Field rotation in alt-az

[opticalpath]

Hi. It's been several years since I did any planetary/ lunar imaging and I'd like to have a go again, but this time it would be on an alt-az. mounted SCT and not my equatorially mounted scope (which is set up strictly for deep sky now).

I haven't tried it before in alt-az, so I thought I'd come here and ask if it's practical, given the inevitable field-rotation in this tracking mode:

- Does any of the popular stacking software for lunar/ planetary imaging correct for rotational mis-alignment?

- In practice, how long a video clip can be recorded in alt-az before field rotation becomes a visible problem?

I realise the answer to the second question depends on a lot of factors (I've seen the maths ) but any practical guidance on this would be really appreciated.

Thanks,

Adrian

[Pete Presland]

not sure to be honest, but i will be fun finding out

[JamesF]

If you look in the planetary imaging section there are some excellent results from cameras in GOTO dobs, so clearly it's possible. I have always understood that the stacking applications handled rotational alignment, but I image from an EQ mount so it's never been an issue for me. One thing you can do is put the mount on a wedge to align the az axis with the Earth's axis, which effectively makes the field rotation problem go away.

James

[Space Cowboy]

I've never had a problem with it on my Dob but its fascinating to watch the field rotation on an animation but the stacking software seems to cope ok even with long avis i.e up to 6 mins.

WINJUPOS has a facility in its de-rotation application for field rotation correction if you want to make sure everything is good and proper but to be honest I can't see much difference.

1 hr 34 mins of field rotation:

3 mins of rotation :

3 min stacked image:

[opticalpath]

Thanks Stuart. Looks like 2-3 mins. would be the limit anyway before rotation of the planet changes the view (for Jupiter anyway). On your 3 min clip you can also just see a little field rotation, but WINJUPOS seems to have done a good job correcting that in the alignment. I wonder if AutoStakkert and Registax do the same. I would just go out and try it but for the weather!

Adrian

[Space Cowboy]

The stacked image I've posted was not de-rotated in Winjupos , thats just a straight stack using Autostakkert. As you say even in a 3min avi there is some field rotation but the stacking can cope with this small level of rotation it seems.

[opticalpath]

The stacked image I've posted was not de-rotated in Winjupos , thats just a straight stack using Autostakkert. As you say even in a 3min avi there is some field rotation but the stacking can cope with this small level of rotation it seems.

That's interesting, Stuart. I just got an answer to the question I posted on the AutoStakkert Yahoo! group .... According to the author, the program doesn't do rotation correction! So that's a bit baffling. Can't deny the good results you got with it anyway and that's what matters.

Adrian

[Space Cowboy]

In my experience the stacking and aligning of both Autostakkert and Registax does correct a small amount of rotation as I demonstrated in this thread :

http://stargazerslounge.com/topic/169898-jupiter-dec-1st-de-rotate-comparrison-update/page__hl__%2Bde%26%238208rotation

[opticalpath]

That's fascinating, Stuart, and your results speak for themselves. Emil is adamant, though, that he has not written rotational correction into AutoStakkert ....... because it would be highly processor intensive and would slow the program down significantly.

I wonder if what's happening is this: Over a short period, the small rotational error would show itself as tiny 'local' angular change in features plus an x-y displacement of features that would get worse the further you are from the centre of rotation. AutoStakkert is doing a good job of removing the larger effect - the x-y displacements - by its multi-point (linear) realignments. The residual tiny angular errors remain, but are now multiple small errors, each one small and localized to a (correctly centered) MAP box.

So my theory is that we can't really see the tiny change of *angle* as such in the features of a rotated image; what we detect is the consequence of the angle, i.e. the x-y displacements of features ... especially as they get bigger further away from the centre of rotation. When you take away all the gross x-y errors by correctly repositioning each MAP box (via linear alignment), you have removed the most visible effect of the rotation and all that's left is the angular error in each box. We don't notice these errors because the radius of rotation in each box is tiny and the resulting local x-y displacement errors are negligible.

Anyway, just a theory! It would be interesting to try this on a series of larger scale images - say the Moon imaged with a large format CCD camera - to see if it still works this way when the image is spread over many more pixels.

Adrian

[Space Cowboy]

Great theory Adrian! You clearly know a lot more about the mechanics of this software than I do. All I go by is my images and I've never noticed a real problem caused by my alt/az tracking. I know some people say the limbs of Jupiter tend to be softer with az tracking but I can't see any difference.

I've been meaning to post my rotation results on the CN forum where Emil is a member.

There is a guy in the far east who uses the same Dob as mine and his images are terrific :

http://www.cloudynig...t/1#Post5630230

[opticalpath]

There is a guy in the far east who uses the same Dob as mine and his images are terrific :

http://www.cloudynig...t/1#Post5630230

gulp! 'Nuff said.

Adrian

[opticalpath]

Emil is adamant, though, that he has not written rotational correction into AutoStakkert .......

I spoke too soon. Emil just confirmed on the Y! group that he is looking at implementing an optional field-rotation correction feature in AS!2.

Adrian

[Space Cowboy]

Excellent!

[stev74]

Great theory Adrian! You clearly know a lot more about the mechanics of this software than I do. All I go by is my images and I've never noticed a real problem caused by my alt/az tracking. I know some people say the limbs of Jupiter tend to be softer with az tracking but I can't see any difference.

I've been meaning to post my rotation results on the CN forum where Emil is a member.

There is a guy in the far east who uses the same Dob as mine and his images are terrific :

http://www.cloudynig...t/1#Post5630230

That image is simple stunning!! :eek: I wish we could get anywhere close to his weather and seeing conditions in the UK!

As for field rotation, I do notice it over a couple of hours imaging as in - each image I stack the main bands will be slightly further up on the right and down on the left but I'm not sure it is moving that much over over a 3 min AVI. I do now tick the 'Compensation of field rotation in Alt/Az mounts' box on Winjupos when I de-rotated the AVI just in case though.