blender-cameras
Skill2 starsChanged 3 days agoMITView on GitHub
Set up Blender cameras with cinematic intent — focal length, depth of field (f-stop / focus object), composition (rule of thirds, leading lines), animated cameras (orbit, dolly, push-in), tracking constraints. Use whenever the user asks to "set up the camera", "frame the shot", "make it look cinematic / hero / portrait / wide-angle / telephoto", "add depth of field", "orbit the camera", or any composition/framing request. Make sure to use this skill even if the user does not say "camera" — also covers "hero shot", "close-up", "from above", "shallow focus", "85mm portrait look".
1 other project uses a version of this, 1 of them edited.
Adopted by
Most recently taken up by
- keysforthewin/thisofficeedited, first seen Jul 24, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 1
- keysforthewin/thisoffice★ 098% same
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository33 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .github/copilot-instructions.mdCopilot instructions
- .cursor/rules/3d-fabric.mdcCursor rule
- 3d-fabricSkill
- animation-quality-gateSkill
- atlas-uv-fittingSkill
- blender-animationSkill
- blender-exportSkill
- blender-lightingSkill
- blender-materialsSkill
- blender-modelingSkill
- blender-pro-workflowSkill
- blender-renderingSkill
- blender-skill-harmonizerSkill
- blender-uv-texturingSkill
- closed-surface-uv-coverageSkill
- contour-to-meshSkill
- fit-repair-optimizerSkill
- landmark-fit-repairSkill
- mascot-logo-reconstructionSkill
- multiview-constraint-solverSkill
- multiview-fit-loopSkill
- orbital-hud-motionSkill
- orthographic-registrationSkill
- quality-refinement-autoloopSkill
- reference-analysis-validatorSkill
- reference-look-calibrationSkill
- reference-to-3dSkill
- source-part-segmentationSkill
- text-to-blenderSkill
- texture-driven-mesh-fittingSkill
- texture-state-animationSkill
- wireframe-to-3dSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 6 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈6
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 10
Use when: Any camera placement, framing, focal length, DoF, or animated camera setup in Blender.
- Distinctive: line 12
Before the first Blender call, read ../text-to-blender/references/mcp-backends.md.
- Distinctive: line 44
Use this when you have a specific subject.
- Distinctive: line 273
Rule of thirds: place the subject at one of the 4 intersection points, not center.
- Distinctive: line 274
Headroom: leave ~10% empty above the head.
- Distinctive: line 275
Nose room: if subject faces left, leave space on the left for them to "look into".
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
MCP backend
Use when: Any camera placement, framing, focal length, DoF, or animated camera setup in Blender.
Before the first Blender call, read ../text-to-blender/references/mcp-backends.md. Prefer official MCP for shared capabilities; preserve the selected backend for the whole task. Tool names in recipes are client/legacy examples: map them through that reference and its agent-host guide.
Blender Cameras
Set up cameras with the same decisions a real cinematographer makes: focal length for feel, f-stop for focus, composition for storytelling.
Focal length cheat sheet
| Length | Feel | Use |
| 14–24mm | Very wide, distorted | Architecture, claustrophobic interiors, exaggerated perspective |
| 28–35mm | Wide, "documentary" | Establishing shots, environments |
| 50mm | Neutral (≈ human eye) | Default storytelling |
| 85mm | Short telephoto | Portraits, character close-ups (flattering) |
| 100–135mm | Telephoto | Hero product shots, isolated subjects |
| 200mm+ | Long tele | Wildlife, surveillance look, heavy compression |
Quick rule: 85mm for intimacy, 24mm for spectacle, 50mm for neutral.
Aperture / f-stop
| f-stop | DoF | Use |
| f/1.2–2.0 | Razor thin | Hero portraits, dreamy |
| f/2.8 | Shallow | Standard portrait |
| f/4 | Moderate | Two subjects in frame |
| f/5.6–8 | Medium-deep | Group portraits, environments |
| f/11+ | Very deep | Landscape, "everything sharp" |
Recipes
Recipe 0 — Bbox-aware hero camera (preferred for orchestrator chains)
Use this when you have a specific subject. Computes the subject's bounding box, places camera at a distance that fits the subject in ~80% of the frame vertically, and aims via Track-To.
import bpy, math
from mathutils import Vector
# Choose subject — all meshes named GEO-* by default, or pass a specific list
subject_meshes = [o for o in bpy.data.objects if o.type == 'MESH' and o.name.startswith('GEO-')]
if not subject_meshes:
raise RuntimeError("No subject meshes found (looking for GEO- prefix)")
# World-space bbox
deps = bpy.context.evaluated_depsgraph_get()
all_verts = []
for o in subject_meshes:
eo = o.evaluated_get(deps); em = eo.to_mesh()
for v in em.vertices:
all_verts.append(o.matrix_world @ v.co)
eo.to_mesh_clear()
xs = [v.x for v in all_verts]; ys = [v.y for v in all_verts]; zs = [v.z for v in all_verts]
center = Vector(((min(xs)+max(xs))/2, (min(ys)+max(ys))/2, (min(zs)+max(zs))/2))
height = max(zs) - min(zs)
width = max(xs) - min(xs)
biggest = max(height, width)
# Frame fit: at distance D, vertical frame = D × (sensor_h / focal). Solve for D.
focal_mm = 60 # 60mm gives a flattering not-too-wide hero shot
sensor_h_mm = 24 # full-frame
frame_per_meter = sensor_h_mm / focal_mm # 0.4 m vertical frame per metre of distance
target_fill = 0.80
camera_distance = biggest / (frame_per_meter * target_fill)
# Camera positioned in front (negative Y) with slight X offset for a 3/4 angle
cam_pos = Vector((center.x + camera_distance * 0.3, center.y - camera_distance, center.z))
# Empty for tracking
empty_name = 'Empty-camera_target'
empty = bpy.data.objects.get(empty_name) or bpy.data.objects.new(empty_name, None)
if empty.name not in [o.name for o in bpy.context.collection.objects]:
bpy.context.collection.objects.link(empty)
empty.location = center
# Camera
cam_data = bpy.data.cameras.new('CAM-hero')
cam_data.lens = focal_mm
cam_data.dof.use_dof = True
cam_data.dof.aperture_fstop = 4.0
cam_data.dof.focus_object = subject_meshes[0] # focus on first/main subject
cam = bpy.data.objects.new('CAM-hero', cam_data)
bpy.context.collection.objects.link(cam)
cam.location = cam_pos
track = cam.constraints.new('TRACK_TO')
track.target = empty
track.track_axis = 'TRACK_NEGATIVE_Z'
track.up_axis = 'UP_Y'
bpy.context.scene.camera = cam
print(f"camera:bbox_aware center={tuple(round(v,2) for v in center)} dist={camera_distance:.2f}m focal={focal_mm}mm")For elongated vertical subjects (sword, flag, candle): biggest dimension is height; the framing math fits height to 80% of vertical frame, which is what you want.
For wide horizontal subjects (car, table): biggest is width; it fits width to 80% of vertical frame too which over-zooms — for those, swap to frame_per_meter_h = (sensor_h_mm * aspect_ratio) / focal_mm or adjust target_fill down.
Recipe 1 — Hero portrait camera (85mm + shallow DoF)
import bpy, math
subject = bpy.data.objects.get('GEO-subject') # change to your subject
# Camera
cam_data = bpy.data.cameras.new('CAM-hero')
cam = bpy.data.objects.new('CAM-hero', cam_data)
bpy.context.collection.objects.link(cam)
bpy.context.scene.camera = cam
cam.location = (3, -4, 1.6)
cam_data.lens = 85
cam_data.sensor_width = 36
# Depth of field
cam_data.dof.use_dof = True
cam_data.dof.aperture_fstop = 2.8
if subject:
cam_data.dof.focus_object = subject
# Track-to constraint (auto-aim at subject)
if subject:
track = cam.constraints.new('TRACK_TO')
track.target = subject
track.track_axis = 'TRACK_NEGATIVE_Z'
track.up_axis = 'UP_Y'
print('camera:CAM-hero set')Recipe 2 — Wide environmental establishing shot (24mm)
import bpy, math
cam_data = bpy.data.cameras.new('CAM-establish')
cam = bpy.data.objects.new('CAM-establish', cam_data)
bpy.context.collection.objects.link(cam)
bpy.context.scene.camera = cam
cam.location = (8, -10, 2.5)
cam.rotation_euler = (math.radians(80), 0, math.radians(35))
cam_data.lens = 24
cam_data.sensor_width = 36
cam_data.dof.use_dof = False
print('camera:CAM-establish (24mm wide)')Recipe 3 — Product hero (100mm + macro DoF)
import bpy, math
subject = bpy.data.objects.get('GEO-product')
cam_data = bpy.data.cameras.new('CAM-product')
cam = bpy.data.objects.new('CAM-product', cam_data)
bpy.context.collection.objects.link(cam)
bpy.context.scene.camera = cam
cam.location = (0.4, -1.5, 0.2) # close-in
cam_data.lens = 100
cam_data.sensor_width = 36
cam_data.dof.use_dof = True
cam_data.dof.aperture_fstop = 2.0
if subject:
cam_data.dof.focus_object = subject
if subject:
track = cam.constraints.new('TRACK_TO')
track.target = subject
track.track_axis = 'TRACK_NEGATIVE_Z'
track.up_axis = 'UP_Y'
print('camera:CAM-product (100mm hero)')Recipe 4 — Composition guides (rule-of-thirds overlay)
import bpy
cam_data = bpy.data.cameras['CAM-hero']
cam_data.show_composition_thirds = True
cam_data.show_composition_golden = False
cam_data.show_composition_center = False
print('composition:thirds_on')These overlays show in viewport only; no effect on render.
Recipe 5 — Orbit camera animation (10-second 360° turntable)
import bpy, math
target = bpy.data.objects.get('GEO-subject')
# Empty as pivot
pivot = bpy.data.objects.new('Empty-orbit_pivot', None)
bpy.context.collection.objects.link(pivot)
if target:
pivot.location = target.location
# Camera child of pivot
cam_data = bpy.data.cameras.new('CAM-orbit')
cam = bpy.data.objects.new('CAM-orbit', cam_data)
bpy.context.collection.objects.link(cam)
cam.parent = pivot
cam.location = (0, -5, 0.5)
cam.rotation_euler = (math.radians(85), 0, 0)
cam_data.lens = 50
bpy.context.scene.camera = cam
# Animate pivot's Z rotation: 0 → 360° over frames 1..240 (10s @ 24fps)
pivot.rotation_euler = (0, 0, 0)
pivot.keyframe_insert('rotation_euler', frame=1)
pivot.rotation_euler = (0, 0, math.radians(360))
pivot.keyframe_insert('rotation_euler', frame=240)
# Set linear interpolation for constant orbit speed
def get_fcurves_compat(action):
if hasattr(action, 'fcurves'):
return list(action.fcurves)
return [fc for layer in action.layers for strip in layer.strips
if hasattr(strip, 'channelbags') for bag in strip.channelbags
for fc in bag.fcurves]
if pivot.animation_data and pivot.animation_data.action:
for fc in get_fcurves_compat(pivot.animation_data.action):
for kp in fc.keyframe_points:
kp.interpolation = 'LINEAR'
print('camera:CAM-orbit (10s turntable)')Recipe 6 — Push-in / dolly (camera moves forward, no zoom)
import bpy
cam = bpy.data.objects['CAM-hero']
# Start position
cam.location = (3, -8, 1.6)
cam.keyframe_insert('location', frame=1)
# End position (closer to subject)
cam.location = (3, -4, 1.6)
cam.keyframe_insert('location', frame=120)
print('camera:dolly_5s')A push-in (physical move forward) is visually distinct from a zoom (focal length change). Use push-ins for cinematic feel, zooms for surveillance/news look.
Composition rules — enforce via positioning
- Rule of thirds: place the subject at one of the 4 intersection points, not center.
- Headroom: leave ~10% empty above the head.
- Nose room: if subject faces left, leave space on the left for them to "look into".
- Foreground/midground/background: three depth layers feel more cinematic.
Sensor sizes
| Sensor | Width (mm) | Notes |
| Full frame DSLR / 35mm cinema | 36 | Default |
| APS-C | 22.5 | 1.5–1.6× crop |
| Super 35 | 24.89 | Most cinema |
| Micro Four Thirds | 17.3 | Mirrorless |
| iPhone 15 Pro | 9.8 | Smartphone reference |
Set with cam_data.sensor_width = 36.
Common pitfalls
| Symptom | Fix |
| Distorted face on portrait | Use 50mm+ for human subjects |
| Subject blurred, background sharp | Set dof.focus_object to the subject |
| Camera dead-center on subject | Apply rule of thirds; offset subject |
| Orbit camera tilts wildly | Use Track-To constraint, not manual rotation |
| Camera below ground in animation | Add Floor constraint or check Z keyframes |
| DoF very slow in Cycles | Acceptable for finals; viewport may use simpler approximation |
When to load references/overview.md
Load when:
- Cinematic effects beyond defaults: anamorphic, lens flares, vignette
- Multi-camera scenes (camera markers for editing)
- Camera shake / handheld noise
- Stereo / VR camera setup
The reference covers: full focal length theory, aperture/f-stop tables, composition guides, sensor variants for matching real cameras (iPhone, cinema, DSLR), animated camera patterns, cinematic effects.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

