294MC Pro, L-eXtreme and a bunch of flats......

[scotty38]

Not sure where to start but it did start with an image of NGC281 that I posted in another thread discussing guiding. I noticed odd colouration/gradients and thought I'd narrowed it down to the flats but I think it's the filter. I know this camera's sensor and this filter has been discussed all over the place with regards to flats etc but I'm not sure how to fix it. I've been messing about all day with both the ASCOM and ZWO drivers and different exposures and each time the flats seem to look ok but the resulting light is always odd looking. I have so many files and tests my head is in a spin but below is a screen shot showing a light with no flats, a light with flats, the masterflat and a debayered single flat frame,.

If I go through the whole process using my UV/IR filter then all seems to be fine so it's definitely when the L-eXtreme is in play. I've taken other images with this filter and I think the subject has masked things somewhat but if I go back the "issue" is there I think. I've not had the filter long and given the weather I've not really done much with it until now so only really just become apparent.

As I said I've tried the ZWO and ASCOM drivers, flats at less than a second and up to 4 seconds but they all result in the same strange lights to some degree. I don't know what other info to provide so what are my next steps to trying to resolve this? I can post files as .xisf, .fits or .jpeg if needed.

Please fire away with questions, hopefully I am missing something fundamental 🙂

[david_taurus83]

You calibrating with raw images before you debayer? What light source you using for flats? I get over correcting and different colour balance if I use a cheap LED tracing pad.

[Adam J]

57 minutes ago, scotty38 said:

Not sure where to start but it did start with an image of NGC281 that I posted in another thread discussing guiding. I noticed odd colouration/gradients and thought I'd narrowed it down to the flats but I think it's the filter. I know this camera's sensor and this filter has been discussed all over the place with regards to flats etc but I'm not sure how to fix it. I've been messing about all day with both the ASCOM and ZWO drivers and different exposures and each time the flats seem to look ok but the resulting light is always odd looking. I have so many files and tests my head is in a spin but below is a screen shot showing a light with no flats, a light with flats, the masterflat and a debayered single flat frame,.

If I go through the whole process using my UV/IR filter then all seems to be fine so it's definitely when the L-eXtreme is in play. I've taken other images with this filter and I think the subject has masked things somewhat but if I go back the "issue" is there I think. I've not had the filter long and given the weather I've not really done much with it until now so only really just become apparent.

As I said I've tried the ZWO and ASCOM drivers, flats at less than a second and up to 4 seconds but they all result in the same strange lights to some degree. I don't know what other info to provide so what are my next steps to trying to resolve this? I can post files as .xisf, .fits or .jpeg if needed.

Please fire away with questions, hopefully I am missing something fundamental 🙂

@PadrePeace may be able to provide some insight here. 

Adam

[scotty38]

18 minutes ago, david_taurus83 said:

You calibrating with raw images before you debayer? What light source you using for flats? I get over correcting and different colour balance if I use a cheap LED tracing pad.

I'm using a Geoptik flat panel so I can adjust to alter the exposure, ADU as needed. I then create a bunch of flats and flat darks and either create masters to use separately or I'll stick the raw files in with everything else. Same end result though. Oh I'm using the WBPP script in PixInsight.

[PadrePeace]

22 hours ago, Adam J said:

@PadrePeace may be able to provide some insight here. 

Adam

Scotty38, I have the 294MC and have had similar FLATs difficulties with the Optalong E-extreme Dual band filter (Ha and OIII 7nm) that being overcorrection down the left and right sides. I went through the same diagnostic tests you have and had gigabits of flats at various exposures and from different light sources including sky flats. My conclusion was that the filter is either leaking additional bright light source frequencies (suspected with sky flats) or the flats are compromised by trying to take a flat at one optimum exposure setting across two specific frequencies. As an illustration my other specific Ha and OIII filters used for NB with my mono need very different treatment to each other to produce working Flats, so why would there be one flat exposure to rule them all so to speak? Someone please come back and explain why I’m wrong if that’s indeed the case.

So What?

Having tried to get a single master flat that worked (using the Flats tool in NINA) for the 294MC and the e-extreme, I ended up stacking my image subs using a let’s say ‘compromised’ master FLAT taken at 5.63s exp and then removing the over corrected corruption around the edges of the integrated image using the light pollution tool in NINA. I’m sure the Flat is introducing correction where it was not required hence the need to remove it as LP. This worked well and as yet I have not seen a more definitive solution posted.
 

Some report not to have this issue at all but then the variables from poor filter quality control through to those that just don’t see poor correction as an issue could explain that away. All I know is my 294MC has issues with my dual band filter or indeed the other way around.
 

Bottom line is my technique produced the following image with a ‘compromised’ flat of 5.63s exp for an ADU of 23k also posted below. 
 

you may also wish to check out this thread 

Edited January 9, 2022 by PadrePeace

[scotty38]

Thanks @PadrePeace that's a great image and I'd be more than happy with it. I had another play a little while ago and created flats at almost 6 seconds with just over 32k ADU so about 50% but same issues again. With my light (top right in the screenshot) it appears I have more than just edge issues as you can see the red and green banding running round the image so I'd be losing 60% of the frame getting rid of that.

I think I have read though that thread, there are probably none I've not seen but I've read so much I don't know which way to turn 🙂 Just off to have another read through it and then I will also have another play, maybe tomorrow now, to see if I can replicate your method.

Thanks again....

[PadrePeace]

12 minutes ago, scotty38 said:

Thanks @PadrePeace that's a great image and I'd be more than happy with it. I had another play a little while ago and created flats at almost 6 seconds with just over 32k ADU so about 50% but same issues again. With my light (top right in the screenshot) it appears I have more than just edge issues as you can see the red and green banding running round the image so I'd be losing 60% of the frame getting rid of that.

I think I have read though that thread, there are probably none I've not seen but I've read so much I don't know which way to turn 🙂 Just off to have another read through it and then I will also have another play, maybe tomorrow now, to see if I can replicate your method.

Thanks again....

I think you’ll find that the red and green banding is a function of this sensor. It’s a common theme on 294 flats posted here and like a fingerprint varies between sensor examples. It is in each sub (though not visible in an unstretched RAW single sub) and needs to be calibrated out. It’s the bright stuff down each side that is my issue cos it’s not in the image subs. Try reducing your flats exposure to 5s at 23ADU and cool the sensor too. 

Good luck. 

Edited January 9, 2022 by PadrePeace

[PadrePeace]

Forgot to mention that I split the Ha and the OIII out from the dual band subs when I integrated them and calibration files in NINA, then applied the LP correction to the Ha and OIII output files (output in greyscale). Splitting the two bands may be key to processing the dual band subs. Just saying.

Edited January 9, 2022 by PadrePeace

[scotty38]

Thanks I do cool the sensor and I've tried various ADU values from maybe 10% all the way to 90%. I do have APP too and have extracted Ha and Oiii so will try that again with this image.

[scotty38]

Tried the Ha extraction alone and the resulting file is pretty much the same. I'm going to reinstall and redo all my calibration frames and make sure my next set of lights are all using the same driver etc as I've messed about so much now I need to start again with a fresh setup.

If that doesn't work I may buy a new (mono) camera 🙂 🙂 I'm not even sure separate Ha and Oiii filter swould be an (inefficient) answer.

294MM - fewest changes needed but does it suffer the same as the 294MC

2600MM - nice but expensive and new filterwheel and (more) expensive filters needed

533MM when it arrives but smaller sensor plus a bit of an unknown..

 

[The Lazy Astronomer]

18 hours ago, scotty38 said:

If that doesn't work I may buy a new (mono) camera

Any excuse for a new purchase 😁

18 hours ago, scotty38 said:

294MM - fewest changes needed but does it suffer the same as the 294MC

The 294MM does calibrate well in my experience, but I feel like it's a bit of a fussy sensor. I've only really done narrowband imaging with it, and it took a little bit of fiddling to find the right settings for good flats, but now that I have that, calibration has so far been very repeatable.

I'm yet to really try much broadband imaging as my current (cheapo) rgb filters have some weird reflection artifacts which do not calibrate out, but the Astronomik luminance filter I use seems to calibrate fine as well.

One odd thing (and this is probably due to the cheap tracing panel I use) is that Ha and Sii calibrate perfectly, but Oiii has a brighter central region in the flat which then overcorrects the lights - nothing I tried worked and in the end I had to resort to a different light source (which worked fine).

[scotty38]

52 minutes ago, The Lazy Astronomer said:

Any excuse for a new purchase 😁

The 294MM does calibrate well in my experience, but I feel like it's a bit of a fussy sensor. I've only really done narrowband imaging with it, and it took a little bit of fiddling to find the right settings for good flats, but now that I have that, calibration has so far been very repeatable.

I'm yet to really try much broadband imaging as my current (cheapo) rgb filters have some weird reflection artifacts which do not calibrate out, but the Astronomik luminance filter I use seems to calibrate fine as well.

One odd thing (and this is probably due to the cheap tracing panel I use) is that Ha and Sii calibrate perfectly, but Oiii has a brighter central region in the flat which then overcorrects the lights - nothing I tried worked and in the end I had to resort to a different light source (which worked fine).

Thanks so what is your process for getting good flats? I ask as I have a Geoptik light panel and it's the one thing I've not swapped out although it seems fine with my UV/IR filter. At the moment no matter what I try the end result with l-extreme is best if I don't use flats at all.

I have read in other places that the 294MM has similar, if not as bad, issues as the MC version. They are different sensors so I guess there's no reason they should suffer the same.... The 2600MM doesn't seem to have any bad points apart from the cost itself plus I'd need a new filter wheel as well as bigger, more expensive filters.

[The Lazy Astronomer]

3 hours ago, scotty38 said:

Thanks so what is your process for getting good flats? I ask as I have a Geoptik light panel and it's the one thing I've not swapped out although it seems fine with my UV/IR filter. At the moment no matter what I try the end result with l-extreme is best if I don't use flats at all.

I have read in other places that the 294MM has similar, if not as bad, issues as the MC version. They are different sensors so I guess there's no reason they should suffer the same.... The 2600MM doesn't seem to have any bad points apart from the cost itself plus I'd need a new filter wheel as well as bigger, more expensive filters.

Flats temperature matched to lights, with matching flat darks. For Ha, Sii and luminance, I use a cheap tracing panel with a few sheets of white paper, target adu value of around 40% (so around 26 - 27k adu), exposures somewhere in the 3 - 10s range.

For Oiii, no amount of variation in adu value, exposure time, brightness or distance of the tracing panel would get me good flats, so on an overcast day (no shortage of those in the UK 😅), I covered the scope with a white tshirt and pointed it at a plain white section of wall (in a room which already has quite soft and diffused lighting in the mid to late afternoon) and fired off some flats using the same settings as above. Worked perfectly.

As I say, I haven't really tried with rgb properly yet - I'm going to get some better filters soon and I guess I'll find out what works.

[Jamgood]

I have the same camera and had issues with flats when I first got it. After reading everything about the cameras problems and needing longer flats I figured that they would work but I could never get long flats to work. 

What I do, using APT, is use the Auto Flats Wizard to get the recommended time for the flats and then take that number over in the a normal Light Plan with the same settings as my lights, just the different timing and they work perfect. My flats are usually 0.03secs with the L-Enhance which completely negates everything I've read about needing to do long flats. I haven't tried with N.I.N.A though as I don't use it. For a light source I use an old TV Tablet with a blank white jpg open.

Have you tried short flats?

[scotty38]

Thanks @The Lazy Astronomer and @Jamgood yes flats match everything such as temperature, gain, offset etc. I've tried exposures from 0.x seconds to about 7 seconds and everything in between. I've tried ADU from 5000 to 55000 and points in between and they all give the same sort of result in the master light.

I tried with flats created using the UV/IR filter and got a totally different result as expected so I am going to try and create them with a different tool, maybe APT or by not using my flat panel and see what happens.

[The Admiral]

If I'm not mistaken, I think what you are discussing is a well known issue with this camera when using NB. There is a long thread here on Cloudy Nights that you might want to look at and compare notes. Start towards the end of p7.

Also some discussion here.

Ian

 

Edited January 11, 2022 by The Admiral

[scotty38]

1 hour ago, The Admiral said:

If I'm not mistaken, I think what you are discussing is a well known issue with this camera when using NB. There is a long thread here on Cloudy Nights that you might want to look at and compare notes. Start towards the end of p7.

Also some discussion here.

Ian

 

Yes agreed thanks. Thought I'd mentioned, maybe not in this thread, that I've pretty much read all the various posts on this subject. What I was aiming for here was to try and fix the issue if I could but so far not much joy....

[The Admiral]

49 minutes ago, scotty38 said:

that I've pretty much read all the various posts on this subject.

Yes you did, looking back. Have you seen anything to suggest a solution had been found? I don't have an ASI294, but I recently acquired an ASI533 and did a quick trial of taking a flat with the L-Extreme in situ, and it looks as though there are colour variations with a stretched flat. Unfortunately I haven't had the opportunity to test it out properly under real conditions. Fingers crossed that it should be useable, given that I've seen results using that combination.

Ian

[scotty38]

47 minutes ago, The Admiral said:

Yes you did, looking back. Have you seen anything to suggest a solution had been found? I don't have an ASI294, but I recently acquired an ASI533 and did a quick trial of taking a flat with the L-Extreme in situ, and it looks as though there are colour variations with a stretched flat. Unfortunately I haven't had the opportunity to test it out properly under real conditions. Fingers crossed that it should be useable, given that I've seen results using that combination.

Ian

Not really no and I've tried t-shirt with my panel, ipad and paper and flats created in Sharpcap this morning and they're all the same. I am wondering if it's the lights that the flats just cannot compensate for and just make things worse. With no flat applied you can see the gradients anyway so I wonder if I'm barking up the wrong tree.

I may call it a day for now and see what happens on the next clear night and new lights......

[The Lazy Astronomer]

2 hours ago, scotty38 said:

Not really no and I've tried t-shirt with my panel, ipad and paper and flats created in Sharpcap this morning and they're all the same. I am wondering if it's the lights that the flats just cannot compensate for and just make things worse. With no flat applied you can see the gradients anyway so I wonder if I'm barking up the wrong tree.

I may call it a day for now and see what happens on the next clear night and new lights......

Have you tried flats using only natural light too? And when you say they're all the same, have you tried calibrating the lights with all of these flat variants, or are you just looking at a stretched master?

[scotty38]

8 minutes ago, The Lazy Astronomer said:

Have you tried flats using only natural light too? And when you say they're all the same, have you tried calibrating the lights with all of these flat variants, or are you just looking at a stretched master?

No flats with natural light yet, just an ipad and a flats panel. Yes, every time I create new flat files I calibrate etc etc again and generate a new master light.

Later this evening I'll come back to this and post some more files:

1. L-eXtreme master light with NO calibration files

2. L-eXtreme masterlight with darks only

3. L-eXtreme masterlight with all calibration files

4. UV/IR masterlight with all calibration files

From what I've seen today messing about in between calls is that only option 3 has any issues.

[The Admiral]

I'm sure that you've read the same, but the overall impression I get from posts is that some 294's cameras suffer more from this than others. A lot of the CN contributors do seem to just throw in the towel and return their camera for one of a different model. If you can find the holy grail I'm sure it'd be appreciated . Though I suspect there'll need to be a lot of hard work to find it, if at all. Good luck.

Ian

[scotty38]

1 hour ago, The Admiral said:

I'm sure that you've read the same, but the overall impression I get from posts is that some 294's cameras suffer more from this than others. A lot of the CN contributors do seem to just throw in the towel and return their camera for one of a different model. If you can find the holy grail I'm sure it'd be appreciated . Though I suspect there'll need to be a lot of hard work to find it, if at all. Good luck.

Ian

Yes I have read all that too.

I am speaking with the supplier too (not FLO) to see what can be done and in my dreams an exchange for something else would be wonderful. Either way it works out my options are:

294MM - could be in the same boat as a 294MC

533MM - not here yet, could be great, could have issues

2600MM expensive plus expensive filters

071MC same price as 294MM, older tech but no worse than the 294. Still expensive filters but would only need dual band anyway....

 

[scotty38]

Here we go. All files have had an STF stretch only. Seems the lextreme filter and flats conspire to muck things up.

1. L-eXtreme master light with NO calibration files

2. L-eXtreme masterlight with darks only

3. L-eXtreme masterlight with all calibration files

4. UV/IR masterlight with all calibration files

[Jamgood]

@scotty38 Have a read through this thread and see if you can find anything helpful. @vlaiv helped me solve my problem when I got my 294mc. There's quite a bit of info in there, if you haven't read it already.