Return to the Cloudy Nights Telescope Reviews home pageAstronomics discounts for Cloudy Nights members
· Get a Cloudy Nights T-Shirt · Submit a Review / Article

Click here if you are having trouble logging into the forums

Privacy Policy | Please read our Terms of Service | Signup and Troubleshooting FAQ | Problems? PM a Red or a Green Gu… uh, User

Astrophotography and Sketching >> DSLR & Digital Camera Astro Imaging & Processing

Pages: 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | (show all)
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: Learning PixInsight new [Re: pfile]
      #5886559 - 05/27/13 04:29 AM Attachment (15 downloads)

Thanks Rob.

Another thing I noticed is image (FIT) is decent looking upon integration (via STF) but looks quite granular/noisy after dynamic crop (via STF); is it supposed to be that way? Not sure why would there be so much disparity pre & post dynamic crop in how STF shows the same image? Has anyone notice this behavior? Thx


Post Extras: Print Post   Remind Me!   Notify Moderator  
jsines
sage
*****

Reged: 09/06/11

Loc: Berkley. Michigan
Re: Learning PixInsight new [Re: mmalik]
      #5886993 - 05/27/13 12:17 PM

Quote:

Thanks Rob.

Would same be true of 'Batch DeBayer Script', i.e., applicable to all types of DSLR RAWs (in-camera as well as out-camera NR RAWs)?




I don't think so. I think the bayer/mosaic pattern is camera specific. I think most Canons are RGGB, for example, but I don't think all DSLRs have the same pattern.


Post Extras: Print Post   Remind Me!   Notify Moderator  
jsines
sage
*****

Reged: 09/06/11

Loc: Berkley. Michigan
Re: Learning PixInsight Need help ! new [Re: mmalik]
      #5887003 - 05/27/13 12:26 PM

Quote:



I would suggest following 'complete' order:

1. RAW File Processing & Calibration [Linear data]
2. Alignment (Registration), Integration and Crop [Linear data]
3. DBE (Gradient Removal) [Linear data]
4. Color Calibration [Optional] [Linear data]
5. SCNR (Noise Reduction) [Linear data]
6. ATrous Wavelet Transform (Noise Reduction) [Linear data]

---Linear/NON-Linear demark---

7. Histogram Transformation
8. HDR Multiscale Transform
9. Local Histogram Equalization (Contrast)
10. ACDNR (Noise Reduction)
11. Curves Transformation (Saturation)






There really isn't a "complete" order for processing images in PixInsight. Once you get past the integration phase and have a stacked image, you need to identify specific problems and find solutions to those specific problems. Each image will have it's own problems. There's noise, and ways to remove it. There's a background gradient, and ways to remove it, etc., etc. PixInsight isn't a tool where you follow steps 1 through 10 and get a processed picture. It seems like you're looking for that type of program, and PI isn't it.

For example, you also have the option of ABE in addition to DBE to remove gradients. Inside ABE and DBE, you have the option of division or subtraction. For division and subtraction, you've also got the option of normalizing the image after. It's like a tree that branches out in different directions.

SCNR is usually done after stretching the data, usually right after Histogram Transform.
Background Neutralization is usually done before Color Calibration.
Some also do Atrous Wavelet Transform after removing the gradient and before Background Neutralization.
Local Histogram Equalization is usually the last step for me.
You're also missing the masks that are applied to the star cores when doing HDR, etc.


Post Extras: Print Post   Remind Me!   Notify Moderator  
jsines
sage
*****

Reged: 09/06/11

Loc: Berkley. Michigan
Re: Learning PixInsight new [Re: pfile]
      #5887012 - 05/27/13 12:30 PM

Quote:

those flows are okay but not sure why you'd use only part of BPP. might as well let it do the whole thing, except for ImageIntegration which should always be tuned by hand.




I haven't found a way to run cosmetic correction inside BPP using only CR2 files, since the Cosmetic Correction tool won't accept them when I try to create a process icon. I think the only way to use Cosmetic Correction when you're starting out with CR2 files is to stop at Calibrate Only. I may be wrong, though.

The Cosmetic Correction script is one more reason why you should not do ICNR, though. You don't get the option of removing hot/cold pixels without a master dark.


Post Extras: Print Post   Remind Me!   Notify Moderator  
jsines
sage
*****

Reged: 09/06/11

Loc: Berkley. Michigan
Re: Learning PixInsight new [Re: mmalik]
      #5887018 - 05/27/13 12:35 PM

Quote:

Thanks Rob.
Another thing I noticed is image (FIT) is decent looking upon integration (via STF) but looks quite granular/noisy after dynamic crop (via STF); is it supposed to be that way? Not sure why would there be so much disparity pre & post dynamic crop in how STF shows the same image? Has anyone notice this behavior? Thx





..because you don't understand how Screen Transfer Function works. You're removing all those black pixels during the crop, and that'll adjust what STF uses for an auto-stretch. You're actually seeing a "more realistic" image with the STF after the crop.


Post Extras: Print Post   Remind Me!   Notify Moderator  
pfile
Post Laureate


Reged: 06/14/09

Re: Learning PixInsight new [Re: jsines]
      #5887023 - 05/27/13 12:38 PM

Quote:

Quote:

those flows are okay but not sure why you'd use only part of BPP. might as well let it do the whole thing, except for ImageIntegration which should always be tuned by hand.




I haven't found a way to run cosmetic correction inside BPP using only CR2 files, since the Cosmetic Correction tool won't accept them when I try to create a process icon. I think the only way to use Cosmetic Correction when you're starting out with CR2 files is to stop at Calibrate Only. I may be wrong, though.

However, Cosmetic Correction is also one more reason why you should not do ICNR. You don't get the option of removing hot/cold pixels without a master dark.




good point, i was not aware of that. i should probably just never say anything about BPP since i don't use it. i'm okay with doing it all by hand - i have a lot of saved process icons that help.

fits has this 'incremental reading' feature that other image formats don't support. i think this is why II requires fits files - it works on the images in slices and that requires incremental reading. maybe CC works the same, not sure.


Post Extras: Print Post   Remind Me!   Notify Moderator  
pfile
Post Laureate


Reged: 06/14/09

Re: Learning PixInsight new [Re: mmalik]
      #5887027 - 05/27/13 12:40 PM

Quote:

Thanks Rob.

Another thing I noticed is image (FIT) is decent looking upon integration (via STF) but looks quite granular/noisy after dynamic crop (via STF); is it supposed to be that way? Not sure why would there be so much disparity pre & post dynamic crop in how STF shows the same image? Has anyone notice this behavior? Thx




this is normal. STF uses the statistics of the image to compute the auto-stretch. if you have a bunch of black pixels around the edge, that affects the stats and you get a different STF computation.

the noise is there in your first image, you just don't see it because it is not stretched as hard. if you apply the STF that was computed on the cropped image to the uncropped image, you'll see the same thing.


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: Learning PixInsight new [Re: jsines]
      #5887376 - 05/27/13 03:58 PM

Quote:

Quote:

Would same be true of 'Batch DeBayer Script', i.e., applicable to all types of DSLR RAWs (in-camera as well as out-camera NR RAWs)?




I don't think so. I think the Bayer/mosaic pattern is camera specific. I think most Canons are RGGB, for example, but I don't think all DSLRs have the same pattern.




If you re-read, my question was different. I was asking if Batch DeBayer script can be or should be used for both, ICNR as well as for OCNR image acquisition scenarios. (Not asking about Bayer pattern per se)


Note: On a side note I would like to confirm though if Canon 60Da is RGGB Bayer pattern?


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: Learning PixInsight new [Re: jsines]
      #5887411 - 05/27/13 04:20 PM

Quote:

Quote:



I would suggest following 'complete' order:

1. RAW File Processing & Calibration [Linear data]
2. Alignment (Registration), Integration and Crop [Linear data]
3. DBE (Gradient Removal) [Linear data]
4. Color Calibration [Optional] [Linear data]
5. SCNR (Noise Reduction) [Linear data]
6. ATrous Wavelet Transform (Noise Reduction) [Linear data]

---Linear/NON-Linear demark---

7. Histogram Transformation
8. HDR Multiscale Transform
9. Local Histogram Equalization (Contrast)
10. ACDNR (Noise Reduction)
11. Curves Transformation (Saturation)






There really isn't a "complete" order for processing images in PixInsight. Once you get past the integration phase and have a stacked image, you need to identify specific problems and find solutions to those specific problems. Each image will have it's own problems. There's noise, and ways to remove it. There's a background gradient, and ways to remove it, etc., etc. PixInsight isn't a tool where you follow steps 1 through 10 and get a processed picture. It seems like you're looking for that type of program, and PI isn't it.

For example, you also have the option of ABE in addition to DBE to remove gradients. Inside ABE and DBE, you have the option of division or subtraction. For division and subtraction, you've also got the option of normalizing the image after. It's like a tree that branches out in different directions.

SCNR is usually done after stretching the data, usually right after Histogram Transform.
Background Neutralization is usually done before Color Calibration.
Some also do Atrous Wavelet Transform after removing the gradient and before Background Neutralization.
Local Histogram Equalization is usually the last step for me.
You're also missing the masks that are applied to the star cores when doing HDR, etc.




You are killing me; where were you when I asked about all this 10 pages back at the start of this thread .

For learners like me and others this was the core point; what basic steps to perform and in which order. This was the order we all agree to including Harry. Read this... to refresh your memory . I am not mad but little upset when experts like you take their leisurely time to chime in when things have progressed this far. Sorry if I am sounding harsh but I am not; I just would like folks to chime in when help is needed the most and from ordering perspective that was at the start of this thread. For the life of me, put it in 1.2.3... order what you think is the correct BASIC order instead of verbalizing what's wrong or right. Remember such order would be for "learners" and "starters", not "experts". Regards


Post Extras: Print Post   Remind Me!   Notify Moderator  
jsines
sage
*****

Reged: 09/06/11

Loc: Berkley. Michigan
Re: Learning PixInsight Need help ! new [Re: mmalik]
      #5889066 - 05/28/13 02:16 PM

Quote:


You are killing me; where were you when I asked about all this 10 pages back at the start of this thread .

For learners like me and others this was the core point; what basic steps to perform and in which order. This was the order we all agree to including Harry. Read this... to refresh your memory . I am not mad but little upset when experts like you take their leisurely time to chime in when things have progressed this far. Sorry if I am sounding harsh but I am not; I just would like folks to chime in when help is needed the most and from ordering perspective that was at the start of this thread. For the life of me, put it in 1.2.3... order what you think is the correct BASIC order instead of verbalizing what's wrong or right. Remember such order would be for "learners" and "starters", not "experts". Regards




Harry said in that comment on page 1 - "...depends on your image".

I said this on 3/28/13, 2 months ago (on page 7)-

Quote:


You're starting to get into the stage of PixInsight where an image can branch out in different directions. You can do some steps on an image and get one result, and you can do different steps on the same image and get different results. It's not really a 10 step process where all 10 steps are followed in a certain way. One of the benefits of PI is that you can have multiple copies of the same image open, even in different windows (the 4 at the bottom), and then try different things on the same image. Harry does this in his video on HDRWavelets, and I do this all the time for different steps.




It depends on the image. You've already been given a "basic" order of processing on page 1, and you've been referred to Harry's tutorials, and I've provided you a sample work-flow, using your own data (pages 7-8). People have used your data in this thread to create a final image using different steps. But the steps you take depends on the image and the specific problem you want to solve. I'm not familiar with any post-processing program where you perform the same steps 1-10 in order on different images and you get final processed images.

Quote:

I am not mad but little upset when experts like you take their leisurely time to chime in when things have progressed this far.




In no way do I consider myself an expert in PixInsight. I bought the program only about 6-8 months ago, and I still consider myself a noob. The users at the PI forum are the experts. I try to help here when I can, and I admit when I don't know if I'm right about something.


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: Learning PixInsight new [Re: jsines]
      #5889880 - 05/28/13 10:14 PM

Quote:

In no way do I consider myself an expert in PixInsight. I bought the program only about 6-8 months ago, and I still consider myself a noob.




Understood; as the subject of this thread says "Learning PixInsight", let's all learn then with an open mind and leave the assertions for the experts .


Another request I have for all, not just jsines, that let's stop the mantra "every image is different", "depends on your image" etc.; we all know that. Plus we all also understand there are multiple ways of doing things in PixInsight. There are basics that we all learners need to know/learn before we get to the next stage.


Everyone, with that said, is following general flow acceptable? Additions/modifications welcome!


1. DSLR_RAW Format Preferences [ONE time/FIRST time setting] (applicable to both, ICNR/OCNR)
2. Calibration [Linear data] (applicable to OCNR only)

•Batch Preprocessing Script ('Calibrate only' option; masters to be created...)
OR
•Image Calibration (masters available...)

3. Batch Debayer Script (applicable to both, ICNR/OCNR)
4. Star Alignment (Registration), Integration and Crop [Linear data]
5. DBE (Gradient Removal) [Linear data]
6. Color Calibration [Linear data]
7. ATrous Wavelet Transform (Noise Reduction) [Optional?] [Linear data]

---Linear/NON-Linear demark---

8. Histogram Transformation
9. SCNR (Noise Reduction) [Optional?]
10. HDR Multiscale Transform [IMPORTANT]
11. Local Histogram Equalization (Contrast) [Optional?]
12. ACDNR (Noise Reduction)
13. Curves Transformation (Saturation)

Edited by mmalik (05/29/13 02:54 AM)


Post Extras: Print Post   Remind Me!   Notify Moderator  
pfile
Post Laureate


Reged: 06/14/09

Re: Learning PixInsight new [Re: mmalik]
      #5890005 - 05/28/13 11:29 PM

before #6 you must do BackgroundNeutralization or else the results of #6 will be garbage...

on #6 it's really important to set the background and foreground thresholds properly - turn on "show white reference" and "show background reference" while you are tuning that. if you get two white squares (or two black squares) then your thresholds are set wrong. same goes for BackgroundNeutralization; you need to make sure you are only picking up background pixels.

for galaxies the best thing to do for the white reference is to draw a preview over the center of the galaxy and then tell CC to use a region of interest, taking the coordinates from the preview you just created. be sure to turn off structure detection in this situation.

for nebula and other non-galaxy DSOs, you should turn on structure detection and don't use a preview. again tune the thresholds so that the white reference is a bunch of white dots on a black background - you are isolating the stars because on average, the color of all the stars together will be white.

in both situations, for the background reference, draw a preview over a part of the background and take the region of interest from that preview. again tune the background thresholds so that you have a white field with black stars - you want to exclude the stars from the background reference.


color calibration is pretty much mandatory if you are dealing with RGB images. for narrowband the color is kind of arbitrary, so it's not important.

rob

p.s. the reason to use regions of interest rather than the previews directly is to make the CC process re-usable. if you use the previews directly then the CC process will not work on another image unless you define exactly the same previews on that image. this may not make sense at first, but say you collect 5h of data on a subject and then process the image. later you collect 5 more hours and you have a new image. if you saved a process icon for your BN and CC you can just use them again on the new image without worrying about replicating the previews.


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: Learning PixInsight new [Re: pfile]
      #5890226 - 05/29/13 06:00 AM

Thanks Rob, good info; what do you think if this general flow?


1. DSLR_RAW Format Preferences [ONE time/FIRST time setting] (applicable to both, ICNR/OCNR)
2. Calibration [Linear data] (applicable to OCNR only)

•Batch Preprocessing Script ('Calibrate only' option; masters to be created...)
OR
•Image Calibration (masters available...)

3. Batch Debayer Script (applicable to both, ICNR/OCNR)
4. Star Alignment (Registration), Integration and Crop [Linear data]
5. DBE (Gradient Removal) [Linear data]
6. Background Neutralization [Linear data]
7. Color Calibration (for RGB images) [Linear data]
8. ATrous Wavelet Transform (Noise Reduction) [Optional?] [Linear data]

---Linear/NON-Linear demark---

9. Histogram Transformation
10. SCNR (Noise Reduction) [Optional?]
11. HDR Multiscale Transform [IMPORTANT]
12. Local Histogram Equalization (Contrast) [Optional?]
13. ACDNR (Noise Reduction)
14. Curves Transformation (Saturation)


Post Extras: Print Post   Remind Me!   Notify Moderator  
jsines
sage
*****

Reged: 09/06/11

Loc: Berkley. Michigan
Re: Learning PixInsight new [Re: mmalik]
      #5890558 - 05/29/13 11:32 AM

Quote:

Thanks Rob, good info; what do you think if this general flow?





I think you're really just re-inventing the wheel at this point, since Harry's video tutorials are considered the beginner's general work flow. Everyone who is learning PI gets referred to them, and I think they do a great job of explaining all the steps you're trying to re-create.

For beginners learning PixInsight, I'd skip the Calibration, Batch Debayering, and Star Alignment steps and just use the BatchPreProcessing script. This means you'll need to take separate darks, but the creators of PixInsight ("the experts") say you should do this anyway. The BatchPreProcessing script is also covered in one of Harry's videos.


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: Learning PixInsight new [Re: jsines]
      #5894205 - 05/31/13 12:35 PM

Harry does SCNR right after color calibration in the video; that's how I had it all along and also how Ajay had suggested at the start and Harry seemed ok with it until you and some folks suggested otherwise. Wouldn't SCNR make more sense after color calibration while one is fixing color in the linear mode? Thx

Post Extras: Print Post   Remind Me!   Notify Moderator  
harry page 1
sage


Reged: 07/25/09

Re: Learning PixInsight new [Re: mmalik]
      #5894435 - 05/31/13 02:39 PM

Hi
indeed I do use scnr most of the time at the linear stage , but it also works very well when the data is stretched

So use as you please

Regards Harry


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: Learning PixInsight new [Re: harry page 1]
      #5896792 - 06/01/13 09:17 PM

Thanks Harry; it has been a while we all heard from you. Please do provide your insights once in while, they are greatly appreciated. Regards

Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: Learning PixInsight new [Re: harry page 1]
      #5955548 - 07/05/13 02:40 AM

I notice PixInsight goes through quite a rapid succession of revisions or updates; was wondering what happens once one purchases a particular version; would they be eligible for free updates and for how long? Is one locked into the version they buy like other software? In short, how are minor or major updates handled after one purchases a version of PixInsight?

Post Extras: Print Post   Remind Me!   Notify Moderator  
harry page 1
sage


Reged: 07/25/09

Re: Learning PixInsight new [Re: mmalik]
      #5956016 - 07/05/13 12:18 PM

Hi
I have had Pi for 4 years now and received all updates for free
There is a bit here from PI as we get closer to ver 2 of pixinsight which might be 2014 -2015

Quote from the FAQ page

When we release version 2.0, the price of a commercial 2.x license for existing users will be directly proportional to the time of use of their 1.x licenses.

For example, a user who purchased his or her license a few months before the 2.0 release date will have access to the new version at no cost. A user who purchased a license a few years before the 2.0 release will have to pay the full 2.x license price. These examples should not be taken literally; the exact terms of our upgrade policy will be determined when appropriate, but you get the idea.

Regards

Harry


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: Learning PixInsight new [Re: harry page 1]
      #6174051 - 11/03/13 04:27 PM Attachment (5 downloads)

I get following error when re-activating PixInsight; I had to reinstall the OS and PixInight but original activation code is not working that worked before. Any ideas how to re-activate PixInsight after a fresh re-install of the OS/PixInsight? Note: I have a purchased license and Yes, I DID forget to backup the original license file after I activated first time. Regards

Edited by mmalik (11/03/13 04:43 PM)


Post Extras: Print Post   Remind Me!   Notify Moderator  
Pages: 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | (show all)


Extra information
12 registered and 33 anonymous users are browsing this forum.

Moderator:  Dave M, WOBentley, tecmage 

Print Thread

Forum Permissions
      You cannot start new topics
      You cannot reply to topics
      HTML is disabled
      UBBCode is enabled


Thread views: 11188

Jump to

CN Forums Home


Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics