SESSION 16 · 10/26 · PARTLY WORKED
Thirty-five panels, and a dimming I never explained
The 127 on the AZ-GTi, run from the mount's hotspot. A 35-panel crescent mosaic at 19:44, anchored 11 arcminutes off the disc centre with a new shift option; a second, softer one after the image dimmed twentyfold at 21:33 and I blew into the tube thinking the mirror had fogged. Saturn reached the Mak at 23:02, and the recording of it was too small to use.
127 Mak on the AZ-GTi, mount on its own hotspot, SV205 in the finder in daylight then the Mak. Claude drove the night through rig-ui; Codex had set up the hotspot networking that afternoon. Written up on 18 October from the logs, the files and the session transcript.
Finder first, then the Mak. The Moon was centred and focused in the finder at 19:12 and aligned on. rig-ui's Calibrate refused: on the finder its step is 0.53 degrees, over the 0.4 limit. The camera went to the Mak at 19:30 and the Moon was 15 to 20 arcminutes off the finder's centre. Mak scale measured on the limb: 0.194 arcseconds a pixel.
Run 1, 19:44. 35 panels of 120 frames at full size, 2.5 GB each, anchored 11.4 arcminutes off the disc centre with the new shift option, since a 20% crescent's centre is dark. The highlights clipped because rig-ui kept re-enabling auto-exposure; that was fixed the same night.
rig-ui's own stitch failed twice with doubled limbs: it assumed a camera orientation about 13 degrees wrong for the 127. Stitching just the nine lit panels with plain shifts worked; the keeper above is the restitch done after midnight.
21:33, everything dimmed about twenty times and went soft. I thought the rear mirror had fogged and blew into the tube. Never explained. Run 2, 21:46, 36 panels of 45 frames at 115 ms with the Moon 25 degrees up: softer than run 1, but nothing clipped.
Saturn, 22:25 to 23:05. At 22:43 the sensor read all zeros; a replug and a torch test cleared it. Saturn was in the Mak at 23:02. At 23:50 something immovable stopped the mount and the motors were switched off.
| time | run | frames | result |
|---|---|---|---|
| 19:44 | mosaic 8x6, 35 panels | 120 | keeper after restitch; limb clipped |
| 21:33 | mosaic, aborted | - | image dimmed, folder empty |
| 21:46 | mosaic 8x6, 36 panels | 45 at 115 ms | soft, unclipped |
| 23:04 | Saturn 640x480 | 1802 | 25 px wide, unusable |
What I got wrong
- Recorded Saturn at 640 by 480. On the SV205 that is the whole sensor scaled down, not a crop. Planets need full size, or a real crop.
- Blew into the tube. Whatever dimmed the image, breath was not going to clear it.
Left for next time
- Find out what dimmed the image at 21:33: dew, the camera, or cloud.
- Fix rig-ui's stitch for the 127 (it assumed the 102's camera angle).
