Mirror/flip/rotate/digital zoom are baked into the stored frame (2026-09-12 pass), which is right: the frame on disk is the record the timelapse and the coming plant regions read. The consequence is that changing the framing re-frames the record from that moment on – frames either side of the change differ in geometry and a timelapse spanning it jumps. Today the drawer only says so. A later pass could stamp a framing epoch per frame or day folder and have assemble_timelapse either refuse a mixed range with a clear reason or normalise by letterboxing to a common frame. Also relevant to S7 plant regions, which are pixel regions in a fixed frame.
Journey event · 12 Sep 2026
A camera framing change splits the timelapse geometry mid-run
Mirror/flip/rotate/digital zoom are baked into the stored frame (2026-09-12 pass), which is right: the frame on disk is the record the timelapse and the coming plant regions read.
● state copied from the tracker, never improved
● event date 2026-09-12
| Event date | 2026-09-12 |
| Version found | |
| Version fixed | |
| Expected kit firmware | |
| Release file | |
| SHA-256 | |
| Served count | |
| Source | https://app.notion.com/3d92b4cda37081838e47c05b0951d098 |
All events · DSC · CannaLib · Site · Firmware releases
Empty rows are honest
A blank cell means the tracker has no value for it. Nothing on this page is filled in to look complete.
