Case No. 2 Turntable
A 24-frame turntable of headphones
orbit=True circles the camera around the object and returns an MP4. No keyframes in the script.
This 24-frame headphone turntable was rendered as an MP4 on a JANCTION GPU with orbit=True. The orbit camera circles every non-flat object and ignores the thin pedestal when it frames the shot, so the script needs no camera animation. It took 58.3 GPU seconds on the production GPU, 5.8 JPY at 0.1 JPY per GPU-second.
What is in the frame
A stylised pair of headphones on a low pedestal, seen from every side in one loop. Turntables are the standard way to show a product or a model in a portfolio, and they are the first thing people ask an AI to make once a model exists.
How the scene is built
The script contains the headphones and the pedestal, nothing else. orbit=True adds a camera that circles the scene; orbit_frames sets how many frames make up the turn. The pedestal is flat, so the framing logic ignores it and keeps the headphones filling the shot.
Adapt it
Swap in your own model: a glTF, FBX or USD file is imported into an empty scene, so the same call works without writing any Blender code. This one took 58.3 GPU seconds, about 5.8 JPY.
5.8 JPY for this render. The free welcome credit of a new key covers about 17 of them, with no card.
Render this example with your AI
ChatGPT · Claude · Grok · Perplexity
Use JANCTION Render (MCP server https://render.janction.jp/mcp) to render this example: render_preview(scene_url="https://render.janction.jp/samples/ex02_turntable_headphones.py"). Then suggest how to change the colors, shapes and camera for my own subject. Ask me before the final render.
Each link opens the app with the text below filled in (if it is not, copy and paste it). If JANCTION Render is not connected in that app yet, add https://render.janction.jp/mcp as a connector first.
Blender warnings
- orbit: an orbit camera circles the scene once over 24 frames (elevation 18 deg); the scene's own camera is not used
- environment preset 'studio' replaced the scene's world (HDRI lighting)
The script (bpy, runs as is)
ex02_turntable_headphones.py. It builds the scene, the camera and the frame range; resolution, samples, denoising and the GPU are set by the service.
"""JANCTION Render example 02: turntable of a stylized pair of headphones (orbit=true, MP4).
Reproduce: render_final(scene_url="https://render.janction.jp/samples/ex02_turntable_headphones.py",
orbit=True, orbit_frames=24, width=1280, height=720, samples=64,
environment="studio", output="mp4")
The orbit camera circles the non-flat objects; the thin pedestal is flat and is ignored for framing.
"""
import math
import bpy
from mathutils import Vector
for ob in list(bpy.data.objects):
bpy.data.objects.remove(ob, do_unlink=True)
scene = bpy.context.scene
scene.frame_start = 1
scene.frame_end = 1
def material(name, color, roughness=0.5, metallic=0.0, **inputs):
mat = bpy.data.materials.new(name)
mat.use_nodes = True
bsdf = mat.node_tree.nodes["Principled BSDF"]
bsdf.inputs["Base Color"].default_value = (*color, 1.0)
bsdf.inputs["Roughness"].default_value = roughness
bsdf.inputs["Metallic"].default_value = metallic
for key, value in inputs.items():
bsdf.inputs[key].default_value = value
return mat
def smooth(ob):
for p in ob.data.polygons:
p.use_smooth = True
return ob
def add(op, name, mat=None, **kw):
getattr(bpy.ops.mesh, op)(**kw)
ob = bpy.context.object
ob.name = name
if mat is not None:
ob.data.materials.append(mat)
return ob
def look_at(ob, target):
ob.rotation_euler = (Vector(target) - ob.location).to_track_quat("-Z", "Y").to_euler()
def camera(location, target, lens=50.0):
data = bpy.data.cameras.new("Camera")
data.lens = lens
cam = bpy.data.objects.new("Camera", data)
scene.collection.objects.link(cam)
cam.location = location
look_at(cam, target)
scene.camera = cam
return cam
def light(name, kind, location, energy, target=None, size=1.0, color=(1.0, 1.0, 1.0)):
data = bpy.data.lights.new(name, kind)
data.energy = energy
data.color = color
if kind == "AREA":
data.size = size
ob = bpy.data.objects.new(name, data)
scene.collection.objects.link(ob)
ob.location = location
if target is not None:
look_at(ob, target)
return ob
plastic = material("Plastic", (0.07, 0.07, 0.08), roughness=0.45)
accent = material("Accent", (0.85, 0.62, 0.3), roughness=0.25, metallic=1.0)
cushion = material("Cushion", (0.11, 0.11, 0.12), roughness=0.95)
base = material("Pedestal", (0.75, 0.75, 0.73), roughness=0.5)
band = add("primitive_torus_add", "Headband", plastic, major_radius=1.0, minor_radius=0.075, major_segments=96,
minor_segments=24, location=(0, 0, 0.9), rotation=(math.pi / 2, 0, 0))
smooth(band)
for sign, side in ((-1, "L"), (1, "R")):
cup = add("primitive_cylinder_add", "Cup" + side, accent, radius=0.42, depth=0.26, vertices=96,
location=(sign * 1.05, 0, 0.9), rotation=(0, math.pi / 2, 0))
smooth(cup)
pad = add("primitive_torus_add", "Pad" + side, cushion, major_radius=0.34, minor_radius=0.1, major_segments=96,
minor_segments=24, location=(sign * 0.9, 0, 0.9), rotation=(0, math.pi / 2, 0))
smooth(pad)
arm = add("primitive_cylinder_add", "Arm" + side, plastic, radius=0.05, depth=0.5, vertices=32,
location=(sign * 1.0, 0, 1.25), rotation=(0, sign * 0.35, 0))
smooth(arm)
pedestal = add("primitive_cylinder_add", "Pedestal", base, radius=1.7, depth=0.04, vertices=128, location=(0, 0, 0.02))
smooth(pedestal)
light("Key", "AREA", (2.5, -3.0, 3.5), 300, target=(0, 0, 0.9), size=2.0)
camera((0.0, -6.0, 1.7), (0.0, 0.0, 0.9), lens=50)

