Derotation Manual

Jupiter turns about 0.6° a minute. Capture for longer than the usual limit, then turn the images back to a common time: Derotate Stacks before deconvolution, Combine Decons to Image and Interpolate Decons to Movie after it. Derotation is part of the solar system editions (SSE Lite, SSE Pro and Universe).

All manuals Dewarp Movie manual Compare editions

Derotation: what do you want to make?

Jupiter turns about 0.6° a minute, Saturn a little slower. In a long capture, or between captures minutes apart, the clouds move. BiggSky has three functions that turn images back to a common time, one used before deconvolution and two after it:

Capture ─► Stack (Multi Time) ─► Derotate Stacks ─► Decon ─┬─► Combine Decons to Image
                                                           └─► Interpolate Decons to Movie

Where to find them. On the Stack tab, open the Functions tab. The Dewarp/Derotate panel has a button for each, in the order you use them:

ButtonWhat it does
Create Dewarp Movie…A movie of the seeing (see the Dewarp Movie manual).
Derotate Stacks…Derotates a series of stacks before deconvolution.
Combine Decons to Image…Turns finished images to one time and combines them.
Interpolate Decons to Movie…Makes evenly timed frames and a movie from finished images.

Each works on the dataset selected in the project. The same four are in the menu bar under Functions › Stacking, which also has Set WinJUPOS Observer… (see below).

Derotation is part of the solar system editions (SSE Lite, SSE Pro and Universe). In the Free and DSO editions these items are hidden.

You have…You want…Use
One Jupiter video longer than about a minuteOne sharp image from all of itStack with Multi Time and Derotate, or Derotate Stacks afterwards
A continuous capture of 20 minutes or moreA movie in which every frame is as clean as a 10-minute captureDerotate Stacks (moving average), Decon, then Interpolate Decons to Movie
Stacks a minute or two apart, from BiggSky or AutoStakkert!One image, or a moving average, to deconvolveDerotate Stacks
A few finished images 15 to 25 minutes apart through a nightA smooth animationInterpolate Decons to Movie
Several finished images over 20 to 30 minutesOne image with more signal, moons and shadows in the right placeCombine Decons to Image
Mono R, G, B (and IR, CH4 …) images taken in turnOne color image, and LRGB, at one momentCombine Decons to Image
Saturn through filtersA color image or movie with sharp ring spokesCombine Decons to Image or Interpolate Decons to Movie

The rule:

  • Derotate Stacks is for many stacks close together in time (no more than a couple of minutes apart), before deconvolution. It carries each stack's noise into the result, so deconvolution works on it as it does on any stack.
  • Combine and Interpolate are for finished images after deconvolution: a few images far apart, images through filters, or anything with a moon shadow on the disk.
  • When the stacks are close enough for Derotate Stacks, use it rather than deconvolving each stack and combining the results: deconvolution then works once on the full signal.

What the outputs are

NameWhat it isGood forDon't
DRC, combinedOne image at one time, from every inputThe image to deconvolve (Derotate Stacks) or to publish (Combine)
DRA11, moving averageOne frame per input time, each averaging 11 neighborsDeconvolving frame by frame, then a movieStack the frames again: neighbors share most of their data
DRN5, nearestOne frame per stack, each from the 5 stacks nearest it in timeRarely needed
DRI1min, interpolatedFrames every minute between the key imagesAnimationMeasure or stack them: an in-between frame has no more signal than the two keys it came from

Things you may not know you can do

  • Capture longer than the rotation limit. The usual advice keeps a Jupiter capture to a minute or two. Capture for longer and split it into 60 s clips: a 3-minute capture derotated this way gave 9 to 25% more detail than the same 3 minutes stacked as one.
  • Capture continuously and get a movie. An hour of capture becomes an animation in which each frame carries the signal of about 10 minutes.
  • Deconvolve derotated images properly. Each Derotate Stacks output has its own noise model, which a derotation made in another program does not keep.
  • No WinJUPOS needed. The key images are timed by their rotation and by Jupiter's moons. A WinJUPOS measurement is used when there is one, but none is required.
  • Smooth animations. Each frame is derotated to its own time, so the planet turns evenly instead of jumping at every key image.
  • Color from mono filters at one moment, and LRGB with IR or L as luminance.
  • Saturn's spokes stay sharp: each part of the rings turns at its own orbital speed.
  • Wrong time labels are found. A key image whose label disagrees with its rotation is reported.
  • Field rotation from an alt-az mount is measured and removed (filter sets: Saturn, Mars and mono Jupiter).

Stacks named for WinJUPOS (WinjuposFormat)

When you stack with Multi Time, each clip's stack is named with the clip's own time in the form WinJUPOS reads:

YYYY-MM-DD-HHMM_T-Observer[-Image info]_Stack2_T<nn>_F<frames>.tif
2026-01-17-0524_0-U-R-Jup_Stack2_T01_F0250.tif
2026-01-17-0524_4-U-R-Jup_Stack2_T02_F0250.tif

The time is the middle of the clip, in UT, to the tenth of a minute (the tenth is truncated, as FireCapture does it). The deconvolved and View images keep the stack's name, so each time point's images carry their own time all the way through. Then:

  • WinJUPOS opens each image at its own time, and the .ims measurement files it saves beside them are used by Combine and Interpolate.
  • Combine and Interpolate can use the decon images of a time series as key images.

Where the time comes from. The capture's own frame time stamps, which FireCapture and SharpCap write into every SER file (also in .ser.mj2 files compressed by BiggSky). If a SER has no frame time stamps, its UTC start time and frame rate are used. For captures with no time (AVI, image sequences), the stacks keep the capture's name, as before.

Your observer name. Choose Functions › Stacking › Set WinJUPOS Observer… and enter your name or initials as you want them in WinJUPOS file names, for example DB. Letters and digits only: hyphens and underscores separate the parts of a WinJUPOS name, so they are removed. The setting is saved with your BiggSky settings, and BiggSkyCLI uses it too with --settings user or --config. Leave it blank to clear it.

Observer and image info. A capture already named the WinJUPOS way, as FireCapture names them (2026-01-17-0524_4-U-R-Jup), keeps its observer and image info, with the time replaced by the clip's. FireCapture writes U when it does not know the observer; that U is replaced by your observer name when you have set one (2026-01-17-0524_4-DB-R-Jup). Any other observer in the capture's name is kept.

Any other name, such as SharpCap's 22_03_09, takes your observer name, else the observer recorded in the SER file, else U (SharpCap records the placeholder "Observer", which counts as none). The image info is the object, from the capture program's folder or names, followed by the capture's own name:

2025-10-10-0503_3-DB-Sat-22_03_09_Stack2_T01_F0300.tif

The WinjuposFormat stacking parameter turns this on (1, the default) or off (0: every time point's stack is named after the capture, as in earlier versions).

Derotate Stacks

Derotate Stacks turns a series of stacks to a common time with a model of the rotating planet and averages each with its neighbors. The result has the signal of the whole series and the sharpness of one short stack. The derotated stacks are added to the project as a new dataset. Each one has its own noise model and noise map, so deconvolution treats it exactly as it treats a normal stack.

Derotate Stacks is designed for Jupiter. For Saturn, Mars, or images through separate filters, use Combine Decons to Image.

Two ways to run it

When you stack. In the Stacking Options panel on the Stack tab, for a Planetary target:

ControlWhat it does
Clip (sec)The length of each clip, in seconds (the MaxClipTimeSec stacking parameter). 0 stacks the whole video as one.
Multi TimeStack every clip the video holds, not just the first (MultiTimePoint). Derotate and Avg # are available only when this is checked.
DerotateAfter the last clip is stacked, derotate the series into a new dataset (Derotate).
Avg #How many consecutive clips each derotated frame averages (DerotateAvg). See below.

Check Multi Time and Derotate, set Clip (sec), and press Stack Frames. BiggSky stacks each clip as usual into a time-point dataset (<dataset>_Stack_..._TP<n>), then derotates it. If you stack at more than one super-resolution scale, each scale's series is derotated separately. If the derotation fails, the stacks are kept and the reason is written to the log.

Afterwards. Select a dataset of stacks and click Derotate Stacks… in the Dewarp/Derotate panel (Stack tab › Functions), or choose Functions › Stacking › Derotate Stacks…. Use this to try other settings without stacking again, or for stacks made by another program. The dataset must hold at least two stacks, either BiggSky's own time-point stacks (..._T01_F####, ..._T02_F#### …) or stacks named with their capture time (see below); if it does not, a message says so. Stacks you have excluded with Select are left out.

Stacks from AutoStakkert! and other programs. Derotate Stacks also accepts a dataset of stacks made by another program, as long as each name starts with its capture time the way WinJUPOS names images, yyyy-mm-dd-hhmm_t... (for example 2026-03-03-1805_7-U-L-Jup_lapl6_ap82.tif). AutoStakkert! keeps this time when the capture file was named that way. Add the stacks to the project as a dataset, select it, and click Derotate Stacks…. The differences from BiggSky's own stacks:

  • The times come from the names, so the dialog does not ask for a clip length, and the stacks can be unevenly spaced and in any order. Two stacks with the same time (for example a drizzled copy of the same capture) are refused; keep one.
  • Every stack counts the same. Other programs do not save BiggSky's noise model, so the stacks cannot be weighted by their noise, and the derotated frames have no noise model of their own. Deconvolution estimates their noise as it does for any image added from outside BiggSky.
  • The rotation rate. If the rate measured from the stacks is more than 3% from Jupiter's System II rate, System II is used, with the measured direction. A moon shadow on the disk can pull the measurement off; the direction it gives is still reliable.
  • One filter of Jupiter. The stacks must all be one size, of Jupiter, in color or through one filter. A dataset of stacks through several filters (R, G, B … read from the names), or of Saturn or Mars, is refused with a message pointing to Combine Decons to Image.

The dialog

FieldWhat it does
Clip length the series was stacked with (s)Must match the clips the stacks were made from: BiggSky uses it to give each stack its time. It starts at the value the dataset recorded when it was stacked. Not asked for stacks named with their capture time: the title line then gives the span of their times instead.
Half-window on the inner disk (min)How far in time, either side of each output, stacks are turned and averaged near the center of the disk. Default 5 (DerotateWindowMin).
Or the N closest stacksGreater than 0: instead of the window, each output uses the N stacks closest to it in time (DerotateNClosest), and there is one output per stack. 0 uses the window.
Avg #How many consecutive stacks each output averages. 0 is automatic. See below.

The window, N closest and Avg # fields start from the current stacking parameters (the Avg # field on the Stack tab, and DerotateWindowMin and DerotateNClosest in the table).

How many frames you get

Each output frame averages a run of consecutive stacks and is turned to the time at the center of that run. Avg # sets how many stacks are in each run:

Avg #ResultDataset suffix
0 (automatic)Enough stacks to cover the half-window either side: 11 for 60 s clips with the 5-minute window, 21 for 30 s clips. Never more than the series holds.DRA&lt;M&gt;, or DRC if it covers the whole series
M, fewer than the stacksA moving average: N stacks give N − M + 1 frames, each M stacks wide.DRA&lt;M&gt;
The number of stacks or moreOne frame, at the center of the series, from every stack.DRC
(N closest stacks above 0)One frame per stack, each from the N stacks nearest it.DRN&lt;N&gt;

So with the automatic setting, a long session of 30 one-minute clips gives a moving average of 20 frames (DRA11), and one 3-minute capture split into three clips gives a single center frame (DRC). The stacking parameter DerotateSingle (Show advanced) set to 1 always gives the one center frame, whatever Avg # says.

What it does

  • The planet's shape and the rotation are measured from your stacks. The disk's outline (center, equator, flattening) is fitted to the mean of all the stacks. The rotation rate, including its direction, is measured from how much each stack has turned against the others, so a mirrored or flipped image is handled. If the stacks span less than 2 minutes there is too little rotation to measure the rate, and Jupiter's System II rate is used with the measured direction.
  • The window narrows toward the limb. Near the center of the disk the clouds stay recognizable for ±14 minutes after derotation; near the limb new longitudes rotate into view within a couple of minutes. So the half-window you set applies to the inner disk and shrinks to 1.5 minutes at the limb.
  • Each point is weighted by how squarely each stack saw it, and points that have turned onto the far side in a stack are left out. Brightness is corrected for limb darkening.
  • Stacks are weighted by their noise. A stack with more frames or less noise counts for more.
  • Off the disk nothing is turned. The sky and the limb glow are averaged as they are. Moons off the disk are taken from the output's own time point only, so they do not smear.
  • Each output gets its own noise model, made from the stacks' noise in exactly the way the output itself was made.

When it finishes, a message gives the number of stacks, the rotation rate, and the new dataset's name.

Where the files go

BiggSky's time-point stacks are derotated into the Stack_16b folder they came from, with their noise models in NoiseMap, named as a time series with the suffix added:

<prefix>_Stack2_[<series>_]<suffix>_T<nn>_F<frames>[_SuperRes<nn>].tif

When the stacks are named for WinJUPOS, the prefix takes each output's own time, so a DRC frame from three 20 s clips is named at the middle of the minute. The new dataset is named <dataset>_<suffix>, for example Frames_Stack_TP30_DRA11. Running Derotate Stacks again on the same dataset with the same settings replaces that dataset's frames.

Stacks named with their capture time (AutoStakkert!) are derotated into a subfolder of their own folder, named after the suffix. Each output keeps the name of the stack at the center of its run, with the time replaced by the output's own:

<stacks' folder>\DRA11\2026-03-03-1805_3-U-L-Jup_lapl6_ap82_DRA11.tif

Tips

  • How long to make the clips. For Jupiter, about 60 seconds. Splitting a capture of a minute or less into shorter clips loses detail: the dewarp already absorbs the rotation within about a minute, and shorter clips just have fewer frames each.
  • Long sessions. Capture continuously and use Avg # 0 for a moving average. Deconvolve the frames, then make the movie with Interpolate Decons to Movie (choose the View images).
  • "Fitted rate is not a gas giant's" in the log means the measured rotation is far from 0.6° per minute. The usual cause is a clip length in the dialog that does not match the stacks.

Note: a moon or moon shadow crossing the disk is not tracked: it is derotated with the clouds and comes out smeared along its path. Moons off the disk are handled. When a shadow is on the disk, deconvolve the stacks and use Combine Decons to Image, which tracks moons and shadows.

Combine Decons to Image and Interpolate Decons to Movie

Both work on finished images taken minutes apart, normally your deconvolved results:

  • Combine Decons to Image turns every image to one time and combines them: more signal in one image, with moons and moon shadows where they were at that time. For a filter set (R, G, B …), each filter is combined and the filters are composed into color.
  • Interpolate Decons to Movie makes frames at even times between the images, for example one a minute from images taken every 20 minutes, and an MP4 movie. Each frame is made by turning the images either side of it to its time and blending them, so the planet turns smoothly.

They also handle Saturn (the globe and the rings, each at its own rate), Jupiter in mono through filters, and Mars.

The images

The images must be named with their capture time in the form WinJUPOS uses:

yyyy-mm-dd-hhmm_t...      for example 2026-01-17-0524_4-U-R-Jup_lapl3_ap120_Decon_View_16b.tif

where t is tenths of a minute, in UT. Captures from FireCapture are named this way, and so are their AutoStakkert! stacks, and BiggSky's own time-point stacks when WinjuposFormat is on. Decon keeps the stack's name, so the decon images carry the time too.

Select the dataset of stacks and click Combine Decons to Image… or Interpolate Decons to Movie… (Stack tab › Functions › Dewarp/Derotate, or the same items under Functions › Stacking). If the stacks have been deconvolved, a list asks which images to use:

ChoiceWhat it is
Decon_View_16b (and any other View folder)The deconvolved images as BiggSky shows them: stretch, gamma and color already applied, and what WinJUPOS measures. The default, and usually the right choice for a final image or movie.
Decon_32bThe deconvolved images before display settings. Use these to apply your own rendering to the result afterwards.
StacksThe stacks themselves, not deconvolved.

The list shows how many of the dataset's frames each folder has an image for; frames without one are left out. Frames you have excluded with Select are left out too. If the selected dataset's frames carry no time, a folder dialog asks for a folder of images instead (PNG and TIFF).

For View images, render every image of the set with the same settings. Jupiter's images are matched in brightness to each other, so small differences do not flicker, but different sharpening does show.

Jupiter color images must all be the same size. For Saturn, filter sets and Mars, images of different sizes are centered on one canvas, and when the same time appears more than once (for example a super-resolution copy), the copy nearest the others' size is used.

WinJUPOS measurements. If you have measured any of the images in WinJUPOS, BiggSky uses the measurement (.ims.xml or .ims), found automatically beside the images, in folders below them, or in folders next to theirs. One measurement is enough: it gives the absolute time, the outline and the tilt of the planet's axis, and the rotation carries these to the other images. For Mars at least one .ims.xml measurement is required.

WinJUPOS saves its measurements in the folder it last used, which may belong to another project. If yours are not beside the images, enter their folder in the dialog's WinJUPOS .ims folder field. When the same capture time appears more than once (a stack and its super-resolution copy), the measurement whose outline fits the image is used.

Which dialog you get

BiggSky decides which planet the images show from a WinJUPOS measurement, or else from the words in the file names (Jup or Jupiter, Sat or Saturn, Mars):

  • Jupiter in color, or through one filter, gets the Jupiter dialog, with moon tracking.
  • Saturn, Mars, and mono Jupiter through two or more filters (R, G, B, IR, L, CH4, UV, Ha …: the filter is read from the file name) get the filter dialog. Each filter is derotated on its own and the results are composed into color.

The first field of both dialogs is Planet. It says auto with the planet it found and how, or is blank when nothing says, in which case you must enter Jupiter, Saturn, Mars or none. If you choose a planet that needs the other dialog, the other dialog opens.

Combine Decons to Image: the dialogs

Jupiter:

FieldWhat it does
Planetauto, Jupiter, Saturn, Mars, or none.
Time, UTThe time of the combined image (hh:mm or hh:mm.m). Blank is the middle of the images' span, shown in the prompt.
Track moons and shadows1: moons and moon shadows are found in each image, cut out, and pasted back where they are at the combined image's time. 0: everything is derotated with the clouds.
Also moons in transit on the disk1: also look for moons crossing the disk. These are kept only when the same moon is also seen clearly off the disk in another image.
Time the images by1: by the moons, or by the rotation when no moons are found. 2: by the WinJUPOS measurements, each measured image taking the time WinJUPOS gives it.
Color BoostColor gain on the color images, as the Decon panel's Color Boost does. 1 is none. It starts at the Decon panel's value when its Color Boost box is checked, else 1.
SuffixAdded to the output folder, files and dataset name, for example _DRC_test (see Where the files go). Letters, digits, - and _. Blank is none.
WinJUPOS .ims folderWhere WinJUPOS saved its measurements, when that is not beside the images. Blank looks beside the images only.

Filter sets (Saturn, Mars, mono Jupiter): the header lists the filters with how many images each has, and the span of capture times.

FieldWhat it does
Planetauto, Jupiter, Saturn, Mars, or none. none turns nothing: each filter's images are registered and averaged.
Time(s), UTBlank is the middle of the span. Give a time for one image at that time (for example with a storm on the central meridian), or several times for one image at each.
Luminance for LRGBThe filter used as luminance for an LRGB image: auto, none, or one of the filters listed. auto takes L, else IR (any IR filter, such as IR685), else Ha, but only when that filter has an image within the time window of every output time. Otherwise there is no LRGB and the closing message says why; choose the filter by name to use it anyway.
Time window (minutes)Images within this time of the output are used, weighted less the further they are from it, falling to zero at the window. Default 25. A filter with no image that close uses its nearest image, so no output loses a filter; a window of 1 uses each filter's nearest capture.
Weight each image by its quality1: each image is weighted by its fine detail against the others of the same filter, so a soft image counts for less. Filters are never compared with each other.
Color Boost, Suffix, WinJUPOS .ims folderAs in the Jupiter dialog. Color Boost applies to the color images (RGB and LRGB), not to the single filters.

Interpolate Decons to Movie: the dialogs

Jupiter: Planet, then Frame spacing (minutes) (default 1), Track moons and shadows, Also moons in transit on the disk, Video frame rate (frames per second, default 10; 0 makes no movie), Time the images by, Color Boost, Suffix and WinJUPOS .ims folder, as for Combine. Color Boost applies to the frames and the movie.

Filter sets: Planet, Frame spacing (minutes) (default 1), Frames from-to, UT (hh:mm-hh:mm; blank is the whole span; a span across midnight is allowed), Luminance for LRGB, Time window, quality weighting, Video frame rate (default 8; 0 makes no movie), Color Boost, Suffix and WinJUPOS .ims folder.

The filter-set answers, apart from Planet and Suffix, are kept until BiggSky is closed, separately for Combine and for Interpolate. The WinJUPOS folder is remembered for both dialogs until BiggSky is closed; Suffix starts blank each time.

What they do

Jupiter (color).

  • The images are timed by their rotation, not their labels. The rotation between neighboring images is measured, and each is given the time its rotation says. An image whose label disagrees by more than a minute is reported in the closing message. When Jupiter's moons are found, they are matched to their predicted positions, and their positions set the images' times.
  • Interpolate: each frame blends the two images either side of it, turned to its time and weighted by how squarely each saw each point, so the limb rotating into view comes from the later image and the one rotating away from the earlier. Other programs make in-between frames by cross-fading the two images, which covers only part of each gap and then jumps at every image.
  • Combine: every image is turned to the chosen time and averaged, each point from the images that saw it best.
  • Moons and shadows are cut out and pasted back where they are at the output's time.

Saturn. Each filter's images are turned to the output time: the globe at Saturn's System III rate, and each part of the rings at its own orbital speed, so the spokes in the rings stay sharp. Only what differs from the median of the images is turned, so the shadows and the rings' fixed structure stay exactly where they are. Without a WinJUPOS measurement, the outline and tilt are fitted to the rings.

Jupiter through filters. As Saturn, but the globe only, with no rings and no median. The moons are not tracked.

Mars. As Jupiter through filters, at Mars' rotation rate. A WinJUPOS .ims.xml measurement of at least one image is needed.

For filter sets (Saturn, Mars, mono Jupiter, and none):

  • Field rotation from an alt-az mount is measured and removed when it is significant: the axis of each image is measured (Jupiter's bands, Saturn's rings), or taken from two or more WinJUPOS measurements, and each image is turned to match before the images are combined. Mars and none need the WinJUPOS measurements for this.
  • Color products (RGB, and LRGB such as IRRGB) are white balanced as BiggSky's planetary auto white balance does, with one balance for the whole set so a movie does not flicker, and then given the Color Boost.

Jupiter in color keeps the images' own color: it is not white balanced again, only given the Color Boost. Field rotation is not removed on this path: images from an alt-az mount turn slightly from one to the next, which softens the limb in a combined image.

Progress is shown in the main window's progress bar, and the main Cancel button stops the run. When it finishes, a message lists what was made, the times, the rotation rate, the moons tracked, any images whose labels disagree with their rotation, and where the files went.

Where the files go

Everything is written to a subfolder of the images' folder (for decon images, inside the decon folder). The folder, the files and the new dataset share one name, made from the set of images:

<set>_DRC[_<Suffix>]             Combine
<set>_DRI<spacing>min[_<Suffix>] Interpolate (for example ..._DRI1min)

<set> is the dataset's name and, for decon images, the folder they came from, for example Saturn_Decon_View_16b; for the stacks themselves it is the dataset's name alone. With a folder of images it is the folder's name, plus the SuperRes tag (such as SuperRes15) when every image carries the same one. So a decon and its super-resolution copies, which share a decon folder, no longer write over each other. Use Suffix to keep runs with different settings apart.

The output images are named with their own times, the way WinJUPOS names images, so they can be measured in WinJUPOS:

yyyy-mm-dd-hhmm_t_<Prefix>_<set>_DRC[_<Suffix>].png             a Jupiter combined image
yyyy-mm-dd-hhmm_t_<Prefix>_<set>_DRI[_<Suffix>].png             a Jupiter movie frame
yyyy-mm-dd-hhmm_t_<Prefix>_<product>_<set>_DRC[_<Suffix>].png   a filter-set image: R, G, B, IR ..., RGB, IRRGB ...
yyyy-mm-dd_<Prefix>[_<product>]_<set>_DRI[_<Suffix>].mp4        the movie

The prefix is the word after the time in the image names (for example U-L-Jup), or Sat / Jup / Mar, or Avg for none. 16-bit images (View) give 16-bit PNGs; 32-bit images (Decon_32b) give 32-bit TIFFs. Combine writes every product of a filter set; Interpolate writes only the main one (LRGB if there is one, else RGB, else the last filter). Movies are written at twice the image size, which keeps the fine detail that H.264's color compression would otherwise lose, using ffmpeg if it is installed.

The images are added to the project as a new dataset with the same name, for example Saturn_Decon_View_16b_DRC or Saturn_Decon_View_16b_DRI1min_test, and selected, so they show in the View right away. For a filter set, the dataset holds the color products. Its display white balance is off, because the images are balanced already.

Tips

  • Image spacing. Images 15 to 25 minutes apart work well for Jupiter movies. Farther apart, less of the disk is seen in both images of a pair.
  • Use the View images for the final picture or movie, and Decon_32b when you want to render the result yourself afterwards.
  • Check the closing message. An image reported as disagreeing with its label by several minutes usually has a wrong label; the output uses the time its rotation gives.

Note: a moon crossing the disk that never leaves it during the sequence is not found, and is blended with the clouds. Moons are tracked only for Jupiter color images.

Derotation needs no WinJUPOS measurement: the key images are timed by their own rotation and by Jupiter’s moons. See also the Dewarp Movie manual, the Stacking parameters and the full BiggSky manual.