The folder path is fixed; the tags change as you rate and refine.
Week one: Collect anatomy reference
Anatomy is the foundation. The character is humanoid but needs exaggerated proportion—lean, extended limbs, small head. Start with figure studies:
type:image tag:proportion,lean in:/Volumes/3D-Work/ArchiveThis pulls from a shared archive of anatomy studies rated and tagged from
previous characters. Fifty-eight images. Open them and rate down the ones that
do not fit—muscular builds instead of lean, or static poses instead of dynamic.
Keep twelve keepers rated 5 and tag them tag:for-armour-character.
Save a few to a new canvas: "Anatomy foundation." Six plates—front, back, side view, proportion study, hand detail, head-shoulder proportion. Export the JSON and commit it to your project revision history.
As sculpting begins, you find small anatomy adjustments. Download high-resolution
figure drawing studies. Name them clearly: Figure-Drawing-Lean-Front.jpg,
Muscle-Anatomy-Shoulders.jpg. Drop them into Anatomy/.
The watcher picks them up immediately. Tag them tag:anatomy,sculpt-study and
rate them. Now they are in the searchable index.
Week two: Hard surface construction
Armour is layered geometry. Joints need definition, gaps need logic, and surfaces need visible construction. Collect reference on hard surfaces:
type:image tag:hardsurface in:/Volumes/3D-Work/Archive rating:>=3Thirty-two images. Scroll and rate. Many are good reference but not specific to
this character—knight armour from historical reference is not the same as
sci-fi armour with exposed joints. Tag them differently: tag:hardsurface:historical
versus tag:hardsurface:scifi. Save the sci-fi set:
tag:hardsurface:scifi rating:>=3 in:/Volumes/3D-Work/Projects/armour-character-2026This becomes a view you check every time you model a new joint or plate. Create a second canvas: "Hard surface detail." Plates: joint articulation, plating pattern, material transitions, panel construction, weathering decay, rivets and fasteners.
Week three: Weathering and material finishing
The model is built. Now it needs character—surface corrosion, paint wear, rust bloom, fabric deterioration. This is where material reference becomes texturable.
Search your full archive for weathering reference:
type:image tag:weathering width:>2500 rating:>=3Wide images work best for texture painting—they give pixel resolution. Find seventy-one matches. Now narrow to specific material types:
type:image tag:weathering:metal rating:>=3 width:>2500 in:/Volumes/3D-Work/ArchiveTwenty-four metal weathering samples. Look for images showing rust progression, oxidation patterns, paint loss, and underlying material. Rate the best ones up to 5. Add a note on each one: "Thin rust layer over bare steel" or "Deep corrosion in crevices."
Week three and a half: Creating cross-project material searches
Before you finish the material gathering phase, think ahead. You will do three more characters this year, and each one will need weathering and hard surface reference. Instead of starting from scratch each time, build a permanent search for material types across your entire archive:
tag:weathering:metal type:image width:>2000 in:/Volumes/3D-Work/Archive rating:>=3This searches not just this project, but every past project in your archive drive. Forty-one images show up—metal weathering samples across every 3D model you have shipped in the past two years. Some are from a sci-fi robot, some from a fantasy sword, some from corroded machinery. The material principle is the same. Tag the best ones:
tag:teaching:weathering:metalThese become your reference library for texture teaching. When a junior 3D artist asks how to paint convincing rust progression, pull this search and show them the collection you have built over time.
Save the view. It becomes one of your daily reference queries:
tag:teaching:weathering type:image rating:>=4 in:/Volumes/3D-Work/ArchiveWeek four: Building the working material library
By week four you have twenty-five tagged and rated material reference images
across your project. They live in the three folders: Anatomy/, HardSurface/,
WeatheringDetails/. The app indexes all of them. Rate each image the moment you collect it—do not save rating for later.
Create a new canvas for each material type as you begin texturing:
Canvas one: "Metal weathering"—eight plates showing different corrosion stages, each at 2500+ pixels wide, each with a note describing what you are seeing in the texture.
Canvas two: "Fabric deterioration"—showing how different materials wear over time: leather cracks, cloth frays, canvas fades.
Canvas three: "Joint detail"—close-ups of how real armour articulates, plates overlap, and surfaces transition.
Export all three to your project folder. Commit them to version control. They become your texture painting reference.
Month two: Archive the project and cross-reference for the next one
The armour model ships. Move the entire project folder to /Volumes/Archive/
and tag everything in it:
tag:archive:armour-2026Now your material library grows. Search across all archived projects:
tag:weathering rating:>=4 in:/Volumes/Archive type:imageThis pulls every weathering sample rated as teaching reference from every shipped project. Fifty-one images. The best material studies from six months of work are now one query away.
Next quarter: new character, similar materials. You do not start from zero. Open a saved view from the previous project and run it again:
tag:weathering:metal in:/Volumes/3D-Work/Projects/new-character-2026Same search structure. Same rating system. Different project folder. The result pulls the old project's best metal reference without touching it—it sits in the archive, version-controlled, backed up, and instantly findable.
Building a shared material taxonomy
A challenge in team work is that different artists use different mental models. One modeller thinks of materials as "worn versus new," another thinks "by corrosion type." Both are right; they are just different slices of the library.
The solution is not one perfect taxonomy. It is multiple tags on every image.
That worn leather reference gets tagged both tag:worn and tag:corrosion:oxidative
and tag:leather and tag:used:project-name. An artist searching tag:worn
finds it. An artist searching tag:corrosion finds it. An artist searching
tag:used-* to see what reference actually made it to production finds it.
Build the taxonomy over time. Do not try to decide the perfect system on day one. Tag as you work, and after three projects, a pattern emerges. That pattern becomes the naming convention the studio standardizes on.
Scaling to team work
In a studio, the reference library is shared. Your project folder sits in a shared drive, and colleagues model alongside you. They need the same reference but may think of it differently.
You tagged weathering studies as tag:weathering:metal and tag:weathering:fabric.
Your colleague prefers tag:surface:corrosion and tag:surface:wear. You do
not reorganise the library. You both tag the same images with your own taxonomy.
A search by your tag pulls your collection. A search by theirs pulls theirs. When someone discovers a surface detail that belongs in both categories, they tag it twice. The image stays in one place on the shared drive. The metadata layers on top.
Third project: Scaling to team materials
By your third project, your material library is substantial. Thirty animated characters, each with their own texture sets, archived across three drives. The reference is not just for you anymore. Colleagues need to know where the good cloth deterioration samples are, or what weathering pattern you used on the last metal design.
Make your material library searchable by the team. Create permanent shared searches in a README or in your project documentation:
"Metal weathering reference: tag:weathering:metal rating:>=4 in:/Volumes/Archive"
"Fabric tear patterns: tag:weathering:fabric rating:>=4 in:/Volumes/Archive"
A new 3D artist on the team can search one line and access years of material reference. They do not have to know where files live or how you organize them. They just search by the principle they need: fabric, metal, leather, rust, paint-wear, corrosion. Sync reference across multiple external drives if your archive grows beyond one drive.
This is where a personal material library becomes a studio asset.
Reference management habits
This workflow assumes you rate as you go. Do not collect fifty images and rate them tomorrow. Rate each image the moment you add it:
- 5 star: "This is production-ready reference. I used this or I will definitely use this."
- 4 star: "This is good teaching reference. Different material, different project, but the principle applies."
- 3 star: "Useful. Has specific detail I might reference."
- 2 star: "I looked at it. Not quite right for this project."
- 0 star: "Mistake. Not what I was looking for."
After two projects, a search for tag:weathering rating:>=4 returns your
institution's best weathering reference, rated by use, not by time. That becomes
the starting point for project three.
Canvas files as texture painting guides
A canvas is a JSON file in your project folder. Drop a weathering canvas into your texture painting setup alongside your other project files:
/Volumes/3D-Work/Projects/armour-2026/
Anatomy/
...
HardSurface/
...
WeatheringDetails/
...
Boards/
Metal-weathering.coii-canvas
Fabric-deterioration.coii-canvas
model.blend
textures/
...Open the app, open the canvas, and reference while you paint. Eight plates of high-resolution weathering samples at canvas zoom. The file diffs in your revision history if you annotate them with notes. The texture painting happens in your regular tool; the reference stays organized and searchable.
At export time, the canvas goes with the project folder wherever it ships—to the render farm, to the next artist who continues the model, to your archive when the project finishes.
The difference between reference and asset
One mistake is treating material reference as textures to copy. They are not. A weathered metal texture is visual research. You look at it, understand the corrosion pattern and the colour shifts, and then paint your own version for your specific model, in your specific lighting, at your specific resolution.
This is why rating matters. A 5-star reference is one that taught you something about how a material actually deteriorates. A 2-star reference is one that looked good but would not teach you anything about the principle.
Over time, your rated reference becomes a visual education. Every surface detail you shipped, every material that sold the design, is connected to a rated reference. The best reference gets rated highest because it shaped the best surfaces.
Live reference capture during production
Sometimes the best reference is something you photograph or download mid-project. A client sends a fabric swatch. A colleague finds a weathering example that perfectly matches the brief. You discover a surface detail in a film that solves a texture problem.
Add it to your project reference folder immediately, tag it with today's date, rate it if it solves something, and search is faster than filing it away.
added:2026-09-11 type:imageThis pulls everything added today, so you can re-tag if needed—maybe it is weathered metal for sci-fi, or a fabric texture for a jacket, or a corrosion pattern that informs how you build the rust pass.
The key habit is: the moment you find reference, it lives in your project folder where it gets indexed, tagged, and rated. It does not live in downloads or email attachments. It does not need to be sorted into perfect folders first. It lives in the working reference folder and becomes searchable the moment the watcher picks it up.
Keeping search fast as the library grows
After five years, your archive has three thousand material reference images. Search is still one line because you tagged consistently. But you need to think about search structure as the library grows:
- Tag by material:
tag:metal,tag:fabric,tag:wood,tag:plastic - Tag by finish:
tag:weathered,tag:new,tag:aged - Tag by damage:
tag:corrosion,tag:dent,tag:scratch,tag:burn - Tag by project type:
tag:scifi,tag:fantasy,tag:contemporary
A search for sci-fi metal with corrosion is:
tag:scifi tag:metal tag:corrosion type:image rating:>=4Seven words, all your best reference for that specific material problem, searchable instantly. This is the power of consistent tagging. Learn to search effectively and your reference becomes an asset instead of a folder you scroll through.
How to organize reference images covers the folder structure that scales across multiple projects. How to rate and filter details the rating habit that makes search work. Organizing by project is the structure template. Tags vs. folders explains why metadata scales better than reorganising. And what file formats reference managers support covers the image formats that texture painting needs.