Hang on a minute! stacking question

[GasGiant]

To get sharp images of say Jupiter and the sun the photographer will take loads of images and stack them, & this process can take a while. Am I correct ?

So if something is moving, such as the atmosphere on jupiter how can the stacking software align the best images, would they not be off-set ?  Ive seen images of the sun with defined flares etc, are Sun images stacked or..... ? 

Please enlighten me 😬

 

[Mal22]

It is deep space imaging where you would take lots of exposures, hours and hours worth, and stack them together, which certainly does take a long time.

For planets, the wobbly atmosphere and rotation is more of a challenge.

I’ve not used a DSLR so can’t comment in relation to that process. I use a dedicated astronomy camera to take a short video, maybe 3 to 5 mins depending on the speed a particular planet spins. That video will consist of thousands of frames, so we then use software to select a percentage of the frames where the atmosphere was least wobbly, this might only be 20%, but could be less or more.

The software then stacks those selected frames together to produce a clearer image, but it is a far quicker process than stacking deep space object images, just a few minutes. 

Edited January 1 by Mal22

[michael8554]

There comes a point where the usual Stacking software "runs out of range" so to speak, due to planetary rotation.

The freeware program WinJupos "derotates", correcting for planetary rotation,  allowing you to extend your imaging time by a factor of three or four.

Michael

Edited January 1 by michael8554

[vlaiv]

It all comes down to timescale of particular changes.

In planetary imaging exposures are set so that our atmosphere does not move in this time frame (our atmosphere is "wobbly" and distorts image). Total exposure time is limited to avoid artifact from change on particular target. Some targets change over long periods of time and some on shorter scales.

Stacking software is able to compensate some level of change, but for other things there is dedicated software.

For example - stacking software is perfectly capable of compensating for change in apparent size of the moon during the course of the evening. Moon changes in size because earth rotates and observer is sometimes closer and sometimes further away from the moon (earth's radius is ~6300 Km and distance to the moon is ~384400Km so that is up to 1.6% change in apparent size between moon rise and it being at zenith).

Stacking software is capable of compensating for Jupiter's rotation up to few minutes, but longer than that and you need to use dedicated software.

In any case - exposure duration is order of milliseconds and total recording often ends up being several dozen of thousand of frames - so up to few minutes of recording. If change is longer than this time frame - there is no issue, but if it's shorter - well, shorter imaging run must be used.

[ollypenrice]

Planetary image stacking and deep sky stacking work slightly differently.

In deep sky, the stacking exploits the fact that noise remaining after calibration is random whereas signal is constant, so signal is reinforced and noise cancels itself out. More signal and less noise is the result.

This also applies in planetary imaging but, in planetary, a proportion of the very short subs will have coincided with moments of good, stable seeing. The imager identifies such a sub and asks the software to find and stack other ones like it, ignoring the rest.

In both cases the software will identify key features and align all the subs based on these. In deep sky this will be the stars and, in planetary, it will be features on the disk.

The algorithms used for stacking don't necessarily use simple averaging. They can also be asked to reject outlying pixel values, such as those created by passing satellites, these to be replaced by the average of the other subs.

In other activities we can all obtain more precise measurements by making not one measurement, but several, and averaging the result.

Olly

[GasGiant]

Thanks guys. Very interesting. Lots to learn ! 

So if i take a video of the moon say, software assess each frame its taken and keeps the best. 

 

If I took a one minute video of the moon with NO tracking would software auto juggle the best frames and stack them. Or, as I suspect, it would not as the moon would drift too far away

Edited January 1 by GasGiant

[vlaiv]

4 minutes ago, GasGiant said:

If I took a one minute video of the moon with NO tracking would software auto juggle the best frames and stack them. Or, as I suspect, it would not as the moon would drift too far away

I've seen people do incredible things with manual tracking dobs for example - so it is possible.

Moon is actually very good target to try to capture as it is big and drifts thru FOV - which means that you always capture some part of it. Then you can do what most people do and create lunar mosaic - from each movie of the moon focus on one part of its surface and stack that (there are different ways to "stabilize" the recording and surface features is one of those methods - and I think it will work well for this use case).

Then you can take the same movie and focus on different part of lunar surface and stack that.

In the end you take different panes and stitch those into mosaic of lunar surface.

Doing good lunar image without tracking is not going to be the easiest thing - but I think it is well possible.

[GasGiant]

Thank you @vlaiv