Can´t register pics

Hi Jorge,

Looking at the screenshot you have posted I think you have two problems:

a) APP has only managed to identified only a relatively small number of stars - the highest count I see is c250 - whereas Mabula frequently suggests a target of c500 stars

b) If you look at the column FWHM min max, the detected stars are quite elongated (c2.5x min to max)  which presumably does not help APP to accurately calculate the centroid position of the stars.

I assume also that your frames do largely cover the same target area (i.e. you are not purposely making a mosaic) as this would further reduce the area of overlap between frames.

If you have not already done so I can only suggest you try:

a) playing with the target stars slider on tab 3) Analyse stars to see if you can increase the star detection rate

b) on tab 4) Register stars, switch the pattern recognition parameter to Triangles and increase the value of scale stop to see if you can improve the registration rate.

If you cannot improve matters further then I guess you can still work through the APP process with what you have managed to register and hope for a reasonably decent result at the end.  I also suggest you look carefully at your imaging rig to see if you can increase stability and your polar alignment if applicable so that your stars do not become so elongated. If you are shooting unguided then I think your 20s exposure time itself should not be an issue unless you are shooting with a very long focal length lens.

Hoping you can achieve some improvement.

Mike

Hello Mike! Many thanks for the detailed explanation!

a) Since I´m not in a very polluted area bortle class 4, it´s not that bad for finding stars I think

b) Pictures were taken with a 20mm lens and 20s each, Yes, each picture is covering the same area, I´m not trying to make a mosaic.

on the actions you recommend:

a), you mean, decreasing target stars number with the slider?

Even with this errors I get, you mean the APP gives me a stacked image with the ones that are succesfuly registered? Where is the stacked file at?

I´m not using a startracked, so if I want to decrease elongation of the stars, I should reduce the exposure time (increase ISO settings)

On starry landscape stacker I can stack images without errors, could it help if I come to APP with the stacked result that gives me Starry Landscape Stacker?

Thanks!

Hi Jorge,

I'm guessing that you are not very familiar with APP as yet.

I suggest first that you take a good look at one of your individual images using settings as in the below screenshot.  In particular, left click the image a few times so it is nicely magnified and select eg 30%BG, 2 sigma, o.o% base so you can easily take a good look at your data - I think you will clearly see that your stars are elongated or mishapen - not nice and round.

You say you were using a 20mm lens at 20sec exposure.  I think it very like that there was significant movement of your camera during  exposure.  I suggest you carefully review how you had your camera mounted at the time of exposure and see what you can do to improve stability.  If your camera is firmly mounted I think you should be able to get nice round stars with that lens combination / exposure time even without tracking.

You should try to increase the target number of stars with the slider to see if this helps.

Even if you cannot get better Star Registration results then you can still follow through the remaining Normalisation and Integration steps to get a stacked image with the frames that have been registered.  You do not yet have a staced image as yet if you have not worked through to Integration.   At the end of Integration your stacked image will be saved in the WIP directory that you have selected.  The file will also be listed at the foot of your image list.

I am not familiar with Starry Landscape Stacker but maybe it is better optimised for the type of data you currently have.  As previously I think your problem is that your stars are not very round.  I'm not sure how you might benefit from importing a stacked image from Starry Landscape to APP. I will take a look at the SL tutorials

In my opinion APP is worth the effort of getting to grips with it.

Wishing you success

Mike

Hi again Mike. First many thanks for helping me!

I contacted Mabula before dowloading his APP telling him what my goal was using it. He told me to download trial version and give it a try.

Basically my final goal is to subtract nasty light pollution seen on milky way on my comp with various landscapes from different pics.

Mabula told me that he have not been successful with images containning more than 25% of it being ground. So I cropped the image to only left less than 10% of ground. I edit separately ground and sky always, so does not mind. He liked the idea and told to give a try and report back if I could get something good out of it.

I´m not any familiar with APP as you can see... I´ve noticed what you say, stars are a bit elongated, I think reason is exposure time, my tripod is firmly set up, and I use an intervalometer for not touching camera at all on the process. If I look on photopills app, to have stars as points, with NPF rule I wouldn´t have to go further than 13". 500 rule is 25", so I could have a problem on that side.

You´re totally agree, APP worth the effort of getting to grips with it, so I have to keep on trying.

Well, I finally got a registered image. Worked to reduce light pollution, calibrate background, and calibrate star colors, but I get weird shadows on the horizon line after applying light pollution removal?. Why am I getting this? How can I avoid it?

Hi,

Your image is looking very promising.  I think that what you are seeing is a residual effect from your Remove Light Pollution.

I've not tried an image like this but I think you will probably need to put a largish number of very small selection boxes evenly all along,  and on both sides of  the transistion line between ground and sky until you get an even skyline from side to side.  In my experience the LPC result can change markedly depending on where the selection boxes are placed.

It is possible to open up an image that has already undergone LP correction so you do not have to start again all over.

Good luck

Mike

Thanks Mike, I see the image I got very plain, without color. Is it ok this? I performed star color correction after LP too...

Hi

I'm not quite clear what your question is and I can see only the screen shot that you posted earlier so you may already have tried some of these things:

As above I think you need to refine your LP correction - you can use either the FITS(?) file you originally used as input to LPC or refine any of the subsequent FITS you have produced.  Because of the hard transistion from dark forground to brighter sky I think this will be tricky and need some persistence.

I notice in the screenshot that you have not set the Saturation checkbox on the RH panel.   When you have this checked you will be able to adjust the SAT and SAT TH sliders. 

I do not know the result that you have in mind but if you are happy to have the foreground trees as silhouettes I think I would make the sky a little darker and then increase the sky colours.   If you have any colour in the foreground then I guess you might try to bring that out a little more as well.

For my taste I think I would crop the picture a little on the LHS.  For me the key areas of interest are the Milky Way, Saturn (?) and the silhouetted trees.

Whatever you decide I looking forward to seeing your result.  And how does it compare with your Starry Landscape result?

Mike

@mestutters

Mike, as soon as I get to the computer again I´ll attach more pics of what I was getting on APP after light pollution removal. Besides the ground transition, I didn´t see that kept star colours, I saw a very plain image.

I decided to go to my classic way, lightroom, starry landscape stacker and photoshop for stretching and gradient removal, since my old iMac was almost all the friday afternoon, analizying stars, registering, and all that stuff, with my trial&errors.

Starry landscape stacker gives me at this moment a much better consistent final result, with minor effort for learning it. I know its much simpler and limited program, but I don't even know which way to go with APP for getting some consistent results.

When placing boxes for removing LP which criteria for placing them? I saw video tutorial from Mabula, he explains to try to address small parts of the image  and placing few boxes first, but don't know why he places where it places it, and what's the technical reason. Where should they be for success?

I tried long, placing many boxes, placing few, placing on all the image, placing only by parts like the tutorial (worst result) but didn´t get any decent result at all. Much time consuming this process without success for my side.

Didn´t try thought to place them on foreground.

I don't really care about the foreground, since foreground pics are separated conveniently.

I´ll be back with some pics from both programs.

PD if somebody that knows the functionality of this program better, wants to make a try of removing from my registered file, I could send it without a problem.

@pachanga

Hi,

As I have no new data of my own to work on I am prepared to have a shot at processing it - no promises mind. 

Can you post a link to your file so I can download it?

It would be nice also to see your SL result, to see where you have set the bar.

Mike

Sorry for the delay, I was on a travel, not at home.

I´ll send you a PM with a link where you can download 3 pictures:

1 Astro pixel processor with the stack (integration done)

1 Astro pixel processor with light pollution removal done (myself)

1 Starry landscape stacker TIF image with mean stacking done by SLS

Thanks Jorge