Why won't DSS stack more than one frame?

[Kev11]

Presumably DSS in some sense measures transparency. Same object, same exposure on DSLR unmodified - one night registration score 18000 another night score 33 at which DSS says "only one frame will be stacked do you wish to continue? [you idiot!]"
But the data is there when I wind down the slider bar at top right of screen (contrast? luminance? adjustment).
How do I get DSS to detect that data? I have star detection threshold set at lowest i.e. 2%. All registration and stacking settings standard as far as I can see.
Any suggestions greatly appreciated.

[glend]

I have never ever changed any DSS settings from the standard. I have very rarely seen that message regarding only one frame to stack. Have you run the sub files through the Image Grader first, to assess the relatively quality?

https://www.sequencegeneratorpro.com/download/fits-image-grader/

BTW star detection threshold should not be that low, try 25% min. But it depend on the target and relative population of stars in the field of view. Your effectively selecting the number of stars to register against.

Have you read through the DSS technical guide:

http://deepskystacker.free.fr/english/technical.htm

[Startrek]

Kevin


DSS may not be recognising your stars due to the following possible reasons -



Stars are to out of focus, so saturated



Exposures are too short



For DSLR you should be stacking with Raw Cr files not Jpegs



**Have you loaded Lights , Darks, Bias etc into correct categories and checked all


Don't know if the above helps ?


Martin

[raymo]

When in the advanced tab start out with the star detection slider set at about 60 or 70%, [not 2%], compute, and if you get a huge number of stars detected [hundreds or thousands] go to a higher percentage and compute again, repeat until you have around 50-100 stars detected; that is all DSS
needs. The more stars detected the slower DSS will run, at 2% it will quite
possibly not even attempt to align and/or stack such vast numbers of stars,
and even if you have a premium computer it will run at glacial speed, like
many hours to stack say 50 subs, as opposed to minutes. Hope this helps.
raymo

[Kev11]

Thanks for the replies.
I probably made my question too complicated.
My problem is DSS will not recognise more than 3 stars on frames that show dozens when viewed on any normal photo viewer software.
Is there any adjustment I can set on DSS?

[Nikolas]

yes, the star detection threshold

slide it to the right or left and see if it makes a difference
it's in the settings

[xelasnave]

It's in "advanced" .. when you go to stack...I used it years ago and could not understand why recently I could not find it..it's not anywhere on the left...I notice mine is at 60 but I suggest you select say 8 light frames to stack , even five, and try it from 15 up to 70 and see if you can notice a difference...good luck.
Alex

[Kev11]

I have tried every available setting. Attached file are the one frame that DSS lets me stack - tried RAW and JPEG. Why can't I stack the other 39 similar ones??

[RyanJones]

Hi Kevin,

Zoom into your stars and you will see they are slightly greyed out in the centre. Your images appear out of focus, that's why DSS won't register them.

Cheers

[Kev11]

Maybe Ryan but this is my focus image with mask. DSS will register it just won't stack more than one frame presumably because it scores them all in single digits. It is something to do with the where it puts the lights on the luminance curve, which in turn is something to do with the atmospheric conditions I think.

[xelasnave]

Kevin what percentage do you have the star detection threshold set at?
Alex

[RyanJones]

Well that looks absolutely bang on. Is there any chance your focuser slipped after focusing ? I could be wrong but just going off the images you posted that is exactly what it looks like with doughnut stars.

You didn’t happen to leave the focuser in the same position from the previous night that worked ? That focus shot was from that particular session ?

[raymo]

I don't know if the mask is perfectly symmetrical, but the spike 6.30am
to 12.30pm has in its lower half a smaller gap to its right than to its left,
but in its upper half the gap difference is much less obvious, in fact barely
detectable visually. Whatever the reason, the stars are definitely out of focus,
but not enough IMHO to prevent DSS from doing its thing.
raymo

P.S. I just thought, the image is extremely noisy, I wonder if that could be enough to
confuse DSS, for example, the noise on the target itself looks like hundreds of pinprick stars,
and that pattern would be different on each sub; but maybe I'm overthinking it.

[glend]

Zooming in on that image (in post #8) shows that the stars are out of focus and exhibit coma tails. Did you grade those subs like I suggested? There are regions of the Image where the stars look like donuts. Have you looked closely at more of the subs? I would be closely eyeballing each one, and grading the stack.

[Kev11]

Alex, I set the slider at 2% the lowest available. It isn’t that; it is as if DSS is saying: “Nah, the image is too dark”. The stars are there but you have to wind up the contrast to see em. Interestingly, the faint outline of the galaxy (NGC55) is visible on the camera screen as well as the stars so my theory (built on near total ignorance of digital photography – I learned photography in the old silver halide days) is that DSS requires some minimum level of contrast between sky and stars to do its thing. But then my site does not have any light pollution as such. What it does have is a location with surrounding hills and the Murrumbidgee River about 5kms away: fogs are not all that common but otherwise clear nights can certainly sometimes look “murky”, my best astrophotography has been done on winter nights in drought.


I agree about the focus issues and I should probably do more masked test shots each night. Even with a 2X magnifier on the viewfinder my tired old eyes (even the good one) struggle to get sharp focus and sometimes even tightening of the focus lock seems to shift the focus slightly. Again, I wonder whether these bad seeing and transparency nights might create some distortion in 30s subs.


However, as Raymo says, out-of-focus doesn’t stop DSS stacking just gives you an out-of-focus stack (which is very annoying when it happens).


In summary I guess what I am asking is can I force DSS to stack light frames which have a score below its self-imposed minimum, or somehow pre-process the RAW images to push them up the luminance curve or is there some other software to try? I hate to throw away data from my relatively rare moonless and cloudless nights (proximity to the High Country gives us more cloud than further West). I apologise to city astrophotographers who have more difficulties for my whinging.

[xelasnave]

Hi Kevin
Please try this.
Take only five frames as this is a test and proceed to stack them but before you do go into the advanced setting on one of those squares that appear in the middle of your screen that has "advanced settings" go in there and change your star number settling to say 50 % and try that..irrespective of your various observations that 2% should work I hope you will find 50% does work and you will find all five frames stack...if they do try again using all your subs.
I have found that the lowest setting simply does not work..you must change it ...
Alex

[RyanJones]

Kevin,

Two things;

As frustrating as it is, if you have to throw out an entire night because something went wrong, sadly it won’t be your last. We’ve all done it. It isn’t a total loss because you can learn from it.

Second, begin with the end in mind ( steven covey ). If you stack out of focus subs, you won’t get a good image anyway.

There will always be other nights. The universe isn’t going anywhere :)

[raymo]

Kevin, I should slap myself on the forehead. At 2% DSS is set on its minimum
sensitivity, so no wonder it is not detecting enough stars to do any aligning or stacking. Please start at around 60 or 70% as I said before, and all should be well. As I said 50 to 100 stars is plenty.
raymo

[Wilsil]

What others have said: move the threshold up.
The question have been asked multiple times now: how many stars did DSS tell you if "sees" after moving the threshold?

[xelasnave]

I am stacking at the moment and unable to read what DSS says about the threshold but I do recall that it to me was misleading in that I tried a low setting as that seemed the best idea in view of the DSS notes however I found it really seems the reverse to how I read the notes...I know at the time I thought it seemed contrary to the DSS notes but I finally set it at 70%..I don't know why at the time I went with 70% but I have not changed it to this day and after I have never got the can only stack one frame message.

Looking at the image you posted I can tell you that I have stacked worse...I will bet you ten bucks that at 70% you will have all frames stacked...well no cause I can only take your money.

If I can I will post a photo of the box where it shows "advanced" and also a picture of my setting at 70%...at the moment I am unable as I am stacking 308 light frames of the Spider thru the 80mm with the Nikon.

Please just try it at 70%

Alex

[Sunfish]

My experience is that if you have the sensitivity percentage about right and DSS counts a few hundred stars it should stack 100% of the images if that is set. Raw frames should be stacked and set up correctly. The only circumstance that it will not do that is if the guiding has allowed the frame to move so much that there is too much out of frame or you have selected a bad frame as the reference frame. Make sure you select the frame with the most stars as the reference frame and open each frame and zoom in to check your object is really clear in there and not obscured by noise or cloud or movement.

[Sunfish]

In the register screen if you check advanced and change the percentage you can quickly compute the number of stars. With my OSC shots , increasing the percentage reduces the number of stars computed in DSS.

[xelasnave]

Hi Ray...you really have to go overboard for DSS not to work because of poor framing...these days I don't worry about getting the following nights framing anyways near perfect...and remember I don't use goto just eyeball objects from memory ..I get a half inch off or more on the lap top screen..for Sharpcap ..which is half a lappy screen...I have images upside down to each other stack...mind you I do crop a quarter of most finals;) because of poor framing...but they stack.
My stack has an hour more but I may have to go out but I will post those photos before midnight hopefully.
Alex

[xelasnave]

When you hit register photos you get the first photo..hit advanced settings and you get the second photo...set your slider way up.
Alex

[Sunfish]

Perhaps a sub from a different night has been included accidentally and set as the reference frame. If the sub has been checked to have a good image of the object and the highest number of stars and is right clicked it can be set as the reference frame.

[Kev11]

OK my poor old head is starting to get around some of this. Registration is just that: aligning the frames or rather the stars in the frames. So Alex and Raymo you are right: changing the registration threshold has no effect on my stacking problem, despite the cryptic message from DSS on the "only one frame to stack" warning. The second part of that says "and help DSS find a transformation between the reference frame and the others" which is Greek to me but seems again to apply to registration only.


So, if I stack the one frame I find the RGB/K window shows an S curve and a tall, narrow bell curve right at the bottom left hand end. If I move the bell curve up to the lower point of inflection of the S curve (which is my standard practice with stacked images anyhow) extra details pop up (which is the standard result).


All of which brings me back to the original problem: the registration process has not found enough stars to allow the stacking process to proceed. Is this perhaps a "fail safe" function to avoid stacking possibly misaligned frames?


Thus my question: can I override that and stack more than one frame because visual examination of the individual frames and the registration scores (low as they may be) does not indicate any alignment problem.

[xelasnave]

Kevin..please just act on my last post and your problems will just disappear.
The solution is simple you just need to do what I suggest...please just try it.
Alex

[Outcast]

If the threshold for star detection is too high (ie: 2%) it will not detect enough stars to frame & therefore cannot align & register frames as it has no point of reference. The 2% figure is not a descriptor of low threshold it is setting a percentage of stars to use. Too low & it has insufficient point of reference.

Set the percentage higher (50% or more) as others have recommended & your problem should be resolved.

[xelasnave]

Why didn't I think of that?

A wonderful explanation Carlton.

Alex

[Kev11]

Thanks Alex and Carlton but you are missing the point. It doesn't matter whether the threshold is set at 2% or 98% or anywhere in between, DSS cannot find more than 4 stars on any of the frames and none on most. It appears to be something to do with the amount of (lack of?) contrast. Consequently, DSS decides there are insufficient points of registration and therefore it should not proceed with stacking.
Cheers

[xelasnave]

Kevin.

What is your current threshold setting?

Have you changed it up from 2% to 50%?

It can't get points of registration if you are at 2%.



Alex

[RyanJones]

My last attempt to explain this.

DSS identifies stars as points of light where the center is the brightest point and the brightness reduces from the center outwards. If your stars are enough out of focus, the center of the star is not the brightest point. It is true that sometimes the smaller stars that are out of focus are so small that the darker center of the star is not resolved and therefore can still be detected as a star but that is not the case with the subs that you posted. This is why DSS cannot find enough stars to align the frames and hence why it refuses to do it. It doesn’t matter what percentage of doughnuts you tell it to look for, they’re still doughnuts.

Cheers

[Startrek]

I replied to Kevin in the 3rd post and mentioned one of the main reasons why DSS won’t stack is due to stars being out of focus

[Startrek]

Another problem could be you have tracking issues or coma issues with your stars if unguided or even guided
Your stars could be elongated slightly or odd shaped so DSS will reject them
DSS has a very rigid algorithm for assessing good and bad stars ( out of focus stars, eggy stars, slightly elongated stars etc....... )
It could be just poor data that’s causing rejection ??
Another thing to check , have you ticked the box rejection of hot pixels in the set up ??
Cannot think of anything else ?

[Kev11]

Thank you everyone for your input. I detect a level of frustration with me commensurate with the frustration I am having with DSS!


The out-of-focus idea I shall have to follow up with past data. In fact Martin, I recall you pointed out that my stars in the images I posted on IIS a while ago were out of focus and elongated. However, DSS registered and stacked those without any problem.


I am holding on to my theory about lack of contrast for now and will contact the DSS developer about that. The thing that should counter that theory is that I would have thought it would be a constant problem for photographers in light polluted sites, but perhaps you all use LP filters or some other technique?


I suppose I should try re-installing DSS in case I have accidentally reset some basic function.
Cheers

[Startrek]

Kevin
Here’s my DSS procedure I’ve been using for nearly 4 years in light polluted Sydney and dark skies South Coast using both DSLR and Cooled OSC cameras with and with out narrowband filters ( not light pollution filters )

Deep Sky Stacker basic procedures

Open Deep Sky Stacker

All tasks and settings are on the LHS column

“Open pictures files” and load your Raw (Cr) files or Fits files from your PC ( make sure you select the correct file type in the drop down to load into DSS )
Open “dark files” and load your darks
Open “flat files” and load your flats
Open “offset/bias files” and load them

Press Check all ( all files will be ticked or checked in the file list at the bottom ready for registering

Press “Register checked pictures”

Register Settings box opens up
“Actions tab”
Tick box - already registered picture
Tick box - Automatic detection of hot pixels
Tick box - Stack after registering
Select the best 90% to 95% pictures and then “untick this box” so you can check your Stacking Settings later ( I prefer to check later )
Under “Advanced tab”- star detection threshold make sure you tick the box “Reduce the noise by using a Median filter”. Then move the Star detection slider to 10% at first but try other settings like 30% or 50% as there may not be enough stars for DSS to detect and align. In some cases where you have poor data due to focus issues or bad tracking you may have move slider down to 5% or even the lowest setting at 2%.
I recommend between 50 and 150 stars , so increase or decrease the star detection threshold to achieve enough stars

NB: if your stacking FITS files from a dedicated Astro CMOS camera then you will find that Star Detection will be low say 50 to 100 stars when the star detection threshold is set at its highest or 2%
This will be worse when you use a narrowband or light pollution filter with your camera. As long as DSS detects at least 8 or 10 stars then it will register the frames and stack for you

“Recommended Settings button”
DSS give you the recommended setting highlighted in green based on the number and type of files to stack
“Stacking Parameters button”
Intermediate files
Only check FITS files box and Use all available processors box
Leave other check boxes clear or unchecked
Press OK then DSS will start registering files

After files are registered main screen will be blank then press “Stack checked pictures”
Stacking Steps box appears
Recommended Settings button
DSS gives you the recommended setting for stacking highlighted in green ( you can change them if you want to )
“Stacking Parameters”
Result tab - check Standard mode
Light tab - check Kappa Sigma clipping
Alignment tab - check Automatic
Intermediate Files tab - check FITS files
Cosmetic tab - don’t do anything
Output tab - check Create Output file, check Append a number to avoid file overwrite, check Create Output file in the folder of the reference frame
Check Use available processors
Press OK
Now Press OK in Stacking Steps
DSS will now stack files based on selected settings

After stacking, DSS will create an output image ready to save
“Important” - do not use DSS RGB level adjustments , leave stacked image as it is ( untouched linear stacked image )

After stacking, in “Processing section” click “Save picture to file” option
Save as window appears , select where you want it saved in your computer and file name etc
Save as type will be FITS image ( 32bit integer )
Under “Compression” below leave as none
Under “Options” below , bullet check - do not apply adjustments to the saved image

Click save and stacked fits file should be saved to your file location ready for Startools or other processing software

Cheers
Martin

[RyanJones]

Martin and I ( amongst others ) have and still do image in Bortle 8 skies. Both with 8” newts at f/5 and f/4. Yes I use filters but have without and until recently Martin has not. Contrast is not the issue.

I will admit to a level of frustration yes. I’m trying to not to come across rude and my assistance is to serve no other purpose than to help a fellow astrophotographer.

[xelasnave]

My frustration turns on the fact you will not answer my simple question which I will repeat...

What is your setting ...the threshold setting???

Have you tried a stack at 50% at all or are you still stuck on "2% is ok"?...


I looked at your images you posted and even though you have stars that are out of focus I am confident there are sufficient "good" ones that DSS will register...to register you need a higher threshold than 2%..It has nothing to do with contrast.

All I would like from you is a simple reply to my question...if you insist on 2% then contacting DSS will confirm your view re contrast is wrong...

What is your threshold setting? Have you tried a stack with the setting at 50%?

If you do not give me a direct answer I will take it that you are trolling and wasting everyone's time...if you are genuine change your setting..



Alex

[glend]

Perhaps it is time to let this thread go, as all the questions have been answered and reasonable advice provided. If there is a lack of understanding at the OP end getting upset is not going to fix that. Let it go.

[Wilsil]

Maybe try another stacking program like Sequator and see what it does.

[Kev11]

Thank you Wilco. SEQUATOR produced a stacked image see attached.


My humblest to everyone, particularly Alex, for not sending the screenshot of DSS message which might have made my problem plain. Apparently no one else has ever got this message from DSS. (incidentally this is with threshold 50% but any other threshold gives the same answer).


I am sorry, Martin to have put you to the trouble of writing out your whole DSS procedure, however I think it is worthy of putting in the IIS handy hints for beginners.


There is more to say about my experience with all this and what I have learnt, but best perhaps saved for another day and a fresh thread.


Avagoodweekend (we have thunderstorms and rain here)

[raymo]

That is the standard message you get when DSS cannot find enough stars to
work with, nothing unusual there; just increase the setting to 80% or more, see how many stars it finds,[will almost certainly be far too many at that setting], and reduce the setting until it finds 50-100 stars,[ usually somewhere between 50 and 75%], and all will be well.
This is absolutely normal, I have done it many times over many years.
As Ryan stated, contrast is not the problem.
raymo

[Startrek]

Kevin
I’ve had this DSS procedure for years as well as many other procedures
i’ve built up during my Astronomy and Astrophotography journey over the past 4 years and still do . Happy to share them on IIS to help others
Martin

[Wilsil]

As Sequator stacks the images, it has to be the settings in DSS.

After setting the slider to 50%, how many stars did DSS says it was seeing?

[xelasnave]

There as no need to post a photo of the message as I am totally aware of the message and I have seen it in the past...(many people have..that is when they find out about setting the threshold setting) .... and at that time I made a post here where I asked how to fix it.... and Raymo gave me an answer like his last post to you ... reading his post brought it all back to me and made me remember why I have it set at 70%.

If SEQUATOR stacks, as you tell us it has, then clearly your setting in DSS as to its threshold is too low...by the way what is your current setting for threshold in DSS?.....perhaps you could post a photo of that as it could answer the question you seem to be missing an answer to...however thinking more about the situation I thought that Ryan really gave the best advice in that you throw out the subs you are trying to stack because really they are extremely poorly focused ...you cant expect to do much with such out of focus stars even when stacked..the main object will be very blurred after a stack and probably be beyond processing...that is the point ...it seems that you are not using a focuser mask ...do you use a mask to focus?...without one achieving focus can take forever..as you need to take a photo then blow it up and check it and repeat until your stars are in focus..it would take me 15 minutes and longer when I had a dslr without enlargement in the viewer screen.....however hopefully you have learnt that paying attention to your threshold setting in DSS is important and that poor attention to this aspect will have DSS unable to register the stars necessary for it to stack...and that contrast is entirely irrelevant..
Good luck and I for one look forward to you posting your next completed image.
Alex

[Sunfish]

There are some pretty weird looking stripy stars and artefacts in that sequator stack . Some of your subs must be bad . The only time I have had that message is when my subs have been bad , too dark or too noisy in Ha .