Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

QHY183M & Flats = Frustration

  • Please log in to reply
94 replies to this topic

#1 terry59

terry59

    Cosmos

  • *****
  • topic starter
  • Posts: 9169
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 14 September 2019 - 05:24 PM

I've been struggling with flats until today when I decided to forego using them. What a huge difference. I once again have the issue with calibrating amp glow out but I had to use the dark flats and I think I need to shoot some regular bias frames now. 

 

Here is the result...cropped to remove the amp glow area, lightly processed in PI and PS

 

Only three months in the making

 

smile.gif

 

GM-8/SV70T/QHY183M/Baader 3.5nm Ha filter

Attached Thumbnails

  • trunk.jpg

Edited by terry59, 14 September 2019 - 05:27 PM.

  • Jim Waters, elmiko, artem2 and 5 others like this

#2 Stelios

Stelios

    Cosmos

  • *****
  • Moderators
  • Posts: 7731
  • Joined: 04 Oct 2003
  • Loc: West Hills, CA

Posted 14 September 2019 - 05:59 PM

Very nice image, but can you elaborate a bit on the issues with flats and dark flats? 

 

I was *just* going to order an 183MM when I saw this post. 

 

I have read before that for some reason you can't use flats + superbias with the 183MM (you definitely *can* with the 1600MM), but I thought that flats + dark flats was "advertised" (on CN) as a perfect solution. 



#3 Hondo

Hondo

    Mariner 2

  • -----
  • Posts: 298
  • Joined: 28 Oct 2017
  • Loc: NB, Canada

Posted 14 September 2019 - 06:03 PM

I was *just* going to order an 183MM when I saw this post. 

I ordered the QHY183M today.  


  • bmhjr likes this

#4 kingjamez

kingjamez

    Vanguard

  • *****
  • Posts: 2069
  • Joined: 03 Oct 2006
  • Loc: Fairfax, VA

Posted 14 September 2019 - 06:08 PM

What are your challenges with flats and darks?

 

I have a 183mm and an 183mc and have no problems calibrating with flats, dark flats, and darks. I got used to not using bias frames so haven't tried yet, but right now I don't see a reason to go back to bias frames. 

 

Can you describe your process of making darks / flats / dark flats? How do you apply them?

 

-Jim



#5 Stelios

Stelios

    Cosmos

  • *****
  • Moderators
  • Posts: 7731
  • Joined: 04 Oct 2003
  • Loc: West Hills, CA

Posted 14 September 2019 - 06:23 PM

What are your challenges with flats and darks?

 

I have a 183mm and an 183mc and have no problems calibrating with flats, dark flats, and darks. I got used to not using bias frames so haven't tried yet, but right now I don't see a reason to go back to bias frames. 

 

Can you describe your process of making darks / flats / dark flats? How do you apply them?

 

-Jim

Perhaps, since whatever you are doing works waytogo.gif , you could help us by telling us what *you* are doing to acquire them. (Flats and dark flats, I mean--what can go wrong with darks?)



#6 kingjamez

kingjamez

    Vanguard

  • *****
  • Posts: 2069
  • Joined: 03 Oct 2006
  • Loc: Fairfax, VA

Posted 14 September 2019 - 06:48 PM

Perhaps, since whatever you are doing works waytogo.gif , you could help us by telling us what *you* are doing to acquire them. (Flats and dark flats, I mean--what can go wrong with darks?)

That’s just it, I don’t do anything special at all.
But here is what I do anyway.

I take darks at the same gain, offset, and temperature.

I take flats and dark flats using the NINA flats wizard. The wizard takes images at 0 gain which is fine for flats. I take dark flats right after flats. In general my flats take 1-5 seconds and I vary the brightness of my flat panel (LED tracing pad from amazon) by adjusting its brightness or adding layers of fabric/translucent plastic if it needs to be dimmer.

I use APP for calibration.

-Jim

Edited by kingjamez, 14 September 2019 - 06:50 PM.

  • OldManSky likes this

#7 terry59

terry59

    Cosmos

  • *****
  • topic starter
  • Posts: 9169
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 14 September 2019 - 07:01 PM

Very nice image, but can you elaborate a bit on the issues with flats and dark flats? 

 

I was *just* going to order an 183MM when I saw this post. 

 

I have read before that for some reason you can't use flats + superbias with the 183MM (you definitely *can* with the 1600MM), but I thought that flats + dark flats was "advertised" (on CN) as a perfect solution. 

Hi Stelios....First, I am really liking the camera. A lot of my problem has been bad weather and having the wrong camera driver installed. So far each of my attemps at flats have not worked. I can upload the latest example later if you would like to see it. I will get things sorted out and I am planning to put this on the FSQ-106 before too long


  • Swordfishy likes this

#8 terry59

terry59

    Cosmos

  • *****
  • topic starter
  • Posts: 9169
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 14 September 2019 - 07:10 PM

What are your challenges with flats and darks?

 

I have a 183mm and an 183mc and have no problems calibrating with flats, dark flats, and darks. I got used to not using bias frames so haven't tried yet, but right now I don't see a reason to go back to bias frames. 

 

Can you describe your process of making darks / flats / dark flats? How do you apply them?

 

-Jim

 

I take the frames, create masters and calibrate with those. I'm pretty sure my flats are mismatched but I haven't found the right exposure time/diffusion. Gain 11 Offset 30


Edited by terry59, 14 September 2019 - 07:10 PM.


#9 kingjamez

kingjamez

    Vanguard

  • *****
  • Posts: 2069
  • Joined: 03 Oct 2006
  • Loc: Fairfax, VA

Posted 14 September 2019 - 07:23 PM

I take the frames, create masters and calibrate with those. I'm pretty sure my flats are mismatched but I haven't found the right exposure time/diffusion. Gain 11 Offset 30

Go deeper on how you take the frames.

Offset 30 is really high for the 183mm and gain 11 is an unusual choice, but I’m not sure that would cause calibration issues. I use Gain 56 (half unity) and 111(unity) both with offset 10.

Flats don’t need to match lights as long as the match the dark flats. How are you taking flats? The super easy way is just tightened T-shirt at dusk. I shoot for 30,000ADU. How long are your flats and are you being very careful to match dark flats?

Just to be certain:
Your darks should be time/temp/gain/offset matched to your lights. EXACTLY.
Your flats should be time/temp/gain/offset matched to your dark flats.
Your flats should shoot for roughly 30,000 ADU with quite a bit of slop as long as it’s consistent.

Every time I have a calibration issue, it always goes back to something not matching. Generally, it’s an accident that I chose the wrong files from my library of master calibration frames.

What is your process for creating masters?

-Jim

Edited by kingjamez, 14 September 2019 - 07:27 PM.

  • elmiko likes this

#10 terry59

terry59

    Cosmos

  • *****
  • topic starter
  • Posts: 9169
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 14 September 2019 - 07:30 PM

Go deeper on how you take the frames.

Offset 30 is really high for the 183mm and gain 11 is an unusual choice, but I’m not sure that would cause calibration issues. I use Gain 53 (half unity) and 111(unity) both with offset 10.

Flats don’t need to match lights as long as the match the dark flats. How are you taking flats? The super easy way is just tightened T-shirt at dusk. I shoot for 30,000ADU. How long are your flats and are you being very careful to match dark flats?

Just to be certain:
Your darks should be time/temp/gain/offset matched to your lights. EXACTLY.
Your flats should be time/temp/gain/offset matched to your dark flats.
Your flats should shoot for roughly 30,000 ADU with quite a bit of slop as long as it’s consistent.

Every time I have a calibration issue, it always goes back to something not matching. Generally, it’s an accident that I chose the wrong files from my library of master calibration frames.

What is your process for creating masters?

-Jim

This is a QHY...gain 11 equals gain 111 for you

 

Yes

Yes

This I've been experimenting with. I think the last set of flats was at 30k but not 100% without seeing one

 

Edit: I use an Alnitak Flatman

 

E2: With the 3.5 nm Ha filter it takes a while to reach 30k. I know the current flats (and dark flats) are 50 seconds and I used a t-shirt as a diffuser


Edited by terry59, 14 September 2019 - 07:42 PM.


#11 kingjamez

kingjamez

    Vanguard

  • *****
  • Posts: 2069
  • Joined: 03 Oct 2006
  • Loc: Fairfax, VA

Posted 14 September 2019 - 07:46 PM

Right... 10x on the gain conversion. Good call.

Can you find a place to upload so we can take look at a light, dark, flat dark and flat.

I’ve never had to go 50 seconds on a flat. I’m at F/2 and F/4 and have never had to go past 5 seconds. However for the purpose of a flat as long as the DF matches length /shouldn’t/ matter.

-Jim

#12 terry59

terry59

    Cosmos

  • *****
  • topic starter
  • Posts: 9169
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 14 September 2019 - 07:58 PM

Right... 10x on the gain conversion. Good call.

Can you find a place to upload so we can take look at a light, dark, flat dark and flat.

I’ve never had to go 50 seconds on a flat. I’m at F/2 and F/4 and have never had to go past 5 seconds. However for the purpose of a flat as long as the DF matches length /shouldn’t/ matter.

-Jim

I will share one of each via OneDrive and post a link. It will be a while before I have access to the files though

 

I appreciate any ideas

 

smile.gif 



#13 Stelios

Stelios

    Cosmos

  • *****
  • Moderators
  • Posts: 7731
  • Joined: 04 Oct 2003
  • Loc: West Hills, CA

Posted 14 September 2019 - 08:33 PM

What about what I'd heard about flats needing to be at least 2 seconds in length? 

 

With my Spike-A-Flat at 50%, I use ranges from 0.01" for Lum to 1.2" for Sii. That brings the ADU to around 28000-31000 (for the ASI1600MM-C)

 

I do hope this will work, although I'm quite adaptable--but get cranky if I need to adapt too much smile.gif

 

Terry, I just ordered it, since you liked it despite your problems. Let's see how long free shipping takes smile.gif


  • kingjamez and terry59 like this

#14 OldManSky

OldManSky

    Surveyor 1

  • *****
  • Posts: 1969
  • Joined: 03 Jan 2019
  • Loc: Valley Center, CA USA

Posted 14 September 2019 - 09:11 PM

Stelios, it will work fine.

I gave the ZWO version. Never had any problems calibrating.

I do t-shirt flats at dusk or dawn, and let NINA figure the exposure. Sometimes (often) that’s less than 2 seconds. No problem.

Interesting thing...I thought the “don’t take flats under 2 seconds” thing was applicable to the 1600mm...:)


  • kingjamez likes this

#15 kingjamez

kingjamez

    Vanguard

  • *****
  • Posts: 2069
  • Joined: 03 Oct 2006
  • Loc: Fairfax, VA

Posted 14 September 2019 - 09:26 PM

 

Interesting thing...I thought the “don’t take flats under 2 seconds” thing was applicable to the 1600mm...smile.gif

And I thought it was "don't take flats under 0.2 seconds" for the ASI1600mm....



#16 terry59

terry59

    Cosmos

  • *****
  • topic starter
  • Posts: 9169
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 14 September 2019 - 11:00 PM

I will share one of each via OneDrive and post a link. It will be a while before I have access to the files though

 

I appreciate any ideas

 

smile.gif

Here is the link

 

thanks

 

https://1drv.ms/u/s!...C_yh52?e=UAm3BA



#17 jdupton

jdupton

    Surveyor 1

  • *****
  • Posts: 1950
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 14 September 2019 - 11:07 PM

Terry,

 

   If you are in a mood to experiment, I can suggest something new to try for your Flat Frame Calibration issue.

 

   Reprocess your Master Flat Frame as follows as an experiment.

  • Integrate your Flat-Darks as usual to make a Master Flat-Dark Frame
  • Calibrate your Flat Frames with the Master Flat-Dark (again as usual)
  • Integrate the Calibrated Flat Frames using the following unorthodox settings in PixInsight
    *  Under Image Integration, set:
       -- Combination = Average,
       -- Normalization = None, (You would normally use Multiplicative for Flats -- Don't for this experiment)
       -- Weights = Don't Care (All = 1)
    * Under Pixel Rejection (1), set:
       -- Rejection = Windorized Sigma Clipping
       -- Normalization = Equalize Fluxes
     
  • Try calibrating your Light Frames with this new experimental Master Flat Frame and your normal Master Dark Frame (The one that matches the Light Exposures)

   This modified process is what I am using until the next maintenance release of PixInsight. I hit a problem where my Flats from my ASI294MC-Pro would not work well at all and gave very odd results. I noticed that it was very dependent on the two things -- 1) the overall level of the Flat Frames (above 30,000 ADU was more prone to the problem), and 2) the number of and spread of hot pixels that were present in the Flat Frames.

 

   After Juan C. Investigated, he found a problem in the PixInsight code for Image Integration when Multiplicative Normalization is used (as for Flats). It was a very rare combination of the relative presence of high brightness pixels in the Reference Frame compared to other Frames. In my case, many frames used as the Reference produced bad Master Flats while others worked. Juan promised a fix in the next release of PI. Until then, I am using the above settings for PI Master Flat Integration. (If you do not use Multiplicative Normalization, the issue cannot arise.) (Note that Juan did not suggest that. It is simply my approach to avoiding the problem as it can show up in lesser degrees compared to the insidious case I had when I discovered the bug. I didn't trust myself to recognize when the bug was being activated so just avoid it completely for now.)

 

   This may not be the problem you are having, but it could be worth the time to try rebuilding your Master Flat with the above settings to see if it helps any or not. It's probably not the cause of your issues with Flats but is easy to check.

 

   As I mentioned, I am using the No Normalization setting for Master Flats now and also reducing the brightness of my Flats to the ~23,000 ADU range just to be safer. I will revert to more usual settings once Juan works his magic in the next PixInsight release.

 

 

John


  • H-Alfa likes this

#18 Stelios

Stelios

    Cosmos

  • *****
  • Moderators
  • Posts: 7731
  • Joined: 04 Oct 2003
  • Loc: West Hills, CA

Posted 15 September 2019 - 01:31 AM

Terry,

 

   If you are in a mood to experiment, I can suggest something new to try for your Flat Frame Calibration issue.

 

   Reprocess your Master Flat Frame as follows as an experiment.

  • Integrate your Flat-Darks as usual to make a Master Flat-Dark Frame
  • Calibrate your Flat Frames with the Master Flat-Dark (again as usual)
  • Integrate the Calibrated Flat Frames using the following unorthodox settings in PixInsight
    *  Under Image Integration, set:
       -- Combination = Average,
       -- Normalization = None, (You would normally use Multiplicative for Flats -- Don't for this experiment)
       -- Weights = Don't Care (All = 1)
    * Under Pixel Rejection (1), set:
       -- Rejection = Windorized Sigma Clipping
       -- Normalization = Equalize Fluxes
     
  • Try calibrating your Light Frames with this new experimental Master Flat Frame and your normal Master Dark Frame (The one that matches the Light Exposures)

   This modified process is what I am using until the next maintenance release of PixInsight. I hit a problem where my Flats from my ASI294MC-Pro would not work well at all and gave very odd results. I noticed that it was very dependent on the two things -- 1) the overall level of the Flat Frames (above 30,000 ADU was more prone to the problem), and 2) the number of and spread of hot pixels that were present in the Flat Frames.

 

   After Juan C. Investigated, he found a problem in the PixInsight code for Image Integration when Multiplicative Normalization is used (as for Flats). It was a very rare combination of the relative presence of high brightness pixels in the Reference Frame compared to other Frames. In my case, many frames used as the Reference produced bad Master Flats while others worked. Juan promised a fix in the next release of PI. Until then, I am using the above settings for PI Master Flat Integration. (If you do not use Multiplicative Normalization, the issue cannot arise.) (Note that Juan did not suggest that. It is simply my approach to avoiding the problem as it can show up in lesser degrees compared to the insidious case I had when I discovered the bug. I didn't trust myself to recognize when the bug was being activated so just avoid it completely for now.)

 

   This may not be the problem you are having, but it could be worth the time to try rebuilding your Master Flat with the above settings to see if it helps any or not. It's probably not the cause of your issues with Flats but is easy to check.

 

   As I mentioned, I am using the No Normalization setting for Master Flats now and also reducing the brightness of my Flats to the ~23,000 ADU range just to be safer. I will revert to more usual settings once Juan works his magic in the next PixInsight release.

 

 

John

Great info! 

I have, in the past, captured flats with ADU's in the 18K range with no ill effects, although now I aim for 28-33K. I will try to capture 2 sets, one at 50% brightness and another at 25%, and generate different master flats and see which works better or if it makes any difference.

 

Of course, I need to get the camera first :)



#19 terry59

terry59

    Cosmos

  • *****
  • topic starter
  • Posts: 9169
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 15 September 2019 - 06:20 AM

Terry,

 

   If you are in a mood to experiment, I can suggest something new to try for your Flat Frame Calibration issue.

 

   Reprocess your Master Flat Frame as follows as an experiment.

  • Integrate your Flat-Darks as usual to make a Master Flat-Dark Frame
  • Calibrate your Flat Frames with the Master Flat-Dark (again as usual)
  • Integrate the Calibrated Flat Frames using the following unorthodox settings in PixInsight
    *  Under Image Integration, set:
       -- Combination = Average,
       -- Normalization = None, (You would normally use Multiplicative for Flats -- Don't for this experiment)
       -- Weights = Don't Care (All = 1)
    * Under Pixel Rejection (1), set:
       -- Rejection = Windorized Sigma Clipping
       -- Normalization = Equalize Fluxes
     
  • Try calibrating your Light Frames with this new experimental Master Flat Frame and your normal Master Dark Frame (The one that matches the Light Exposures)

   This modified process is what I am using until the next maintenance release of PixInsight. I hit a problem where my Flats from my ASI294MC-Pro would not work well at all and gave very odd results. I noticed that it was very dependent on the two things -- 1) the overall level of the Flat Frames (above 30,000 ADU was more prone to the problem), and 2) the number of and spread of hot pixels that were present in the Flat Frames.

 

   After Juan C. Investigated, he found a problem in the PixInsight code for Image Integration when Multiplicative Normalization is used (as for Flats). It was a very rare combination of the relative presence of high brightness pixels in the Reference Frame compared to other Frames. In my case, many frames used as the Reference produced bad Master Flats while others worked. Juan promised a fix in the next release of PI. Until then, I am using the above settings for PI Master Flat Integration. (If you do not use Multiplicative Normalization, the issue cannot arise.) (Note that Juan did not suggest that. It is simply my approach to avoiding the problem as it can show up in lesser degrees compared to the insidious case I had when I discovered the bug. I didn't trust myself to recognize when the bug was being activated so just avoid it completely for now.)

 

   This may not be the problem you are having, but it could be worth the time to try rebuilding your Master Flat with the above settings to see if it helps any or not. It's probably not the cause of your issues with Flats but is easy to check.

 

   As I mentioned, I am using the No Normalization setting for Master Flats now and also reducing the brightness of my Flats to the ~23,000 ADU range just to be safer. I will revert to more usual settings once Juan works his magic in the next PixInsight release.

 

 

John

Thank you John.....I will give this a go. If I don't get the opportunity this morning it will likely be tomorrow as the wife and I will be occupied this afternoon and evening

 

waytogo.gif 



#20 kingjamez

kingjamez

    Vanguard

  • *****
  • Posts: 2069
  • Joined: 03 Oct 2006
  • Loc: Fairfax, VA

Posted 15 September 2019 - 08:08 AM

Here is the link

 

thanks

 

https://1drv.ms/u/s!...C_yh52?e=UAm3BA

Hi Terry,

I'm headed off to church this morning but will take a closer look when I return.

 

First impressions are:

1. The flat's focus position is about 2800 steps off of the light's focus position. That could make a difference. I don't know if it's the answer, but it's worth exploring. 

2. Wow, that 300 second dark is far brighter than my 300 second darks. It's clear from looking at your 300 second light that the amp-glow is different between the dark and the light. However, according to the FITS header, both were 300 seconds, at -15c at 11 gain. Offset isn't reported but I'd have a hard time believing offset could make that much difference.

 

I'm going to take a look at median ADU when I get back to see if I can figure out what's going on with that dark being so drastically different than the light.

 

-Jim



#21 terry59

terry59

    Cosmos

  • *****
  • topic starter
  • Posts: 9169
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 15 September 2019 - 08:20 AM

Hi Terry,

I'm headed off to church this morning but will take a closer look when I return.

 

First impressions are:

1. The flat's focus position is about 2800 steps off of the light's focus position. That could make a difference. I don't know if it's the answer, but it's worth exploring. 

2. Wow, that 300 second dark is far brighter than my 300 second darks. It's clear from looking at your 300 second light that the amp-glow is different between the dark and the light. However, according to the FITS header, both were 300 seconds, at -15c at 11 gain. Offset isn't reported but I'd have a hard time believing offset could make that much difference.

 

I'm going to take a look at median ADU when I get back to see if I can figure out what's going on with that dark being so drastically different than the light.

 

-Jim

You've hit upon something that has been niggling at me. All of the calibration files were shot when I had the bad camera driver installed



#22 kingjamez

kingjamez

    Vanguard

  • *****
  • Posts: 2069
  • Joined: 03 Oct 2006
  • Loc: Fairfax, VA

Posted 15 September 2019 - 02:24 PM

You've hit upon something that has been niggling at me. All of the calibration files were shot when I had the bad camera driver installed

I think you may have found it.

 

Either way, I think the differences in our cameras is going to make my masters not useful for comparison. 

 

My unity gain darks are around 709ADU, your 50second unity gain dark is ~1900 and the 300s is ~1950. I'm surprised by that difference but don't know enough about the QHY offering to know if that's significant. When I calibrate your light, as I'm sure you've seen, it way overcompensates using the dark supplied.  

 

 

The 300s light has a median ADU of 2160. Even with a 3nm filter, I have a hard time buying that a 300s exposure would only get 200ADU above a dark frame on such a bright object. 

 

I re-wrote the fits header of the 50s dark flat, to 300 seconds to see  what would happen if I used the dark flat as a dark for the 300s light. It did reduce the amp glow as I would expect for only being a 50s exposure. So it does appear that for some reason the 300s dark is seemingly way overexposed even though it was taken correctly.

 

I too couldn't get the flats to calibrate the light, but as I'm sure you've seen the dark flat does look like it's doing it's job. Perhaps it's the focus difference?

 

Wish I could be of more help, but my ASI 183's seem to be too different.

 

-Jim


Edited by kingjamez, 15 September 2019 - 03:05 PM.

  • OldManSky likes this

#23 terry59

terry59

    Cosmos

  • *****
  • topic starter
  • Posts: 9169
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 16 September 2019 - 06:51 AM

Thanks for taking a look Jim....I shot a new set of darks and a set of bias frames overnight so I'll see how those work a bit later 



#24 terry59

terry59

    Cosmos

  • *****
  • topic starter
  • Posts: 9169
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 16 September 2019 - 07:55 AM

This was calibrated with bias and darks only. The glow is "almost" calibrated out now but not quite. This has me puzzled

Attached Thumbnails

  • newdarks.jpg


#25 terry59

terry59

    Cosmos

  • *****
  • topic starter
  • Posts: 9169
  • Joined: 18 Jul 2011
  • Loc: Colorado, USA

Posted 16 September 2019 - 11:14 AM

I am (finally) in business with this camera. Gain 11/Offset 30, darks, bias, no flats or dark flats. It needs a pedestal of 100 added when calibrating the darks.

 

Very lightly processed....it just doesn't seem to need it

 

smile.gif

Attached Thumbnails

  • trunk2.jpg

Edited by terry59, 16 September 2019 - 11:25 AM.

  • Hondo likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics