Silly problem with DeepSkyStacker - help needed please

[Psychobilly]

Dont even think of using Jpegs.. Starfields are just about the worst thing you could thing of storing as a jpeg artefacts city..... It's got to be RAWS for this stuff...

Billy...

[**Space*Hopper**]

Has anyone had any joy with stacking the 10 x full size subs posted within this thread ?

[Psychobilly]

it wont do anything for star shape etc.. its just the Proper way of doing it... why neuter the camera by getting it to reduce the raw data form the CCD to a lossy compressed (noisy) 8 bit format...

[**Space*Hopper**]

Ok understood, I will give the RAW format a go for my next project but that won't help me with this one for which all my subs are in JPEG format.

Surely there must be some clever person out there who can find a method of successfully stacking these

[Psychobilly]

We'll i downloaded them all and whilst DSS does find quite a few stars... registar (not registrax) decide there werent any in there which it could use for stacking.

If you look at the trailing on the individual large subs its not really suprising that the stars all merge together when stacked. Also if you look very carefully at the stars you will see that they are "donut" mode suggesting the focus of the imaeg was a bit soft... Perhaps when you reduced the size of the frame both of these "improved" allowign DSS to a better job

It might be better to repair the individual sub - there was a technique mentioned on the forum while back... and then stacking the repaired images...

I also have another plan which I am working on seeing as it's only 10 subs... tried manually processing the images as a multillayer document in CS3... still no real joy... I can get a better image from just the best sub

Billy...

[**Space*Hopper**]

Cool, Thanks Billy

[Psychobilly]

Ok...

I manually aligned the 10 frames using a multi layer image in CS3 and diff mode.

I then cropped a standard size portion of the aligned image from each layer...

Took these into DSS...

Tried stackign the lot...

1 frame got offset from the others by quite some distance so i removed that from the stack...

Processed the result (which was very noisy) in CS3

[ATTACH]20323[/ATTACH]

Billy...

I still think you can get a better image using only the best 2 or 3 subs...

[Kaptain Klevtsov]

Had a quick play (well, a couple of hours really) with the posted subs and tried some stuff. A Lucy Richardson deconvolution sorted out the individual stars with limited effect, but totally messed up the core. I managed to DC three subs only, the rest crashed the 'puter. Stacking the three deconvoluted subs wasn't much better than a single raw sub as posted wrt the core section.

It looks like the focus was out just enough to mess this up, but at such long focal lengths its a right royal pain to get the focus bang on. The giveaway to the focus is the donut looking stars.

If you ever get the focus cracked, let me know how you did it, its my biggest nightmare with DSLR imaging.

Kaptain Klevtsov

[Psychobilly]

KK for focus use a Bathinov... or one of the other slot type focus mask... typically a 2 s sub at iso 1600 is all thats needed to show a decent image...

Billy....

[Kaptain Klevtsov]

I would Billy, but I've not had a night without cloud for months now. Last time I was out imaging they hadn't been invented. Currently thinking of flogging all the kit as its doing nothing. Hoping it gets better soon or its bargain time for someone!

Kaptain Klevtsov

[Psychobilly]

I hope not KK theres a few other thinking the same way ....

Billy

[mostschaedel]

Hi Space Hopper!

This is not an issue of Software, or your C8, or JPEG versus RAW, etc...

The images are simply FAR out of FOCUS.

The only way to solve the issue in the future is -

take your time to focus properly, after shooting of the first image

check the focus carefully, again and again.

The ED images were better to work with because they do not have

the black center if you are out of focus. SCT images out of focus do have

a black disk in the middle.So the software cannot find the middle of the

star for registering and stacking.

Gerald

[**Space*Hopper**]

Firstly a big thanks to Billy & KK for having a play with the 10 subs I posted and for all of the info they have provided.

Secondly a big thanks to everyone else who also provided useful info within this thread.

I had another go at M3 last night with the camera set to RAW but the resulting subs will not stack either, even though I had taken even more care with the focusing. I guess for the reasons already mentioned in this thread, the focusing of the C8 is more critical than my ED100 and certainly more so than I can achieve on the tiny screen of the 400D, (even when zoomed in). I therefore feel that a Bathinov mask is definitely the way to go.

Has anyone used the below mask generator to create a mask for a 2000mm C8 SCT ?

astrojargon - Bahtinov Focusing Mask Generator: Overview

If so I would be very interested to know how it worked along with any tips you discovered while fabricating.

Thanks again to all of you

Ian

P.S. One thing that I do not fully understand about the mask, my normal proceedure is to center an object with a wide angle eyepiece, then replace the eyepiece with the camera and re-focus. However if the mask works by focusing on a bright star with the camera attached, how do you then locate the much dimmer object that you actually want to image ?

[Psychobilly]

I know a man who used to make these .....

Billy...

[mostschaedel]

Hi Space Hopper!

A mask gets you much faster near to the best focus position

than anything else.

But not always does it help to find the best focus position.

A software like ImagePlus, DSLRFocus, Maxim DL etc which analyzes

the dim star on pixel level gets you into perfect focus.

Everything you estimate with your eyes on the screen is less precise than focusing with star analyzing software on pixel level.

[Reggie]

Is iso speed an issue when focusing? what about the length of exposure used if you don't have a liveview screen?if you are using any of the above software to focus?

[mostschaedel]

Is iso speed an issue when focusing? what about the length of exposure used if you don't have a liveview screen?if you are using any of the above software to focus?

Hi Reggie!

I am using the highest ISO Speed for focusing with the Laptop

and DSLRFocus. The focusing is done on a dim star of the final

view in the center which gives a high curve for the FWHM value.

Usually around 15 seconds exposure per checking.

Liveview is also good for getting fast very near to the optimum focus but it is a visual check

and therefore precision depends on your eyes.

The software delivers curves and values for more precise

focusing.

DSLRFocus is used for the older Canon DSLRs, Imageplus for the newer ones.

[**Space*Hopper**]

Not being one to give up easily, today I made a focus mask from the free C8 template available at Bahtinov Mask by Focus-Mask

I used this to focus on a fairly bright star and it appeared to work exactly as described. I then searched for M3 and this was harder than usual as I left the camera on and had to just take images of the rough area until it appeared on the cameras screen, (normally I would use a wide angle eyepiece to locate M3 and then swap to the camera but having just focused the camera with the mask I could not do this). Anyway eventually I found it and took some subs in RAW mode. As always to my eyes they looked good.

Eagerly I then came inside and loaded them into DSS and guess what, they wont stack either

:crybaby2:Anyone want to buy a C8 ?

[roundycat]

Ian,

I first noticed this thread when you first posted and ignored it as I know nothing about DSS and don't want to know anything about it. I also do not use a DSLR for my pics so, as you can guess, I know nothing about that either.

Can I be honest? I do not want to put you off but I would say you have to learn to focus before you can do anything else. You mention that this is probably your best ever M3 but I am convinced it can be made a lot, lot better with minimal effort.

How do you focus. This mask business may help and it is certainly all the rage at the moment but people have been focusing successfully for years without it. Do you use the camera screen? If so you are flogging a dead horse. Can you set a sub frame for a single exposure and display it on a monitor? Can you then repeat the exposure, which should be no more than 1 second, and examine it with some focus assisting software? If you do not have the s/w but can take repeated very short exposures you should be able to get a better focus just by watching a bright star on your computer screen. I am banging on about focus simply because yours is miles off and you are not going anywhere until you get that right.

Spend a night just playing around with focus and see if you can get a very quick download (of a sub-frame) as that will give you almost real-time focusing. You must either measure the Full Width Half Maximum or do it by eye on a computer screen with the screen display at 100%.

I downloaded four of your frames and I was very surprised at what happened next. I use an Artemis 35mm chipped camera that produces 21MB frames but yours took longer to align then four of them would. By a very considerable time. When I combined the four in Maxim using SD Mask I had enough time to go and make a sandwich. My normal frames would take barely 30s to align and stack on this computer. Something is wrong there but I don't know what. Could it be that they were jpeg's? As others have said above, do not touch jpeg's with a barge pole until you post on the internet.

The focus was out on all four frames but Maxim did align them more or less straight away. The signal level was very, very faint indicating your exposure needs to be much longer. The background was brown but cleaned up in PS easily. I cropped and have posted my attempt on your data below.

Don't give up and don't give up with the C8. It will come right in the end but by the sound of it it will come right a lot quicker if you use different s/w. But then I would say that.

Dennis

[Reggie]

Whilst people have been focusing sucessfully for years (no argument there) if you are new to it and aren't picking up on techniques then it can be a frustrating exercise in futility, especially on long FL scopes with average mounts. You can spend more time waiting for the rig to stop wobbling than you actually do focusing.

If it's a dslr, the simplest tool to use is a bright star (capella, betelgeuse etc.) and the viewfinder. this will get you close in the dark without any extra tools.

Also look at the direction you are turning the focus wheel as you approach being in focus, play with it, rack the foucser in and out so you can see how it behaves.

You should be able to see a sweet spot where the star looks smallest, once you get to that its time to start using some test shots to gauge focus and how small the unavoidable minute movement of the focus wheel is to get true focus. Its these last few minute movements that turn a difficult to process shot into an ok one, practice with software will turn it into a great one(all other things being equal, correct exposure, right iso, etc.). If you don't nail focus all other efforts are extended and wasted.

Some focuser knobs/wheels are way too small for the ota they are on, take my 6" sct, its about 1.5cm in diameter, makes it a real pain to use in its stock state, as a small movement translates to a relatively large focus shift, it wasn't until I got a much larger diameter thing to fit over the focuser shaft that I could actually achieve the focuser resolution I needed.

I am basically suggesting making the diameter of your focus knob/wheel larger in diameter by attaching something to it with.... a larger diameter. Round milk bottle lids have been used, at one point I used the cap from an aftershave bottle, have a look around the house and see if there is anything that might remotely fit Remember the bigger the diameter, the more you have to move it, the greater the accuracy of focus will be with each movement.

You could also look at getting some kind of markings around your focus wheel, this will allow you to better judge how focus is going.

There are plenty of 'online ruler' generating/printing websites, just tell it the circumference of the focus knob and how many markings you want, it will generate one for you which you can then print out (making sure to not have any stretch/resize options active on the printer).

Once you have it printed out you can stick it around the outer edge (or flat face depending on type of focus wheel/knob).

Get yourself at close(ish) focus with the viewfinder and check the focus markings position against a fixed mark on the ota, then you never come back past that point, take a test shot, move the focuser, check whether your target is smaller or larger.

If its larger then move you focuser back to the original point and continue in the opposite direction from your first movement.

If its smaller note the focuser marking again and continue, repeat this until you are happy, all the while noting how much you have to move the focuser to make a difference to the object.

If you go over the focus point simply note the point where you are when you've gone over focus and move back slightly towards the last focus point before you went through focus, rinse, repeat until in focus

I appreciate my methods here aren't anywhere near as sophisticated as the software methods but I feel just getting to grips with focus at its basic level will allow you to appreciate whats going on and get yourself where you need to be, then all of my fellow sgl'rs techniques will become very simple to get your head round and implement.

[**Space*Hopper**]

Hi everyone,

After my disapointing attempt with the focus mask I have made, the very helpful & constructive advice posted in this thread motivated me into having another bash at M3 last night, so here's how it went:

I took everything out into the garden and played around for a while to give the scope enough time to cool down, (or so I thought). Previously I had pointed the scope at a fairly bright star for the purpose of using the focus mask but this time round I used one of the brightest stars I could find and was amazed at how much difference this extra brightness made to the clarity of the lines appearing in the focusing pattern and I was therefore able to easily see if I was turning the focus knob the right or wrong way and quickly obtained perfect focus, (according to the mask).

But then the problems started. As previously mentioned I usually locate a deep space object using a low power eyepiece, center it and then replace the eyepiece with the camera and re-focus. However with the camera now in theory perfectly focused, I obviously did not want to remove it so had to try an locate M3 purely by pointing and shooting, (does anyone have a better way of doing this ?). After around 45 mins of this without even a glimpse of M3 I felt like giving up and started to just point the scope at other stars to at least see how the focus mask had performed and much to my surprise, there were donuts everywhere Pointing the scope back at the original star that I had used to focus with and re-attaching the mask showed that the focus had shifted by a long way and I now believe that I had not allowed anywhere near enough time for the air, mirrors & glass inside the scope to properly cool down. Anyway I re-focused using the mask and thought that I would have another go at finding M3 as that was going to be the best way of proving if the focus mask was doing its job properly. This time after about 10 minutes I found it and quickly started taking subs, (in RAW mode of course ). Even though the LCD screen of my camera is very small, when zooming into the subs I could tell that the focus of the stars in the middle of the cluster looked better than I ever remember seeing them before so I was quietly optimistic.

I took just over 30 x 30 sec subs along with 8 x 30 sec darks and then came in for what I knew was going to be the moment of truth, would any of the subs stack in DSS at full resolution ? I picked 17 of what I considered to be the best subs and loaded them along with the darks into DSS. It proceeded to register 17 and then stack 11 but during this I thought to myself, dare I look at the resulting image on the screen at the end of the process, would all the subs be out of alignment Anyway stacking of the 11 completed and when the image appeared on the screen a very big came across my face as the 11 frames had correctly stacked this time.

So what have I learnt from all this ?

(1) My C8 scope is awesome but requires more attention to detail than my ED100 in terms of critical focusing, cooling down times etc.

(2) While my mount was an absolute bargain for the £100 I paid for it, it is now the weekest link in my system and for every sub I use, I delete around 3 or 4 because of tracking errors, (even some of the subs I use for stacking do not have perfectly round stars).

(3) My focus mask, crudely cut out from a bit of cardboard using the downloaded template from the site previously mentioned in this thread is simply awesome and for people like me who are strugling to obtain perfect focus, I cannot recommend it enough, (especially as it costs nothing but time to make).

(4) You guys were spot on as always thanks, poor focus resulting in donut shaped stars with dark centers was the reason why DSS could not stack the subs but fingers crossed, armed with my focus mask and the lessons learnt from this M3 project, my next project with the C8 should not be such a trying one

For your viewing pleasure I have attached:

(1) An image of the bright star I used to focus

(2) An image of the star viewed through the mask with the scope out of focus.

(3) An image of the star viewed through the mask with the scope in focus.

(4) An image of my crudely produced mask, (In time I will probably take more time to make a better one out of a lino floor tile or similar).

(5) An image resulting from the DSS stack of 11 x full res RAW subs + 8 x darks quickly processed in Paint Shop Pro 8. I will undoubtedly take more time and have a more careful go at the PS Pro processing but at this stage I am so happy just to have an image from full res subs words cannot explain

Anyway thanks again to everyone who contributed to the invaluable help and advice in this thread to get me to this point, I really do greatly appreciate it.

Ian

M3-27-3-09-11 FILES - FULL SIZE STACK - PROCESSED-DIMMER CENTRE.bmp

[Psychobilly]

Nicely Done Ian

The mask are quite good aren't they

Just remember to take them off ... I think most off us who use them have left them on at least once

Peter....

[roundycat]

that focused image does look sharp. You may get it still sharper with software assist. Try to find somewhere cold to store your scope.

Dennis

[Reggie]

The image does look much much sharper, your getting egg shaped stars but you've already pointed out your £100 mount being the weak link, so that'll be the tracking

For your issue of not wanting to move the focuser, leave teh camera on there and take a test shot and compare it to what you can see in a planetarium, or buy a parfocal ring for the EP, this goes around the base of the ep allowing you to draw the EP out of the diagonal until its in focus, you then tighten the screws and both items (EP and camera) should be in focus when you swap them over, the trouble might come if you have a diagonal in the way as this will have added extra distance to the optical path, in which case you'd have to use an extension tube on the camera/T-ring to compensate for the extra distance taken up by the diagonal as well as the parfocal ring.

Donut shaped stars after focusing may not be a cooling issue, it could also be a slipping focuser issue, check whether your scope has a focus lock mechanism (usually a bolt/bolts) so once you have it in focus you just lock it and it won't move. It could be the angle the scope has been pointing (anything at zenith is a good culprit) or it could be the mount being a bit shaky as its moving.

With DSS you can chuck all your images in and let it register them without stacking, you can set a %age of images to keep, or once they've registered you can check them individually and discard anything else.