The "No EQ" DSO Challenge!

[SteveNickolls]

34 minutes ago, Stub Mandrel said:

You ALT/AZ folks shouldn't think using an EQ mount is a bed of roses. Here's what happens if last night's PA is 12 hours out:

 

 

He, he we know EQ mounts aren't beds of roses :-) Good luck sorting out your mount problem though.

Best Wishes,
Steve

[Filroden]

41 minutes ago, SteveNickolls said:

Hi Ken,

I've had a try with StarTools on your downloaded image. The COLOUR module did not like the data and would have washed the nebulosity firther away so I missed that step out. The background is becoming darker and some nebulosity is developing. If it had been a more regular shaped object using isolate/mask in the LIFE module can really bring out an object more. I hope you can get more data to get this to be a good success.

Cheers,
Steve

Thanks for trying it Steve. I'll see what more data I can collect and in the mean time I'm going to try deconvolution again with some better masks to see if I can sharpen it.

[The Admiral]

Hi Ken,

Well, that was fun, thanks for posting. I've got so rusty at using ST. For what it's worth, here's my scrappy attempt at processing using ST. Mind you, there is such a multiplicity of processing options, as I guess with any software, that one doesn't know just how much is 'real'.

Ian

Edited September 12, 2016 by The Admiral

[Filroden]

Another great effort. Thank you for taking the time to process it. I'm glad you had some fun with it.

I've had a second try, this time managing to use Deconvolution - probably the hardest module I've found in Pixinsight so far. I've also reduces the stars and toned back the red. I'm not sure this is better as I've completely killed the background but I can see more of the fainter nebula.

BTW, does anyone else think this nebula looks like a) a baby with its head at the lower end of the photo or b) Elvis, with his quiff to the top?

For interest, I've also included two pictures of my current set up position (one with and without flash). I'm having to use the fence to block the light right next to the garden. There is one at the end of the garden too. Pretty amazing that night seem to affect my images with little discernible pollution and a rough measure of SQM at 20.4. If I also stand by the fence I can just trace the Milky Way overhead.

[Nigel G]

Ken and Steve.

I tried processing Kens soul tif with much the same result,  the colour adjustment would not bring out the emissions like I'd hoped.  Adjusting the dark saturation to 3 bought out a bit but also I think was amplifying any noise in the image.

Was the image white balanced , I'm wondering if I used the correct opening tab in StarTools. 

Nige.

[Filroden]

6 minutes ago, Nigel G said:

Was the image white balanced , I'm wondering if I used the correct opening tab in StarTools. 

It was debayered but I'd not applied white balance. This was straight from the integration. I'd normally remove any gradient then neutralise the background and run colour calibration as my first steps. I don't know what that would be in StarTools terms. 

Edited September 12, 2016 by Filroden

[The Admiral]

Well, I guess that you live at No 20!

Not an ideal lighting arrangement is it, but it does at least appear to be directed at you rather than the sky, so at least you can get out of its direct rays.

Baby or Elvis? Not until you mentioned it!

Ian

[SteveNickolls]

With immediate light overspill like that Ken you are doing extremely well to take the images you do. I've just been watching one of the old Astro Imaging Channel sessions (on You Tube) about light pollution and the 'orange peel effect' noise we all see in images can be attributed to light pollution.

I think you have started the subject of a ramble on this thread Ken-that is what our imaging locations look like at night! Will post up when I get chance to take a night shot.

Can just make out the upturned baby and Elvis too in your image now you point it out.

Cheers,
Steve

[Herzy]

3 hours ago, Filroden said:

Another great effort. Thank you for taking the time to process it. I'm glad you had some fun with it.

I've had a second try, this time managing to use Deconvolution - probably the hardest module I've found in Pixinsight so far. I've also reduces the stars and toned back the red. I'm not sure this is better as I've completely killed the background but I can see more of the fainter nebula.

BTW, does anyone else think this nebula looks like a) a baby with its head at the lower end of the photo or Elvis, with his quiff to the top?

For interest, I've also included two pictures of my current set up position (one with and without flash). I'm having to use the fence to block the light right next to the garden. There is one at the end of the garden too. Pretty amazing that night seem to affect my images with little discernible pollution and a rough measure of SQM at 20.4. If I also stand by the fence I can just trace the Milky Way overhead.

You could probably get an old cardboard box and flatten it out and tape it with a small frame to the fence to keep the scope hidden from the streetlight. The light is most likely spilling into your frames and causing nasty gradients. 

This will be especially apparent if the seeing isn't great (i.e. A lot of dust and fog in the air) the light will start lighting up the surrounding air near it and you will be within that area. 

 

Edited September 12, 2016 by Herzy

[Filroden]

12 minutes ago, Herzy said:

You could probably get an old cardboard box and flatten it out and tape it with a small frame to the fence to keep the scope hidden from the streetlight. The light is most likely spilling into your frames and causing nasty gradients. 

This will be especially apparent if the seeing isn't great (i.e. A lot of dust and fog in the air) the light will start light up the surrounding air near it and you will be within that area. 

 

I think I've been lucky so far. The fence screens it completely already and the Dynamic Background Extractor finds no gradient. You're right though, poorer seem might diffuse the light and interfere more. I may need to increase the fence height with a temporary screen. 

Thankfully all the lights in the neighbourhood are LED and screened to only point down. I think that's why I'm seeing such dark skies. From the top floor (above the lights) I get some amazing views.

[Filroden]

1 hour ago, The Admiral said:

Baby or Elvis? Not until you mentioned it!

And now I can't decide if it's an upturned baby or a side view of ET.

[SilverAstro]

23 hours ago, Filroden said:

And now I can't decide if it's an upturned baby or a side view of ET.

I cant see any of those, but I can see a weasel !

I cant download your fit so I had a fiddle in Gimp with your jpg for a consolation I think I have over-cooked it a bit

Edited September 13, 2016 by SilverAstro

[jimbo747]

4 hours ago, SilverAstro said:

I cant see any of those, but I can see a weasel !

I cant download your fit so I had a fiddle in Gimp with your jpg for a consolation I think I have over-cooked it a bit

I can see at least 3 cat faces in that image, possibly as many as 6.

[Nigel G]

I see Elvis looking up at the sky  ☺

[RobertI]

On 12 September 2016 at 18:24, Filroden said:

Another great effort. Thank you for taking the time to process it. I'm glad you had some fun with it.

I've had a second try, this time managing to use Deconvolution - probably the hardest module I've found in Pixinsight so far. I've also reduces the stars and toned back the red. I'm not sure this is better as I've completely killed the background but I can see more of the fainter nebula.

BTW, does anyone else think this nebula looks like a) a baby with its head at the lower end of the photo or Elvis, with his quiff to the top?

For interest, I've also included two pictures of my current set up position (one with and without flash). I'm having to use the fence to block the light right next to the garden. There is one at the end of the garden too. Pretty amazing that night seem to affect my images with little discernible pollution and a rough measure of SQM at 20.4. If I also stand by the fence I can just trace the Milky Way overhead.

 

 

Those street lamps are so annoying aren't they, sounds like they don't affect the images which is great, but they are irritating when they shine in your face. I have three victorian gas light style lamps within 100feet which prevent me seeing anything visually!  Light does leak into the ota at lower altitude when imaging (video astronomy in my case) i have a portable screen which i use to shield the ota, comprising a large black piece of card attached to a tripod. Works a treat.

Some amazing images on this thread, and I watch with interest.

[Filroden]

7 hours ago, jimbo747 said:

I can see at least 3 cat faces in that image, possibly as many as 6.

If you can see cats and this is the Soul Nebula then surely there must be nine in there? Happy hunting the other three

[Nigel G]

 I'm having trouble loading light frames up to DSS it stops running.  It will load darks and bias and some lights but not  all my lights.  Has anyone experienced this problem.  Is there a limit to amount of frames.  I'm trying to load 220 frames. It lets me load some of them if I do it in batches but not all of them. 

Any ideas

Nige.

[The Admiral]

I haven't had that, but then I don't think I've loaded that many frames. I'm not aware of a limit, but if anything I wonder if it's a size limit. How much RAM do you have, though I'm not sure if this is critical to DSS ? I have similar problems with AS!2, though I know it's not the same beast. Come to think of it, is DSS 64-bit or 32?

Sorry, not much help!

Ian

PS. Do you get any error windows or does it just stop?

PPS. You could try using the Windows task manager to see how much memory is being used.

Edited September 14, 2016 by The Admiral

[Stub Mandrel]

18 hours ago, SilverAstro said:

I cant see any of those, but I can see a weasel !

I cant download your fit so I had a fiddle in Gimp with your jpg for a consolation I think I have over-cooked it a bit

Now it looks like King Neptune from the SpongeBob movie...

[Nigel G]

6 hours ago, The Admiral said:

I haven't had that, but then I don't think I've loaded that many frames. I'm not aware of a limit, but if anything I wonder if it's a size limit. How much RAM do you have, though I'm not sure if this is critical to DSS ? I have similar problems with AS!2, though I know it's not the same beast. Come to think of it, is DSS 64-bit or 32?

Sorry, not much help!

Ian

PS. Do you get any error windows or does it just stop?

PPS. You could try using the Windows task manager to see how much memory is being used.

Thanks for the reply Ian.

DSS 64 bit and trying to load 240 frames

I have sorted it. My task manager said DSS is not responding.  I narrowed it down to 1 corrupt frame which did load last week, I deleted it and the rest uploaded fine. Not sure how it became corrupted. 

Now stacking 2 hours 20 minutes of cats eye neb, hopefully will get more details. 

Cheers

Nige.

[Nigel G]

Exposure times, now I've been told and read in books that ( example ) 120 x 30sec exposures is equal to say 6 x 10min exposures and that you need 2 hours of data for dso's. 

After 9 months of experimenting I have decided this is not strictly true. I believe longer exposures are much better than shorter for finding detail and light gathering but more frames is better for noise elimination .

I have taken images of Andromeda with as little as 15 minutes of data up to  2 hours of data with not much difference in the final images. The longer the exposure times the more detail captured regardless of total time.

My latest image, ngc 6543 , firstly I stacked about 45 minutes  of 30 to 45 second exposures,  second time around I stacked 2 hours 20 minutes of 30 to 45 second exposures,  both with darks and bias. Both lack the emissions of the cats eye, the second image has a touch more detail but was far better to work with in StarTools due to much less noise.  The cats eye is a difficult dso to capture and I'm not disappointed with the result although it lacks most of the nebula.

I think my test images of ngc 7000 showed the power of the longer exp times. Our greatest barrier is the exposure time limits of the Alt-Az mount.

Having said that there are plenty of dso's as we have seen in this thread where short exposures are all you need,  it's the fainter objects that require the long exp.

I have posted both images of ngc 6543 so they can be compared.   

First the 45 minute,  second the 2h 20 minutes. 

There are 2 other dso's within the image which are more apparent in the second image also there is a wiff more emissions  .  The second being far easier to process with far less noise reduction needed, I think due to there being 238 frames. Although noise is present in no 2 thats because I'm trying to extract emissions. 

I'm sure someone else with better processing skills could get more out of it if I knew how to upload a fits image. 

With 2 hours of 600s I believe the nebula would be greatly improved,  I might be wrong.

Cheers 

Nige.

 

 

[Filroden]

4 minutes ago, Nigel G said:

Exposure times, now I've been told and read in books that ( example ) 120 x 30sec exposures is equal to say 6 x 10min exposures and that you need 2 hours of data for dso's. 

After 9 months of experimenting I have decided this is not strictly true. I believe longer exposures are much better than shorter for finding detail and light gathering but more frames is better for noise elimination .

I have taken images of Andromeda with as little as 15 minutes of data up to  2 hours of data with not much difference in the final images. The longer the exposure times the more detail captured regardless of total time.

My latest image, ngc 6543 , firstly I stacked about 45 minutes  of 30 to 45 second exposures,  second time around I stacked 2 hours 20 minutes of 30 to 45 second exposures,  both with darks and bias. Both lack the emissions of the cats eye, the second image has a touch more detail but was far better to work with in StarTools due to much less noise.  The cats eye is a difficult dso to capture and I'm not disappointed with the result although it lacks most of the nebula.

I think my test images of ngc 7000 showed the power of the longer exp times. Our greatest barrier is the exposure time limits of the Alt-Az mount.

Having said that there are plenty of dso's as we have seen in this thread where short exposures are all you need,  it's the fainter objects that require the long exp.

I have posted both images of ngc 6543 so they can be compared.   

First the 45 minute,  second the 2h 20 minutes. 

There are 2 other dso's within the image which are more apparent in the second image also there is a wiff more emissions  .  The second being far easier to process with far less noise reduction needed, I think due to there being 238 frames. Although noise is present in no 2 thats because I'm trying to extract emissions. 

I'm sure someone else with better processing skills could get more out of it if I knew how to upload a fits image. 

With 2 hours of 600s I believe the nebula would be greatly improved,  I might be wrong.

Cheers 

Nige.

 

Interesting findings. I'd always thought that so long as you could exceed read noise then total exposure time was all that mattered (other than more subs for DSLRs is better to help reduce/remove banding).

if you can upload and share the fits file in something like OneDrive or Dropbox I can give it a go in PixInsight.

[SteveNickolls]

Hi Nige,

Thanks for posting this. I very much agree with your observation that the longer the individual exposures are, then other things being equal, the resulting image (S:N ratio) will be superior and is of course why people use EQ gear to get longer individual frames unfettered as they are by the effects of field rotation. With our Alt-Az equipment we are limited by the length of individual exposures possible by field rotation and the ability of our mounts to track and their (lack of) stability. Within this envelope (perhaps 60-70 seconds, sometimes much less) we hope to collect as many photons out of the decent % of frames from both the brightest through to the dimmest parts of a DSO and hope the signal can be pulled out of the noise enough. Local light pollution will add to the task. By repeating the exposures many times we hope to detect as much as we can the faintest part of the DSO which statistically will show up in lower numbers and in some exposures not at all if the object is very dim or the pollution very bad (that's my poor way of explaining the different ways signal and noise are treated mathematically). Joseph Ashley in, "Astro-photography on the Go" mentioned a survey of images posted an the Internet he made and concluded that 120 minutes of exposure was the right amount to get a decent image, so with Alt-Az gear that can amount to hundreds of light frames especially as not all frames are good enough for DSS to use in stacking. In practice I think the most frames I have ever taken on one object was 160.

Your images of NGC 6543 bear out the positive effect of taking more frames-reducing the noise more. As an aside I'm unsure right now about the pro's and con's of using dark frames at all. It's telling that you felt StarTools was happier processing the second batch of exposures with the greater number of exposures. As the object is a difficult (fainter) one you are adding to the difficulty of extracting a good image.

There is also some gain we can make by getting more adept with our processing software. Yesterday for example I reprocessed a cropped image of Kemble's Cascade taken originally in August and through better use of ST's got a much improved colour and dark background. However NGC 1502 (Mag 6.9) is much easier to image than your NGC 6543 (Mag 9.8) and in there hangs a tale.

Original-

Most recent reworking-

I'm really pleased you have posted this subject Nige for us to chew over :-)

Cheers,
Steve

 

[Nigel G]

I have just installed OneDrive and am attempting to upload the fts file to share.

I'm going to a dark spot next available clear night to see what difference dark skys will make to the cats eye. also want to try my luck on M1.

This weekend would be perfect but the weather man he say NO.

Steve, the second image is much sharper, looks like we will have plenty of time to think about this subject with the current forecast.

Nige

This fts file is taking a while to upload, I guess it will at 206 MB

Edited September 15, 2016 by Nigel G
add text

[Herzy]

4 hours ago, Nigel G said:

Exposure times, now I've been told and read in books that ( example ) 120 x 30sec exposures is equal to say 6 x 10min exposures and that you need 2 hours of data for dso's. 

After 9 months of experimenting I have decided this is not strictly true. I believe longer exposures are much better than shorter for finding detail and light gathering but more frames is better for noise elimination .

I have taken images of Andromeda with as little as 15 minutes of data up to  2 hours of data with not much difference in the final images. The longer the exposure times the more detail captured regardless of total time.

My latest image, ngc 6543 , firstly I stacked about 45 minutes  of 30 to 45 second exposures,  second time around I stacked 2 hours 20 minutes of 30 to 45 second exposures,  both with darks and bias. Both lack the emissions of the cats eye, the second image has a touch more detail but was far better to work with in StarTools due to much less noise.  The cats eye is a difficult dso to capture and I'm not disappointed with the result although it lacks most of the nebula.

I think my test images of ngc 7000 showed the power of the longer exp times. Our greatest barrier is the exposure time limits of the Alt-Az mount.

Having said that there are plenty of dso's as we have seen in this thread where short exposures are all you need,  it's the fainter objects that require the long exp.

I have posted both images of ngc 6543 so they can be compared.   

First the 45 minute,  second the 2h 20 minutes. 

There are 2 other dso's within the image which are more apparent in the second image also there is a wiff more emissions  .  The second being far easier to process with far less noise reduction needed, I think due to there being 238 frames. Although noise is present in no 2 thats because I'm trying to extract emissions. 

I'm sure someone else with better processing skills could get more out of it if I knew how to upload a fits image. 

With 2 hours of 600s I believe the nebula would be greatly improved,  I might be wrong.

Cheers 

Nige.

 

 

I remember posting a thread concerning one of my pictures and the lack of detail even though I integrated 4 hours. The object was very faint, but I thought my integration should've showed more detail than it did. 

I was informed by wimvb that my exposures (60s) weren't long enough to do the object justice. He said that the reasoning for longer exposures making a difference is the amount of signal vs the amount of noise.

If I recall correctly, he gave this example:

Consider a REALLY faint object. It gives you roughly 1 photon (1 signal electron) per minute. At the same time, your cameras read noise will contribute 1 electron. It doesn't matter how many frames you take, your signal will remain 1 because stacking doesn't boost signal, it justs removes noise. SNR = 1.

Now consider you are imaging that same object with 10 minute exposures. Your capturing 10 signal electrons/photons and 1 read noise electron. SNR = 10.

Longer exposures give you more signal, and more of them give you less noise. The combination of long exposures and lots of them give you a really good SNR.

So in that regard, an alt/az mount user is limited. They can always remove more noise, but can never boost their signal much without field rotation. That doesn't mean they can't produce wonderful images - they certainly can. They are just limited to bright-ish targets. 

Edited September 15, 2016 by Herzy