M101 LRGB

[Davey-T]

Looking good Rodd, some elongated stars at the top if you zoom in but I tend not to worry about that sort of thing.

A lot of imagers get a bit OCD about the slightest blemish and spend / waste lot's of valuable imaging time messing about in search of perfection, life's too short, well it is at my age 

Dave

[Rodd]

11 minutes ago, Davey-T said:

Looking good Rodd, some elongated stars at the top if you zoom in but I tend not to worry about that sort of thing.

A lot of imagers get a bit OCD about the slightest blemish and spend / waste lot's of valuable imaging time messing about in search of perfection, life's too short, well it is at my age 

Dave

Do you think it might have something to do with too short of a settle down time after dithering?  

[Davey-T]

No, looks more like some slight tilt somewhere, you can use CCD inspector to analyse your images but it's the road to madness

Dave

[Rodd]

Darn it.  I thought it was pretty good.  First time I automated a sequence, first time I dithered using software.  Now I need to decide whether it is worth adding Ha and OIII data to the image--or to scrap it and start over.

[Davey-T]

It is pretty good, I'd carry on and get a finished image.

Dave

[Rodd]

6 minutes ago, Davey-T said:

It is pretty good, I'd carry on and get a finished image.

Dave

The most likely source of the tilt is the junction of the Camera and the scope--the 2 1/4 nose piece slips into the focuser and is held with a brass ring clamp system.  The nose piece has groves and ridges to help establish a firm connection, but it only has to be off by the slightest amount.  The problem is if I correct the tilt (can only be done manually, so no guarantee it will be better), I'll need 2 sets of flats.  Hopefully it doesn't get worse as I move forward.

[gnomus]

On 14/05/2016 at 23:12, Rodd said:

 

.... then a touch of saturation .....  

First off the M101 looks pretty good - no need to beat yourself up over it.  Nonetheless, I quite understand your wish to improve.  I don't know anything, so please take what follows with a large pinch of salt.  

I like PixInsight and use it for a number of things.  I have seen excellent, indeed outstanding, images processed by people who have used PixInsight for everything.  I have tried using PixInsight for everything.  I think I get my best results when using a combination of PixInsight and Photoshop.  If I go the PixInsight-only route, then, try as I might, I cannot get a controlled result - everything looks over-processed.  This is the cardinal error that I keep making - trying to get more from the data than the data will allow me to have.  I still 'over-process' in PS but in PI I feel I have less 'control'.  Sometimes I will hop back and forth from one program to the other whilst at different stages.  

Your subs look good.  I do see some minor elongation in the top corners all of which seem to me to be moving away from the centre.  I thought that this could be caused by incorrect spacing between the reducer and the camera sensor.  Are you certain you have nailed this?  But as Davey-T says they are not a big issue and you could forget about it - if you look around you will find worse corner elongation.  

I do  wonder if you've overstretched the M101 image a little too much.  The background measures ~50 in PS - people like Olly Penrice aim for less than half that.  It looks also like you may have blown some highlights (the histogram on the JPG certainly shows this) - another sign of over-stretching perhaps.   You should not eliminate Bias or Flats.  I find Flats help considerably.  I had some very dodgy flats for an M45 image recently and so I tried a version where I just used DBE.  I found that I didn't get nearly the same amount of 'wispy stuff' in the image without flats.  I concluded that flats were really important.  Biases are of course easy to do.  I have found that Darks seem to produce the last tangible benefit - maybe even make things worse sometimes - but as I say, I don't really know.  

Flats will not, however, get rid of all gradients and I find I still have to use DBE.  There are 'tricks' that you can use to tame backgrounds.  The first is probably to avoid over-stretching in the first place.  But even then, things can be done.  I hope you don't mind, but I had a quick go at your M101 background.  It looks a bit artificial ('plasticky') given the starting point (8 bit JPEG) but..... I'm just trying to show that you can even things out even after an indifferent Flats-DBE result ...  (don't laugh now I only spent a few minutes on it)

There are also a few neat ways of increasing saturation in Photoshop without there being noise-eruptions.    I'm happy to share them with you if you have PS - I stole them from someone else after all!!!! 

You're very nearly there I think.

Good luck

[Rodd]

19 minutes ago, gnomus said:

First off the M101 looks pretty good - no need to beat yourself up over it.  Nonetheless, I quite understand your wish to improve.  I don't know anything, so please take what follows with a large pinch of salt.  

I like PixInsight and use it for a number of things.  I have seen excellent, indeed outstanding, images processed by people who have used PixInsight for everything.  I have tried using PixInsight for everything.  I think I get my best results when using a combination of PixInsight and Photoshop.  If I go the PixInsight-only route, then, try as I might, I cannot get a controlled result - everything looks over-processed.  This is the cardinal error that I keep making - trying to get more from the data than the data will allow me to have.  I still 'over-process' in PS but in PI I feel I have less 'control'.  Sometimes I will hop back and forth from one program to the other whilst at different stages.  

Your subs look good.  I do see some minor elongation in the top corners all of which seem to me to be moving away from the centre.  I thought that this could be caused by incorrect spacing between the reducer and the camera sensor.  Are you certain you have nailed this?  But as Davey-T says they are not a big issue and you could forget about it - if you look around you will find worse corner elongation.  

I do  wonder if you've overstretched the M101 image a little too much.  The background measures ~50 in PS - people like Olly Penrice aim for less than half that.  It looks also like you may have blown some highlights (the histogram on the JPG certainly shows this) - another sign of over-stretching perhaps.   You should not eliminate Bias or Flats.  I find Flats help considerably.  I had some very dodgy flats for an M45 image recently and so I tried a version where I just used DBE.  I found that I didn't get nearly the same amount of 'wispy stuff' in the image without flats.  I concluded that flats were really important.  Biases are of course easy to do.  I have found that Darks seem to produce the last tangible benefit - maybe even make things worse sometimes - but as I say, I don't really know.  

Flats will not, however, get rid of all gradients and I find I still have to use DBE.  There are 'tricks' that you can use to tame backgrounds.  The first is probably to avoid over-stretching in the first place.  But even then, things can be done.  I hope you don't mind, but I had a quick go at your M101 background.  It looks a bit artificial ('plasticky') given the starting point (8 bit JPEG) but..... I'm just trying to show that you can even things out even after an indifferent Flats-DBE result ...  (don't laugh now I only spent a few minutes on it)

There are also a few neat ways of increasing saturation in Photoshop without there being noise-eruptions.    I'm happy to share them with you if you have PS - I stole them from someone else after all!!!! 

You're very nearly there I think.

Good luck

Gnomus--nice job--image looks allot better.  You are right about all of it--I overprocessed trying to compensate for lack of data.  I worked with Televue on spacing.  Maybe after I perfect other things I will take a look at this again.  Televue does sell very tiny spacers so I should be able to tweek it.  But my sensor is only 17 x 13 or something, and I use the large field corrector lens.  So spacing shouldn't be as important as it is with  large format sensors.

I don't have Photoshop--can't get past the monthly subscription.  I have Nebulosity, Gimp, and a few others if IO want to meddle.  I hear Gimp is similar to PS.  To tell you teh truth, I can't habdle another learning curve!  Right now I have Maxim DL, Pixinsigjht, Flat fields, and others on top of these related to the Celestron.  This is tough climbing simultaneously.  

 

I do appreciate your input, and you obviously know more than you give yourself credit for.  I plan on adding data to M101.  I got waylaid by the Crescent Nebula for the time being.  But I have decided not to try to complete images on a nightly basis--there is just not enough time, especially in the Summer.  So, I am building a library (an extensive one).  I add subs as targets become available.  It doesn't give you the immediate satisfaction, but in 6 months or so, I will have accumulated allot of data on allot of targets.  20-25 hours of data on M101 should improve it.  The hardest thing is the flats--they have to be taken with each group of subs and organized so the right flats are used with the right subs.  So my Ha stack might consist of 5-6 Ha sub-stacks each batch preprocessed with their own flats and integrated into stacks for a final stack integration.  I am working on developing a way to simply my goals.  Careful file, folder and directory naming will be crucial.

The deeper I get into this hobby, the deeper I want to go.  I have a book (Burhnams Celestrial Handbook in 3 vol.-great books by the way) that has allot of images taken with professional observatory telescopes (200" 100", 48". 36" etc).  The images we produce as amateurs blow these images away.  Its amazing.

[gnomus]

NP Rodd.

4 minutes ago, Rodd said:

Gnomus--nice job--image looks allot better.  ..... I worked with Televue on spacing.   ....

You can get 'delrin spacers' which allow for very fine adjustments (bit of a pain to get on and off through) - https://www.firstlightoptics.com/adapters/baader-t2-delrin-spacer-ring-set.html

7 minutes ago, Rodd said:

 .... I don't have Photoshop--can't get past the monthly subscription.....  To tell you teh truth, I can't habdle another learning curve!  .....

 I don't like the monthly subscription either.  I bought (and will be sticking with) the pre-subscription version.  If you can handle PixInsight you can certainly handle Photoshop (it's much more intuitive).

9 minutes ago, Rodd said:

......  But I have decided not to try to complete images on a nightly basis--.....

Absolutely.  This was probably the biggest realisation I made moving from DSLR to CCD.  Abandon thoughts of getting a result in one night. 

I agree - whilst I continue to beat myself up over my astrophotographic disasters, I have to keep reminding myself what we are trying to do here.  I am still early on into this process and I haven't lost the 'I cannot believe that I can get these images from my backyard' feeling.  I hope I don't lose that.

Regards

Steve

[gnomus]

16 minutes ago, Rodd said:

But I have decided not to try to complete images on a nightly basis--there is just not enough time, especially in the Summer.  So, I am building a library (an extensive one).  I add subs as targets become available.  It doesn't give you the immediate satisfaction, but in 6 months or so, I will have accumulated allot of data on allot of targets.  20-25 hours of data on M101 should improve it.  The hardest thing is the flats--they have to be taken with each group of subs and organized so the right flats are used with the right subs.  So my Ha stack might consist of 5-6 Ha sub-stacks each batch preprocessed with their own flats and integrated into stacks for a final stack integration.  I am working on developing a way to simply my goals.  Careful file, folder and directory naming will be crucial.

It should be possible to complete a single targets in a few nights of imaging.   I have a permanent set up, so I use one set of flats for all my (say) Lum subs even if the Lum was taken over different nights.  (Whisper: I hear a rumour that some people just use their Lum flats for all their subs).  If you are not permanent then you could shoot your Lum flats and then run the calibration routine in PixInsight (Preprocessing).  This will give you a folder full of calibrated Lum images.  Once you have your Calibrated (not registered) Lum images from your various 'Lum' nights, register them and stack them.  It's not a big deal.  

The other option if you do have to tear it all down every night is to leave the camera and wheel on the telescope - then one set of flats should do you.   (So long as you don't jiggle anything).  As long as you don't change the orientation of the camera and wheel in relation to the scope, I think you can get away with slightly different focuser positions (after all you need to refocus as the night goes on anyway, don't you?).    

[Rodd]

10 minutes ago, gnomus said:

It should be possible to complete a single targets in a few nights of imaging.   I have a permanent set up, so I use one set of flats for all my (say) Lum subs even if the Lum was taken over different nights.  (Whisper: I hear a rumour that some people just use their Lum flats for all their subs).  If you are not permanent then you could shoot your Lum flats and then run the calibration routine in PixInsight (Preprocessing).  This will give you a folder full of calibrated Lum images.  Once you have your Calibrated (not registered) Lum images from your various 'Lum' nights, register them and stack them.  It's not a big deal.  

The other option if you do have to tear it all down every night is to leave the camera and wheel on the telescope - then one set of flats should do you.   (So long as you don't jiggle anything).  As long as you don't change the orientation of the camera and wheel in relation to the scope, I think you can get away with slightly different focuser positions (after all you need to refocus as the night goes on anyway, don't you?).    

I do focus periodically--but the Televue is amazingly stable.  One night I left it set up for the next night and after 24 hours the focus was almost perfect--had to tweek it, but it was close.  I refocus after every filter change, or if I am doing a bunch of subs in 1 filter I refocus about every hour, or if the temp drops noticeably.   A few nights of imaging typically translates to a couple of weeks due to weather.  I do want a permanent setup.  One time last month I got lucky and was able to leave it setup for 7 days.  What a pleasure to turn it on and bam--your focusing on your focus star and get right into it.  I am progressing in leaps and bounds--especially since I got rid of CCDops and switched to Maxim DL.