Stacking Parameters
The parameters in the stacking table, one by one. Each target type – Lunar, Planetary, Solar – keeps its own settings. Values are numbers; 0 means off and 1 means on unless stated.
Eight of them also have a control elsewhere on the Stack tab. The table and the control always agree: change either one and the other follows.
Frames
AutoFrames — let BiggSky choose how many frames to stack, from the quality curve it measures during analysis. On by default. Turn it off to stack exactly the percentage you ask for.
FramePercent — the percentage of the video's frames to stack, best first. With AutoFrames on this is the starting point BiggSky adjusts; with it off it is used as given. A list is allowed (for example 10 25 50), which stacks each count in turn so you can compare them.
MinFrames, MaxFrames — bounds on the automatic count. MaxFrames 0 means no upper bound. Use MaxFrames to keep a long video's stack to a time you are willing to wait for.
MaxClipTimeSec — split a long video into clips of this many seconds and stack each clip separately. 0 stacks the whole video as one. Use it when the subject rotates or the seeing changes through the capture.
MultiTimePoint — how many of those clips to process: 0 processes one, 1 processes as many as the video holds, and a larger number processes exactly that many.
WinjuposFormat — with MultiTimePoint, name each clip's stack with the clip's own time, in UT, the way WinJUPOS reads it (2026-01-17-0524_4-U-R-Jup_Stack2_T02_F0702.tif), so WinJUPOS, Combine Decons to Image and Interpolate Decons to Movie see every time point at its own time. On (1) by default. 0 names every clip's stack after the capture. Needs a SER capture; see the Derotation manual.
Image quality
FixHotPixels — remove hot pixels from every frame before alignment. 1 is the current filter, 2 the older one, 0 off. Leave it on; hot pixels otherwise pull the alignment and survive stacking.
FixPatternNoise — remove the camera's fixed row/column pattern from every frame. On by default.
AutoDark, AutoFlat — apply the dark and flat frames attached to the project. Off unless you have calibration frames for this capture.
Alignment
AlignPasses — 1 aligns each frame as a whole (shift only). 2 adds the second pass that corrects the local distortion seeing leaves across the frame, which is what sharpens fine detail on a planet or a crater field. 2 is the default and is what you normally want.
Dewarp — 1 (default) lets the second pass correct the local distortion. 0 keeps the second pass but aligns each frame by a single sub-pixel shift instead, and skips the dewarp calibration. Use 0 for stars and double stars: there is no surface detail to track, and the dewarp can pull a close pair together. SuperRes still works with Dewarp 0.
AlignChannel — for colour cameras, which channel the alignment measures. 0 green (default), 1 red, 2 blue, 3 all three averaged. Green is the best-sampled channel on most sensors. Choose another only when one channel is spoiled on that capture, for example blue at low altitude.
Super-resolution
SuperRes — the output scale. 1 stacks at the camera's own pixel scale; 1.5 and 2 stack onto a finer grid, which gives deconvolution more room when the image is undersampled. A list is allowed (1 1.5 2) and produces one stack per scale.
Output
BorderCrop — pixels trimmed from each edge of the finished stack, where frames do not overlap after alignment. 8 by default.
Save32bit — save the stack in 32-bit floating point as well as the standard 16-bit file. Use it when the stack will be deconvolved or stretched hard.
SaveFrames — save the aligned frames alongside the stack. They take a lot of disk; this is for checking alignment or for reprocessing without re-aligning.
AnalyzePlotSave — save the frame-quality plot and its CSV with the results.
StackSuffix — text added to the output file names, so several runs of the same video sit side
StackPreview — what the image area shows while the frames are being aligned and added. 1 (the default) shows four views, refreshed every couple of seconds: the raw frame (with its place in the stack), the same frame after dewarping, a turbulence map of how far and which way each part of that frame was pulled (color is direction, brightness is distance, 2 pixels is full color; the caption gives the frame's turbulence RMS in pixels), and the stack so far. 0 shows the stack alone. It does not change the result and costs about 1% of the stacking time. Large frames are shown reduced. by side. Leave it empty for the default naming.
Show advanced
The checkbox under the table adds the expert and debug parameters: the dewarp's internals, the sharpness weighting, the frame-selection tuning, the noise-model diagnostics, and the multi-frame deconvolution settings. They are applied whether or not they are shown, and the defaults are the measured ones. Change them only with a reason to: several trade detail against noise, and a few exist only to compare two implementations.
The same parameters are listed by BiggSkyCLI params and can be set with --stack Name=value; see the BiggSkyCLI manual. See also the Deconvolution Expert settings and the full BiggSky manual.