Tools Field of View & Framing
Field of View & Framing
A target drawn at its catalogue size against your sensor, at any camera angle. You get the field, the image scale and how much of the frame the target fills, and when it does not fit in one frame, the mosaic that covers it. Put a second rig beside the first to compare them on the same target.
Target sizes are OpenNGC’s catalogue extents, as published. The stars and the glow are synthetic, drawn in your browser; no survey imagery is used.
Telescope
766 mm. Source
Camera
19.1 × 13 mm, 4.63 µm pixels. Source
OpenNGC NGC0224: 177.83′ × 69.66′, position angle 35°.
Camera angle
0° lays the sensor's width along the target's long axis. With OpenNGC's position angle of 35°, the width now runs at position angle 35°, east of north.
Compare
Mosaic overlap and seeing
ZWO FF110 APO at 766 mm, f/7.0, with the ToupTek ATR294C.
M31 Andromeda Galaxy does not fit one frame (207 % along the sensor's width). 3 × 2 panels at 15 % overlap cover its catalogue extent, 6 frames in all.
- Turned 30° from its long axis, it spans 184 %: the least of any angle.
- At 90°, 1 × 4 panels cover it, 4 in all.
At 2.5″ seeing, 1.25″ per pixel puts 2.0 pixels across the seeing disc. Practical band for typical seeing.
The arithmetic for this rig
W = 2 × atan(19.1 / (2 × 766)) = 1.4286° = 85.71′
H = 2 × atan(13 / (2 × 766)) = 0.9724° = 58.34′
X = √(177.83² cos²0° + 69.66² sin²0°) = 177.83′
Y = √(177.83² sin²0° + 69.66² cos²0°) = 69.66′
fill = max(X / W, Y / H) = max(2.075, 1.194) = 2.075
panels 3 × 2 at o = 0.15
scale = 206.265 × 4.63 / 766 = 1.247″/pxHow this is computed
scale["/px] = 206.265 × pixel[µm] / focal length[mm]
W, H [°] = 2 × atan(sensor width, height[mm] / (2 × focal length[mm])) × 180/π
X = √(a² cos²θ + b² sin²θ) extent along the sensor's width
Y = √(a² sin²θ + b² cos²θ) extent along its height
frame filled = max(X / W, Y / H)
panels = 1 + ⌈(T − F) / (F × (1 − o))⌉ along each axis, 1 when T ≤ F
least fill tan²θ* = (a²/W² − b²/H²) / (a²/H² − b²/W²)
sampling = seeing FWHM["] / scale["/px]
a, b = OpenNGC's major and minor axes, θ = camera angle from the major axis,
T = X or Y, F = W or H, o = overlap between panelsField and scale use the exact forms; the worked example shows the arithmetic. The focal length is the telescope’s, times a corrector’s factor only where its maker states one. Where the maker publishes no factor, the figures stay at the native focal length and the tool says so.
Frame filled is linear: the target’s extent along whichever sensor axis it fills more, as a fraction of the field along that axis. 100 % means the catalogue outline touches both edges of that axis. It is not an area ratio, and at 0° it is the figure the Rig Reality Check and the rig pages print. Those two sort it into four bands for scoring; this tool reports the number and the panel count instead.
Turning the camera does not always make a long target fit worse. Between 0° and 90° the share along the width only falls and the share along the height only rises, so the larger of the two is smallest where they meet: for M31 on a 23.5 × 15.7 mm sensor at 400 mm, at 29°, where it spans 79 % against 88 % square to the galaxy. The panel count is a step function of the angle and is found by trying every whole degree.
A mosaic here covers the outline’s bounding box on a grid that turns with the camera, centered on the target. Panel steps are angles on the sky, which is what the mount slews; for fields of a few degrees the difference from a projected grid is far below the overlap. The default overlap is 15 %; set it to what your stitching software wants.
Position angles are OpenNGC’s, from HyperLEDA, for the 5 galaxies that have one (M31, M101, M51, M81, NGC 4565). For those the figure is drawn north up and east left, and the tool gives the position angle of the sensor’s width. The other 8 targets have none in the table, so their figure is drawn with the long axis level and no north.
The sampling line is a tradeoff, never a pass or a fail. Every band, as the physics reference sets them out:
| Seeing disc, px | What it means |
|---|---|
| < 0.8 | Badly undersampled: the pixel is wider than the seeing disc, so stars land as blocky squares |
| 0.8 – 1.4 | Undersampled; the standard, deliberate wide-field trade |
| 1.4 – 3.2 | Practical band for typical seeing |
| 3.2 – 4.5 | Oversampled: lower per-pixel signal and tighter guiding; binning recovers most of it |
| > 4.5 | Heavily oversampled: the atmosphere limits resolution long before the pixel does |
Target sizes: OpenNGC by Mattia Verga (tag v20260501), CC BY-SA 4.0. This table is shared under the same license. Source and license
Source: docs/03-physics-reference.md § 1, Optics and framing, and its Framing subsection with worked examples for M31 on an APS-C sensor at 250, 400 and 1000 mm. Implemented in packages/physics (optics.ts, framing.ts), with those examples as unit tests.
Last updated