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

Horsehead and Flame - D5300

dslr
  • Please log in to reply
31 replies to this topic

#1 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 19 January 2021 - 10:15 AM

This is my first experience with a Nikon camera and a modded camera.  The D5300 is more sensitive than the Canon 60D and it shows in the upper left corner of the image where the light from a nearby streetlight was clearly picked up.

 

I am still working out the kinks on collecting and applying flats to this setup.  I am not sure if the compression Nikon applies to its RAW images is a culprit.  I had to raise the average ADU to around 12K rather than the 8K I would expect with a 14-bit camera.  Lower values did not remove dust motes and I saw some weird rings appearing in my work in progress.

 

In any case I am happy to see the enhanced Ha response and am looking forward to capturing more with this setup. 

 

HH and Flame ST small.jpg

 

I am still working out the details of focus with APT, mostly due to my own impatience, I think.  Better days ahead!

 

Oops - I forgot to include details:  47 180 second lights, 21 darks, 81 flats, and 151 bias frames were used for calibration and stacking in DSS, post-processing done with StarTools.

 

Bob


Edited by bobharmony, 19 January 2021 - 10:25 PM.

  • Starman27, edif300, Scott Mitchell and 4 others like this

#2 moxican

moxican

    Viking 1

  • *****
  • Posts: 917
  • Joined: 13 Nov 2011
  • Loc: Toronto, Canada

Posted 19 January 2021 - 11:01 AM

I use the same camera (unmodded). What I'd like to know is, if this image is calibrated. You mentioned you are working on your flats, but did you apply any to this image? Furthermore, did you dark and bias (darkflat) calibrate? What was your equipment? What exposure time?

I had this same issue with my previous camera, hacked and modded D5100, that unless it was calibrated (dark, flat, bias) it was very difficult to dig out details and the dark areas were quite mottled for that reason. All in all the end image was always much noisier and I had to raise ADU values, which in turn really enhanced noise and flat field issues.


  • bobharmony likes this

#3 charlieb123

charlieb123

    Vostok 1

  • *****
  • Posts: 181
  • Joined: 05 Jun 2020
  • Loc: New York

Posted 19 January 2021 - 11:07 AM

Looks great considering your circumstances.

What are you using to focus with APT? An EAF?


  • bobharmony likes this

#4 asanmax

asanmax

    Vendor - DSLR Modifications

  • *****
  • Vendors
  • Posts: 719
  • Joined: 17 Sep 2018
  • Loc: Vancouver BC

Posted 19 January 2021 - 05:45 PM

Nice image! Lots of data and you were able to maintain the star sizes.

The secondary mirror spider artifacts have double lines which means that the scope was out of focus. You can use this as a gauge next time you focus it, just make sure all the lines are perfect.


  • bobharmony likes this

#5 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 19 January 2021 - 10:32 PM

I use the same camera (unmodded). What I'd like to know is, if this image is calibrated. You mentioned you are working on your flats, but did you apply any to this image? Furthermore, did you dark and bias (darkflat) calibrate? What was your equipment? What exposure time?

I had this same issue with my previous camera, hacked and modded D5100, that unless it was calibrated (dark, flat, bias) it was very difficult to dig out details and the dark areas were quite mottled for that reason. All in all the end image was always much noisier and I had to raise ADU values, which in turn really enhanced noise and flat field issues.

I have added the details of the calibration frames to the original post - in short lights were calibrated with darks, flats, and bias frames.  I am going to have another go at it replacing the bias frames with dark flats I have shot since.  Scope was the one in my signature, a C6N.  The ADU value I needed to raise was my flats average ADU.  I was quite surprised when I tried 8K adu flats (1/6 second) and neither the dust motes nor the vignetting was corrected.  12K (1/4") took care of both issues.

 

Looks great considering your circumstances.

What are you using to focus with APT? An EAF?

My focusing is manual at the moment (and for the near future as well - I spent my funds on the camera, there will be no more purchases for a while).  I just need to commit the time to get it working well.  I have traditionally used a baht mask and BYEOS to get good focus.  Since I went to APT (when I was still using the Canon) I have had trouble getting good diffraction spikes from the mask, which causes issues with the baht tool.

 

Bob
 


Edited by bobharmony, 19 January 2021 - 10:33 PM.


#6 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 21 January 2021 - 10:37 PM

I restacked in DSS with 81 dark flats in place of the bias frames.  I think the colors are a little better, although it is difficult to directly compare one image to the next.  I cropped a little bit to reduce the "streetlight effect", but would love to come up with a way to eliminate it entirely. 

 

Autosave021HH ST.jpg

 

Bob


  • Mike in Rancho likes this

#7 Mike in Rancho

Mike in Rancho

    Ranger 4

  • -----
  • Posts: 321
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 23 January 2021 - 05:49 AM

I restacked in DSS with 81 dark flats in place of the bias frames.  I think the colors are a little better, although it is difficult to directly compare one image to the next.  I cropped a little bit to reduce the "streetlight effect", but would love to come up with a way to eliminate it entirely. 

 

attachicon.gifAutosave021HH ST.jpg

 

Bob

That's pretty sharp-looking.  Strong colors.

 

I've been fighting the D5300 rings too.  Have not figured it out, and not sure anyone has.  Mark Shelly has joined, as well as started, some threads about it.  Right now I am trying flats at about 7/8 BOC histogram, which translates to the rightmost main peak about 8-9K ADU.  If I push it farther, where you say you are, I get a huge color cast that Startools cannot wipe out.  I'm trying out ASTAP right now as a stacker, to see how much it might help.  So far it seems a bit better than DSS as far as the rainbow rings go.  It also has a blur function for the master flat, which I will try out.  Alas the program overall is rather clunky and doesn't readily allow for multi-night group stacking where each session has its own calibration files.

 

For streetlights, or the moon for that matter, I've been blocking them using the canopy of my yard swing, or even a curved piece of cardboard taped to the end of the dewshield like the bill of a cap, to keep light from crossing into and reflecting off the inside end of the OTA.

 

After the fact, with some care in mask selection and mask fuzz you could probably kill off and repair that orange corner using by variously using the filter, color, and layer modules.


  • bobharmony and limeyx like this

#8 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 23 January 2021 - 11:16 AM

That's pretty sharp-looking.  Strong colors.

 

I've been fighting the D5300 rings too.  Have not figured it out, and not sure anyone has.  Mark Shelly has joined, as well as started, some threads about it.  Right now I am trying flats at about 7/8 BOC histogram, which translates to the rightmost main peak about 8-9K ADU.  If I push it farther, where you say you are, I get a huge color cast that Startools cannot wipe out.  I'm trying out ASTAP right now as a stacker, to see how much it might help.  So far it seems a bit better than DSS as far as the rainbow rings go.  It also has a blur function for the master flat, which I will try out.  Alas the program overall is rather clunky and doesn't readily allow for multi-night group stacking where each session has its own calibration files.

 

For streetlights, or the moon for that matter, I've been blocking them using the canopy of my yard swing, or even a curved piece of cardboard taped to the end of the dewshield like the bill of a cap, to keep light from crossing into and reflecting off the inside end of the OTA.

 

After the fact, with some care in mask selection and mask fuzz you could probably kill off and repair that orange corner using by variously using the filter, color, and layer modules.

Thanks, Mike.  I haven't looked at the BoC histogram.  In APT the autostrecthed blue channel is off the right side, and the green and red are well to the right of center.  Looking at the flats with APLabs, the median and mean numbers are 12K with a max right around 15K (out of a total of 16K), much higher than what I was used to with the Canon.

 

I also tried ASTAP for stacking but didn't spend a lot of time on it, and wasn't too pleased with the result.  I got the result I was looking for with DSS and stopped messing with it.  I agree that it is a bit kludgy, and slow.  But it does great at plate solving!

 

Thanks for the tips on the streetlight.  I was quite surprised to see it as I have never noticed anything in that area of the sky with the Canon.  I had an extended shield in the front end of the OTA.  I have been chatted with Ivo about what to do for it.  He suggested the Filter module as a possibility.  I also gave him a copy of the DSS stack and he is going to see what he can do on the linear side of things.

 

Bob


  • Mike in Rancho likes this

#9 Mike in Rancho

Mike in Rancho

    Ranger 4

  • -----
  • Posts: 321
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 24 January 2021 - 03:25 AM

Thanks, Mike.  I haven't looked at the BoC histogram.  In APT the autostrecthed blue channel is off the right side, and the green and red are well to the right of center.  Looking at the flats with APLabs, the median and mean numbers are 12K with a max right around 15K (out of a total of 16K), much higher than what I was used to with the Canon.

 

I also tried ASTAP for stacking but didn't spend a lot of time on it, and wasn't too pleased with the result.  I got the result I was looking for with DSS and stopped messing with it.  I agree that it is a bit kludgy, and slow.  But it does great at plate solving!

 

Thanks for the tips on the streetlight.  I was quite surprised to see it as I have never noticed anything in that area of the sky with the Canon.  I had an extended shield in the front end of the OTA.  I have been chatted with Ivo about what to do for it.  He suggested the Filter module as a possibility.  I also gave him a copy of the DSS stack and he is going to see what he can do on the linear side of things.

 

Bob

I do my flats straight to the SD card after unhooking everything and shutting down BYN, so I needed a way to know what the BOC histo meant.  Made a chart of where various exposures compare between BOC and the ADU as measured in the Iris program.  On a few, I too have seen where certain programs will think the blue is clipped off the right (including Gimp in its alleged linear mode), but Iris doesn't think so in the ADU graph.  That said, on my highest flat exposure test, which was maybe 15/16 BOC and per Iris had a peak around 13K or so, a very strong color circular color cast rendered the image trash.  The other weird thing about it was it showed three distinct peaks on the ADU graph in Iris.  Almost all the others only show two, as I believe a couple are overlaying each other.

 

I played with your streetlight a little bit, loading it with tracking on so I could do a minor denoise at the end.  It does take a bit of a dance what with the orange tone and gradient nature.  Because of that, I had to make several fuzzed masks of decreasing size, switching between the filter module for "reject" and the color module to try to match the colors and saturation of the background sky.  I sort of matched it, anyway.  Of course, it's soooo bright orange that who knows what is truly supposed to be hiding behind there, but I think if you put some elbow grease into the original data you could probably make it look like the streetlight never even happened.

 

Let us know though if Ivo gets back to you on other good ways to handle it.

 

Bob H Horsey x-streetlight.jpg


  • bobharmony and limeyx like this

#10 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 25 January 2021 - 12:11 AM

Mike, I heard back from Ivo.  He came up with a way to clean up the streetlight using the Wipe module.  He used the "Uncalib2" preset and reduced the gradient aggressiveness to 25%.  It did a great job of removing the streetlight, but dulled down the colors in the target quite a bit.  I have played around with various settings and processes today and haven't been able to get anything much better out of it.

 

About the Nikon color rings, here is what I got from the "Color" display in "Wipe" when I used 8K flats:

 

rings in Wipe.JPG

 

As you can see it was pretty ugly.  I used bias frames to calibrate the flats, I am wondering if dark flats would do a better job as the exposure was 1/6".  I am going to give it a try after I shoot the dark flats and see if it does any better.  I used dark flats with the 1/4" flats for this process and they did a better job of cleaning the dust bunnies and removing the rings.

 

Thanks for taking the time to work through the streetlights with the Filter module, your result was more natural looking than my first pass at it.

 

I have been reading several of the threads on the rings over the past few days and will continue to do so.  I shot this data at ISO 400 and was hoping to go down to ISO 200 the next time out, but I admit to being a bit fearful of it now.  We'll see how the experiment(s) go over the next few days.

 

Bob


  • limeyx and Mike in Rancho like this

#11 skywolf856

skywolf856

    Viking 1

  • -----
  • Posts: 628
  • Joined: 25 Jan 2008
  • Loc: SE Michigan

Posted 25 January 2021 - 06:15 PM

What was your ISO setting?



#12 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 25 January 2021 - 11:52 PM

What was your ISO setting?

I used ISO 400 for everything; lights, darks, flats, flat darks, and bias.  I never used bias and dark flats in a single stack, just alternated between them to see if one would work better.  I don't see any difference regardless of which is included.

 

Bob



#13 Mike in Rancho

Mike in Rancho

    Ranger 4

  • -----
  • Posts: 321
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 26 January 2021 - 12:59 AM

Mike, I heard back from Ivo.  He came up with a way to clean up the streetlight using the Wipe module.  He used the "Uncalib2" preset and reduced the gradient aggressiveness to 25%.  It did a great job of removing the streetlight, but dulled down the colors in the target quite a bit.  I have played around with various settings and processes today and haven't been able to get anything much better out of it.

 

About the Nikon color rings, here is what I got from the "Color" display in "Wipe" when I used 8K flats:

 

attachicon.gifrings in Wipe.JPG

 

As you can see it was pretty ugly.  I used bias frames to calibrate the flats, I am wondering if dark flats would do a better job as the exposure was 1/6".  I am going to give it a try after I shoot the dark flats and see if it does any better.  I used dark flats with the 1/4" flats for this process and they did a better job of cleaning the dust bunnies and removing the rings.

 

Thanks for taking the time to work through the streetlights with the Filter module, your result was more natural looking than my first pass at it.

 

I have been reading several of the threads on the rings over the past few days and will continue to do so.  I shot this data at ISO 400 and was hoping to go down to ISO 200 the next time out, but I admit to being a bit fearful of it now.  We'll see how the experiment(s) go over the next few days.

 

Bob

Wow, like a roll of Lifesavers!  My NGC1499 was like that, and a few others.

 

For bad rainbow rings I usually keep the standard preset in Wipe, move the second slider (falloff or something like that?) all the way to 100, and the aggressiveness to whatever level short of 100 I can get away with.  Often 90-95.  Worst case I will use the synthetic flats (even though I use flats) and/or the uncalib, to see if it helps.

 

Sometimes I'll wipe a little less, 85-90, if after that there's only one far-flung ring left - which seems to usually be yellow or blue - and I can then either crop it out, or select it easily in mask by color/brightness and then use filter->reject.  Poof, gone.

 

Yeah after today's rainbow ring developments I'm looking forward in trying to dial in Mark's recommended ADU levels for the light frames.  Seems so low!  But I'll give it a try.  Week of rain though, so I may just browse my HDD to see if I have any old data that was unintentionally underexposed.

 

Me too on the bias.  After reading up a bunch on DSS, I've just been doing dark flats.  I had been throwing everything in there.


Edited by Mike in Rancho, 26 January 2021 - 01:01 AM.

  • bobharmony likes this

#14 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 26 January 2021 - 10:34 PM

After reading through the first few pages of Jerry Lodriguss's topic on the D5300, I began to wonder if the 1/4" second flats might be clipped in the blue channel.  The light panel I am using for flats has a blue tint to it, which is what made me suspect that might be possible (that and the blue channel showing past the right end of the autostretched histogram in APT.  So I opened up one of the NEF files in PhotoShop Camera RAW and sure enough the center section was at 255 in the blue channel, and much of it was at 253 and 254 in the red and green channels.

 

Definitely not suitable for flats so I decided to load the master flat DSS generates into PhotoShop and check it as well.  The center was at 213/255 and it dropped off slowly toward the corners at 149/255.  Not a flat that I can trust as there wasn't enough spread in the source files to generate an accurate image.  The 1/10" flats that I took as an experiment come out much closer to what I'd expect so I am going to create a set of them and the darks to go with them and see what that looks like.

 

Bob


  • Mike in Rancho likes this

#15 Mike in Rancho

Mike in Rancho

    Ranger 4

  • -----
  • Posts: 321
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 28 January 2021 - 05:50 AM

After reading through the first few pages of Jerry Lodriguss's topic on the D5300, I began to wonder if the 1/4" second flats might be clipped in the blue channel.  The light panel I am using for flats has a blue tint to it, which is what made me suspect that might be possible (that and the blue channel showing past the right end of the autostretched histogram in APT.  So I opened up one of the NEF files in PhotoShop Camera RAW and sure enough the center section was at 255 in the blue channel, and much of it was at 253 and 254 in the red and green channels.

 

Definitely not suitable for flats so I decided to load the master flat DSS generates into PhotoShop and check it as well.  The center was at 213/255 and it dropped off slowly toward the corners at 149/255.  Not a flat that I can trust as there wasn't enough spread in the source files to generate an accurate image.  The 1/10" flats that I took as an experiment come out much closer to what I'd expect so I am going to create a set of them and the darks to go with them and see what that looks like.

 

Bob

How did it go?  It's really just hard to tell on these things.  I've been following the Nikon ring thread, while also trying to experiment at home as it goes along so as to get a better understanding of it.  Not with the greatest success either.

 

I too have thought that a flat had the blues clipped off the right, as it sure looked like it.  But Iris said it wasn't, in the raw ADU anyway.  That said I'm still trying to get used to an ADU histogram - sort of the photon bucket on its side I guess?

 

Does PS RAW truly work with the raw NEF file?  I've used Gimp to open frames to calibrate them manually, just to understand the formula.  And while it does in fact work, to a certain extent, the end result is nowhere near the same as if I just feed a single light, dark, flat, and dark flat into DSS and hit stack.

 

Weird.


  • bobharmony likes this

#16 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 28 January 2021 - 08:44 AM

Mike -

 

My experiments with shorter flats are going badly.  I tried 1/20" and 1/15" flats and both of them brought the rings back in force, and they don't neutralize the dust donuts at all.  I don't know how effective PS is at producing an accurate histogram with NEF files, but it really looked good for those two exposures.  I had previously used APLabs to measure them and 1/4" had a mean at about 12K there. 

 

I am going to try 1/5 and 1/3 and see what they generate.  I am also looking to see if DSS can adjust the individual color channels to counter the reported scaling of colors that Nikon does with the RAWs.  Finally I am looking into whether Newtons Rings is a possible cause of issues here.  I will have to look at Marks thread to check out the recommended light exposures there.

 

The battle continues!

 

Bob


  • Mike in Rancho likes this

#17 limeyx

limeyx

    Viking 1

  • -----
  • Posts: 869
  • Joined: 23 May 2020
  • Loc: Seattle, WA

Posted 28 January 2021 - 12:58 PM

Mike -

 

My experiments with shorter flats are going badly.  I tried 1/20" and 1/15" flats and both of them brought the rings back in force, and they don't neutralize the dust donuts at all.  I don't know how effective PS is at producing an accurate histogram with NEF files, but it really looked good for those two exposures.  I had previously used APLabs to measure them and 1/4" had a mean at about 12K there. 

 

I am going to try 1/5 and 1/3 and see what they generate.  I am also looking to see if DSS can adjust the individual color channels to counter the reported scaling of colors that Nikon does with the RAWs.  Finally I am looking into whether Newtons Rings is a possible cause of issues here.  I will have to look at Marks thread to check out the recommended light exposures there.

 

The battle continues!

 

Bob

Radical idea: Try 3 second flats even though this seems impossible

 

What I Do: Tracking pad +7-9 sheets of white copy paper ! Leave the pad on brightest to get a fixed light source and potentially avoid flickering with dimming

 

Track the ADUs to 7000-9000 *approx* - depends on your setup (I take 3" and 4" flats and just started dark flats)

 

Anecdotally this has worked *way* better for me than <1 second, but I don't have an explanation or enough data points yet

 

I can visually see them correcting dust much better than the short duration flats


  • bobharmony and Mike in Rancho like this

#18 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 28 January 2021 - 03:32 PM

Radical idea: Try 3 second flats even though this seems impossible

 

What I Do: Tracking pad +7-9 sheets of white copy paper ! Leave the pad on brightest to get a fixed light source and potentially avoid flickering with dimming

 

Track the ADUs to 7000-9000 *approx* - depends on your setup (I take 3" and 4" flats and just started dark flats)

 

Anecdotally this has worked *way* better for me than <1 second, but I don't have an explanation or enough data points yet

 

I can visually see them correcting dust much better than the short duration flats

Limeyx - interesting that you suggest this. I was just going over conversations I had with the guy I bought the camera from. He is honestly perplexed by what I am experiencing. He had sent me one of his flats to look at and mentioned that he was doing 1' flats. I was considering trying longer flats too. What the heck, I've already got two sheets of paper in front of the light panel, adding a few more can't hurt.

 

Being able to turn the panel up to full brightness would be a good thing too. I will give this a try and report back. I am sure this can be solved, it is just frustrating to work through the details!


I was also trying to find info on the Nikon shutter mechanism without a whole lot of success. When I set up flats for the Canon (a much simpler operation) it was pretty easy to find the description of the first curtain/second curtain method.

 

Bob


Edited by bobharmony, 28 January 2021 - 03:32 PM.


#19 Mike in Rancho

Mike in Rancho

    Ranger 4

  • -----
  • Posts: 321
  • Joined: 15 Oct 2020
  • Loc: Alta Loma, CA

Posted 28 January 2021 - 04:52 PM

Mike -

 

My experiments with shorter flats are going badly.  I tried 1/20" and 1/15" flats and both of them brought the rings back in force, and they don't neutralize the dust donuts at all.  I don't know how effective PS is at producing an accurate histogram with NEF files, but it really looked good for those two exposures.  I had previously used APLabs to measure them and 1/4" had a mean at about 12K there. 

 

I am going to try 1/5 and 1/3 and see what they generate.  I am also looking to see if DSS can adjust the individual color channels to counter the reported scaling of colors that Nikon does with the RAWs.  Finally I am looking into whether Newtons Rings is a possible cause of issues here.  I will have to look at Marks thread to check out the recommended light exposures there.

 

The battle continues!

 

Bob

Mark's thread is quite useful.  I posted up what I think is my understanding of things - will see if I have it confirmed.

 

Sticking to your horsey image here though, as Mark said, the rings can be in the lights, the flats, or both.  Flats can possibly be retaken, but the lights are a done deal.  Only less of a stretch and post-processing will work then, though I suppose different flats could change the color casting some.  But bad flats would probably only make it worse.

 

We are probably stuck with it and need to lower our lights exposures, unless we get a new firmware or some kind of after-the-fact batch repair processing of our NEF's.  That's probably hard - how to determine which pixels have gone haywire and how to fix them, without having anything else go sideways as a result.

 

I would do a test to see if the rings are already in your lights.  Maybe can do it with a single sub?  Last night I took some old M81 data and just put like 10 frames, plus darks, into DSS and stacked only those.  That was enough for me to feed into either ST or Gimp and be able to bring the rings out.  In a case like that, I believe you then just want the best flats for vignetting and dust, and without their own rings thrown into the mix.


  • bobharmony likes this

#20 limeyx

limeyx

    Viking 1

  • -----
  • Posts: 869
  • Joined: 23 May 2020
  • Loc: Seattle, WA

Posted 28 January 2021 - 05:07 PM

Limeyx - interesting that you suggest this. I was just going over conversations I had with the guy I bought the camera from. He is honestly perplexed by what I am experiencing. He had sent me one of his flats to look at and mentioned that he was doing 1' flats. I was considering trying longer flats too. What the heck, I've already got two sheets of paper in front of the light panel, adding a few more can't hurt.

 

Being able to turn the panel up to full brightness would be a good thing too. I will give this a try and report back. I am sure this can be solved, it is just frustrating to work through the details!


I was also trying to find info on the Nikon shutter mechanism without a whole lot of success. When I set up flats for the Canon (a much simpler operation) it was pretty easy to find the description of the first curtain/second curtain method.

 

Bob

Yeah it took my months of messing with this. I have a lot of bad data that I had to crop and clip black points to make usable

 

It *might* be related to amount of vignette but I dont know if anyone is sure, and possibly amount of moonlight or other light coming in.

 

I used my refractor for the first time (during a good sized moon) at ISO400 and had no rings but I have only had one clear night all year so more data points are needed


  • bobharmony likes this

#21 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 28 January 2021 - 06:36 PM

I did a stack of just lights and darks - good news (?) is that the background of the Wipe is the the nice pale green I am used to from shooting with the Canon, with some red areas right where they should be, around the horsehead and a little in the Flame area.  No circles at all.  It gives me hope that I can get it better if I can fix the flats - next up is some longer flats.

 

I read through Marks thread in detail today, I see you guys have been pretty regular contributors.  Nice job keeping the discussion going and some great questions and suggestions.  I will be tagging in there soon when I have something useful to add.

 

And so it goes...

 

Bob


  • limeyx likes this

#22 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 29 January 2021 - 12:12 AM

I went with 3 second flats, a fully illuminated light panel, and 11 sheets of paper between the panel and the scope.  The color rings are gone, but the dust motes aren't.  I am going to mess around with longer flats and see if I can get any improvement.  Also shorter flats just to keep things interesting.  It appears that I can't trust the PS histogram.  The one the APLab produces only allows me to display the red or green pixels.  They are both mid-range, which may mean I have saturated the blue.  They both also show the peaks and valleys that we know are a "feature" of the lossy compression. 

 

Interestingly, lights go up to an ADU of a little over 2000, and they are smooth as silk - go figure.  Maybe dithering had a positive effect on them.

 

Bob



#23 limeyx

limeyx

    Viking 1

  • -----
  • Posts: 869
  • Joined: 23 May 2020
  • Loc: Seattle, WA

Posted 29 January 2021 - 12:42 AM

I went with 3 second flats, a fully illuminated light panel, and 11 sheets of paper between the panel and the scope.  The color rings are gone, but the dust motes aren't.  I am going to mess around with longer flats and see if I can get any improvement.  Also shorter flats just to keep things interesting.  It appears that I can't trust the PS histogram.  The one the APLab produces only allows me to display the red or green pixels.  They are both mid-range, which may mean I have saturated the blue.  They both also show the peaks and valleys that we know are a "feature" of the lossy compression. 

 

Interestingly, lights go up to an ADU of a little over 2000, and they are smooth as silk - go figure.  Maybe dithering had a positive effect on them.

 

Bob

So that's a positive, right ? Do you have Iris ? I load mine in and check histogram. If the dust is still there, check that it's actually in the flats by stretching (in photoshop, de-saturate fully and stretch)

 

if Iris says the ADU is much over 9500 then you might need more paper. I can see my flats working when I stack with and w/out


  • bobharmony likes this

#24 bobharmony

bobharmony

    Messenger

  • *****
  • topic starter
  • Posts: 421
  • Joined: 28 Oct 2017
  • Loc: Connecticut, USA

Posted 29 January 2021 - 09:01 AM

So that's a positive, right ? Do you have Iris ? I load mine in and check histogram. If the dust is still there, check that it's actually in the flats by stretching (in photoshop, de-saturate fully and stretch)

 

if Iris says the ADU is much over 9500 then you might need more paper. I can see my flats working when I stack with and w/out

It is indeed a positive, although I will have a better feel when I can get the flats under control.  I don't have Iris (yet), but stand by.  Is version 5.59 (June 24, 2010) really the latest version?  It would be nice to be able to look at the blue channel.

 

Does running longer time flats reduce the effect of screen flicker?  It seems like 3" would handle any power cycle issues. If not soon the stack will be thicker than some of the books I am reading smile.gif

 

Bob



#25 limeyx

limeyx

    Viking 1

  • -----
  • Posts: 869
  • Joined: 23 May 2020
  • Loc: Seattle, WA

Posted 29 January 2021 - 11:39 AM

It is indeed a positive, although I will have a better feel when I can get the flats under control.  I don't have Iris (yet), but stand by.  Is version 5.59 (June 24, 2010) really the latest version?  It would be nice to be able to look at the blue channel.

 

Does running longer time flats reduce the effect of screen flicker?  It seems like 3" would handle any power cycle issues. If not soon the stack will be thicker than some of the books I am reading smile.gif

 

Bob

Yup thats the version I have. I use it for a quick ADU check alongside PixInsight

 

Definitely check that the motes show in your flats - I have some that are uncorrelated and are never going to fix the dust no matter what I do


  • bobharmony 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





Also tagged with one or more of these keywords: dslr



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics