Markarian's Chain

[Rodd]

Hi folks--attached is a pic of Markarian'd Chain I captured with a Televue np101is and STT-8300.  Exposures were 10 5min for each RGB.  As you can see, I have a gradient issues that I was able to reduce in Pixinsight, though not completely remove. Where is the banding/gradients coming from?  Light pollution?  Hazy sky? Anybody know how to do this?  My second issue, is with the color--I have seen images of this object in RGB using the same equipment and exposure times that look way more colorful--stars are yellow, blue, white--mine looks almost like a monochromatic image-especially after I remove the gradients.  I used both channel combination RGB as well as LRGB using a synthetic Lum.  It always comes out like this.  Any ideas?

Thanks,

Rodd

[laudropb]

Very nice capture.

[Rodd]

4 hours ago, laudropb said:

Very nice capture.

Far from it--does that look like sky glow to you?

[ChrisLX200]

Crop about 20 pixels off each side before you start post-processing - some of the routines are picking up the artifacts around the edges.

ChrisH

[symesie04]

Hard to say but doesnt look like skyglow to me, that is usely more evened out surely? Looks like two lights sources entering the frame from below to me. Battery lights? Distant street lights just below frame? Unsure. Did you use darks? flats?

[Rodd]

3 hours ago, symesie04 said:

Hard to say but doesnt look like skyglow to me, that is usely more evened out surely? Looks like two lights sources entering the frame from below to me. Battery lights? Distant street lights just below frame? Unsure. Did you use darks? flats?

Yes--fully calibrated (300 bias, 33 darks and 30 flats).  How about the laptop screen?  can light get into the camera from exterior sources not in front of the scope?  Maybe I'll try covering the laptop?  

[ChrisLX200]

I also note you are using a self-guiding filterwheel? An OAG in other words... Make sure the pick-off prism is not set too far into the lightcone.

ChrisH

[jetstream]

Great capture, there are a pile of galaxies in this image.

[wimvb]

This definitely has the potential of a great image. It certainly is an interesting target. I'm surprised that PI couldn't cope with the gradient. As ChirsH suggested: start by cropping the edges slightly to get rid of stacking artefacts, before you do DBE. You can always try to do several runs with DBE.

There's also another trick, which gets rid of most banding issues: the CanonBanding script. It isn't meant for this kind of faults, but I found that it works on all kind of horizontal banding issues. Fast rotate the image so the bands run horizontally, run the script, and rotate back. Then do DBE.

 

 

Good luck,

[Rodd]

12 minutes ago, wimvb said:

This definitely has the potential of a great image. It certainly is an interesting target. I'm surprised that PI couldn't cope with the gradient. As ChirsH suggested: start by cropping the edges slightly to get rid of stacking artefacts, before you do DBE. You can always try to do several runs with DBE.

There's also another trick, which gets rid of most banding issues: the CanonBanding script. It isn't meant for this kind of faults, but I found that it works on all kind of horizontal banding issues. Fast rotate the image so the bands run horizontally, run the script, and rotate back. Then do DBE.

 

 

Good luck,

Thanks Wim--I did try the canon debanding script--many times.  It tended to spread the haze out--better, but only a little bit.  For some reason I can't use DBE--I keep getting memory read errors--so I have to use ABE.  That didn't work either.  I tried to use masks--but for some reason, when I make a mask and cover the galaxies--they still get eroded away when I do a Histogram transformation on the unprotected areas.  I can remove 90% of the gradients using about a dozen different things-but when I do, the galaxies get small and faint, and the color disappears--so the image looks mono (or pretty close).  

[wimvb]

One further observation: your (large scale) image is a whopping 135 Mpix (13400 * 10100). Could this (partly) be the reason for the memory error? I've never encountered memory issues with PI. My images are only 1/10 or yours (14.6 Mpix).

[MartinB]

Looks like a dodgy flat.  Have you had a look at your uncalibrated date., my guess is that if you stack and process the problem wont be there (you'll have other issues mind!)  Hang onto that data though because it looks great!

[steppenwolf]

Having just resolved an issue with my Flat calibration on images captured with my 8300 sensored CCD camera, I may be a little blinkered here but I'd be very tempted to follow Martin's suggestion and re-stack your lights without the Flat calibration to rule this out before going any further.

[abhoriel]

I removed the gradient manually in photoshop. I hope you don't mind, I just did it because I thought it was a fantastic image - I love galaxies. The image you uploaded was massive, and I'm limited to a 1MB upload so I had to resize the original and compress it to get it to upload.

I'm still a beginner at all this, but if its of any help at all, what I did was:

• I loaded it into photoshop
• Image -> Duplicate Image
• Click on the duplicate image
• Filter -> Noise -> Dust and Scratches (I set the radius to 500, this operation took a long time due to the size of the radius and size of the image)
• Click on the original image
• Image -> Apply Image. Set source to the duplicate image, set blending to subtract, set the offset to something like 30
• Did a limited stretch

If you want the full size one, the png is 158MB and the high quality compressed jpeg is 17.6MB.. Is there anywhere I can upload it or something?

[Rodd]

On 2/14/2016 at 04:54, abhoriel said:

I removed the gradient manually in photoshop. I hope you don't mind, I just did it because I thought it was a fantastic image - I love galaxies. The image you uploaded was massive, and I'm limited to a 1MB upload so I had to resize the original and compress it to get it to upload.

I'm still a beginner at all this, but if its of any help at all, what I did was:

• I loaded it into photoshop
• Image -> Duplicate Image
• Click on the duplicate image
• Filter -> Noise -> Dust and Scratches (I set the radius to 500, this operation took a long time due to the size of the radius and size of the image)
• Click on the original image
• Image -> Apply Image. Set source to the duplicate image, set blending to subtract, set the offset to something like 30
• Did a limited stretch

If you want the full size one, the png is 158MB and the high quality compressed jpeg is 17.6MB.. Is there anywhere I can upload it or something?

Wow--great job.  I am sure there is a way to do this in Pixinsight--but it is obviously beyond me.  I don't have PS.  I am finding the whole processing thing a bit frustrating.  At least I know my data CAN produce a good image.  I still think the background could use a bit of darkening.  In PI you are suppose to be able to put a mask over the galaxies and reduce the background--but I find it erodes the galaxies as well.  BY the way--the file is so large because I dithered.  The non dithered version is not as big.  Doesn';t dithering make your files big

 

Thanks,

 

Rodd

[abhoriel]

I don't know how pixinsight works (I need to have a play with a trial license!) so I can't really comment..

The background is too bright I agree.. I do my image processing on my desktop because its more powerful. It looked fine on there, but on my laptop and my phone it looks too bright. I think I need to make some adjustments to the graphics driver/monitor settings..

Your data is fantastic by the way.. really beautiful. Something I aspire to

[harry page]

Hi

This is very easy for pixinsight DBE to correct , just make sure your tolerance is large enough to make your samples not red ( about 2 for the jpeg)

 

can't do it well with a jpeg but you get the idea

 

Harry 

[Guest]

2 hours ago, harry page said:

Hi

This is very easy for pixinsight DBE to correct , just make sure your tolerance is large enough to make your samples not red ( about 2 for the jpeg)

 

can't do it well with a jpeg but you get the idea

 

Harry 

...and check out Harry's free tutorials on PI on his website*  - I did - its the only way I was able to use it :-)...and it gave me the confidence to experiment with PI and learn more.

David

*although I did donate and encourage others too as they're great tutorials

[Rodd]

4 hours ago, David_L said:

...and check out Harry's free tutorials on PI on his website*  - I did - its the only way I was able to use it :-)...and it gave me the confidence to experiment with PI and learn more.

David

*although I did donate and encourage others too as they're great tutorials

Nice--sometimes I look for a tutorial and I get a 404 error--maybe an old site.  Anyway--I am going to a PI seminar in a few weeks--7 days.  Maybe I'll learn something.

[Paolo De Salvatore]

Hi Rodd, I have the same problem with my stt8300 and filter wheel sbig, in all my light subs i have this refraction.

How you solved the problem?