Corona Version: Corona 14 Update 1
Cinema 4D Version: Cinema 4D 2026.1.3
Operating System: macOS 15.7.3 (Sequoia)
The Corona VFB continuously restarts Interactive Rendering (IR) without ever accumulating passes. Each render cycle completes internally but exits with “Rendered 0/10 passes”, immediately followed by a new “Parsing scene” and “RENDER STARTED”. The issue is reproducible and isolated to the VFB – Viewport IR works correctly.
Curiously, the loop sometimes resolves temporarily when interacting with the Cinema 4D Object Manager – specifically by expanding a hierarchy in the object tree and selecting an object within it. This suggests the root cause is related to an internal dirty/state flag that is incorrectly set and only cleared by certain UI interactions.
Steps to Reproduce
1. Open Cinema 4D with a test.c4d scene
2. Start Interactive Rendering via the Corona VFB
3. VFB shows “Parsing…” → “Rendering” → exits with Rendered 0/10 passes
4. Immediately restarts: “Parsing scene” → “RENDER STARTED”
5. Loop repeats indefinitely
6. (Intermittent workaround): Expand a hierarchy in the C4D Object Manager and select an object → IR sometimes stabilizes temporarily
Intermittent Workaround (Unreliable)
Expanding a group/hierarchy in the Cinema 4D Object Manager and clicking on an object within it sometimes causes IR to stabilize and begin accumulating passes normally. This workaround is not reliable and does not always work.
This behavior is a strong indicator that the bug is caused by a persistent dirty/scene-change flag that is never cleared during VFB IR. The Object Manager interaction appears to inadvertently trigger a state flush or scene re-evaluation that clears the flag, allowing IR to proceed. This mechanism does not exist or is handled differently in the Viewport IR pipeline, which is why Viewport IR is unaffected.
Console Log Analysis
The console output reveals a key anomaly: the render frame logs only Pass 1/10 being rendered, yet the session exits with “Rendered 3/10 passes” – suggesting passes are being counted across restarts rather than within a single stable session. The entire render frame completes in under 1 second ( Rendering took 0.221 seconds ) before immediately terminating and restarting.
Notable observations from the log:
• REPORT: unchanged: 43, Prim: 0, Inst: 0, Mtl: 0, Env: false – the scene reports no changes, yet the render restarts anyway
• Avg samples per pixel: 0.398 – well below 1, indicating the pass is aborted before completion
• UHDCache rejected %: 89.7% – unusually high rejection rate for the secondary GI cache
• DR (Distributed Rendering) slave termination is called twice per cycle, which may indicate an unexpected shutdown sequence
2026/02/19 11:17:53 ====RENDER STARTED====
2026/02/19 11:17:53 Core build timestamp: Dec 9 2025 19:39:09
2026/02/19 11:17:53 Pass 1/10
2026/02/19 11:17:54 CoronaCore::renderFrame: after render
2026/02/19 11:17:54 Terminating DR slaves ← first termination
2026/02/19 11:17:54 Terminating DR slaves ended
2026/02/19 11:17:54 Rendering took 0.221000001 seconds
2026/02/19 11:17:54 Cleaning up
2026/02/19 11:17:54 Terminating DR slaves ← second termination (unexpected)
2026/02/19 11:17:54 Terminating DR slaves ended
2026/02/19 11:17:54 Avg samples per pixel: 0.398539305
2026/02/19 11:17:54 UHDCache rejected %: 89.7482224
2026/02/19 11:17:54 Rendered 3/10 passes ← pass count mismatch
test.c4d (3.07 MB)
