Jump to content

Narrowband

Rising Cam IMX 571 gradient issues


Clarkey

Recommended Posts

Up until now, most of my imaging has been done using a mono camera - ZWO 1600mm pro and I have no issues with gradients / processing. I have also used a DSLR in the past. However, over the summer I decided to buy a OSC camera - in this case the Rising Cam IMX571. I have only had 3 nights to try and get any imaging done since the purchase, but what I am getting is awful and useless. I feel there must be a problem with the camera, but it might just be something obvious I am missing. For information, I am using exactly the same set up with the mono camera without any issues.

There seems to be three main problems. Firstly, the images have really bad gradients across them. Secondly, the flats are not compensating and actually making everything worse. (There is also a lot of dirt in the camera which however much cleaning I do I cannot remove the dust). The histotgram for the flat is very odd, with three distinct peaks. Thirdly, when I try to take flats, NINA often says that the screen is too dim, even though the light panel is on full brightness - almost like day. In addition to this, in Affinity it opens as an 8 bit image and in Astro pixel processor the flats have a fixed value blue channel.

I am at a bit of a loss regarding what is going on. I have updated drivers / firmware where I can, but I still cannot get a decent image. Any guidance regarding what I am doing wrong would be greatly appreciated.

 

FLAT.jpg

NGC2064-RGB JPG.jpg

Link to comment
Share on other sites

I also see the misaligned histogram on individual subs, and  the intense green/yellow cast. This is removed after normalisation and stacking, or at least it is for me after calibration and stacking in APP. I use darks, flats and dark flats for calibration.

I haven’t encountered the flats too dim issue when using NINA’s Flats Wizard. I usually have to dim the panel to get a reasonable exposure time of 2-3 seconds and a peak around two thirds along the axis. This is with an Esprit 150 and both the Risingcam and QHY OSC versions, they behave very similarly. I also see multiple peaks on the flats histogram, which concerned me initially but the subs calibrated OK with the master flat so I was relaxed about it after that.
 

Do you get any warning messages when creating the master calibration frames in APP?

How does your stacked image look after gradient removal? Mine always look very washed out before I apply the ‘Remove Light Pollution’ tool in APP, an essential first processing step for me.

Link to comment
Share on other sites

Do the single subs have gradients? That's a stack you've posted?

The 3 peaks will be the RGB channels. Depending on your light source, the 3 channels will vary. I see this also with my DSLR so nothing new there.

Is that a debayered and stretched flat? It does look very bright.

Link to comment
Share on other sites

Hi,

I'll try to answer some of the questions. Firstly, the green hue I am expecting - its twice as much green as the other colours. 

image.thumb.png.e74ff3aadd5812e1432ea95a1ebe762b.png

This is a screen shot of a flat. The histogram is very odd for an RGB - it has no blue in it.  Below is another flat opened in affinity which looks totally different? As for the 'Flats are too dim' - I am completely lost on this one. The problem I usually have is getting the flat panel dim enough.

image.thumb.png.f0de180d8297d8cd12804b2514e7bffe.png

 

6 hours ago, newbie alert said:

You say that there's lots of dust but your flats aren't showing any?

No - but they really show up of the 'corrected' images.

This is a corrected image - even worse than the one above:

 

image.thumb.jpeg.6aec4616601d0159652e7d89ec0bcfc5.jpeg

FWIW - I have the inbuilt dew heater and an additional dew heater, so I don't think it is due. I am at a total loss as to what is going on with the camera.

5 hours ago, david_taurus83 said:

Do the single subs have gradients? That's a stack you've posted?

The 3 peaks will be the RGB channels. Depending on your light source, the 3 channels will vary. I see this also with my DSLR so nothing new there.

Is that a debayered and stretched flat? It does look very bright

Yes - it is stacked and stretched. I appreciate there will be some difference - but not to the extent it is showing. The gradients can be seen on single and stacked flats (don't worry about the very dark patch on some - that is the OAG which needs adjusting.

This is a single unstretched or altered flat:

image.thumb.png.13853bfe098e5f29c764b9f3038e5cfb.png

 

7 hours ago, tomato said:

Do you get any warning messages when creating the master calibration frames in APP?

How does your stacked image look after gradient removal? Mine always look very washed out before I apply the ‘Remove Light Pollution’ tool in APP, an essential first processing step for me

No error messages. The gradient removal sort of helps - but it is only masking another issue. The quality of the images does vary - but there are none that are even remotely acceptable. The last image of the original post is the best I have managed.

 

image.png

Edited by Clarkey
Link to comment
Share on other sites

1 hour ago, tomato said:

How does the flat look if you tick the  background neutralisation box in APP? This is what my Risingcam master flat looks like with and without BG neutralisation

I'll have to try this. Not something I have done so far. Do I need to do something similar with the images of just create the flat with the box ticked?

 

1 hour ago, newbie alert said:

So this is an image with flats applied?

So like I said they're too bright so they're overcorrecting

Are you choosing the right Bayer pattern? Not a OSC user so not too sure if that's relevant 

Yes.

I'm not sure they are too bright - but yes there does seem to be some over-correction. The image above seems particularly bad so I wonder if something else is going on. Even on other images the correction is poor. And yes, I am using the right Bayer pattern, although I think APP works it out anyway. As I said in a previous post, I am actually getting a warning from NINA saying my flats are too dim and suggesting I need 80 second exposures?!

Link to comment
Share on other sites

Did you use darkflats and darks with calibration? Without those you will have overcorrection of flats and they look like yours.

Alternatively if you have some weird offset issues where some of the calibration frames are taken at different offsets to the lights you will get this kind of deal, or maybe a little bit of overcorrection if the offset were 0.

As for the NINA flats issue, no clue but i actually also ran into this issue. Previously i shot with a Maxfield 0.95x coma corrector and flats exposures were 0.136s give or take a few decimals in the end when using NINA flats wizard, but now i have a TeleVue Paracorr which is a 1.15x barlow, so flats exposures are increased to 0.205s. Thing is, NINA flats wizard refuses to figure out the exposure 3/4 of the time and says something about the panel being too bright and no solution was found to get a good exposure(cant recall exactly what). But manually taking the flats works just fine. Below is what a histogram looks like where flats calibration works with no apparent flaws:

Flats-histogram.JPG.cf414761f7ea063f797dc30101b2eb01.JPG

My panel is a plastic toy - quality amazon tracing LED panel and visually a little bit bluer than neutral white, hence the blues and reds switched places to what you would expect based on the QE curve of the camera. This flat calibrates all the channels just fine so have not bothered looking into a proper panel.

If you want to, you could post one of each calibration frame and a matching light frame preferably straight off the camera in .FITS format and others could have a look. Its easy to overlook something and get left with odd looking data.

Link to comment
Share on other sites

TBH I don’t think ticking the BG neutralisation box changes how the master flat is applied in APP’s calibration process, the frame characteristics are all sorted out by the software. However, if it is then displayed as you would expect the flat to look like, you know there is nothing fundamentally wrong with it.

Link to comment
Share on other sites

8 hours ago, Clarkey said:

 

I'm not sure they are too bright - but yes there does seem to be some over-correction. The image above seems particularly bad so I wonder if something else is going on. Even on other images the correction is poor. And yes, I am using the right Bayer pattern, although I think APP works it out anyway. As I said in a previous post, I am actually getting a warning from NINA saying my flats are too dim and suggesting I need 80 second exposures?!

If they're not too bright how can it be over correcting

If Nina is correct how can a 80 sec sub for a flat be right??

Link to comment
Share on other sites

8 hours ago, ONIKKINEN said:

Did you use darkflats and darks with calibration? Without those you will have overcorrection of flats and they look like yours.

Yes - both.

 

8 hours ago, ONIKKINEN said:

Alternatively if you have some weird offset issues where some of the calibration frames are taken at different offsets to the lights you will get this kind of deal, or maybe a little bit of overcorrection if the offset were 0.

I'm assuming the flat wizard would use the same for both imaging and flats - my offset is currently at 700.

8 hours ago, ONIKKINEN said:

My panel is a plastic toy - quality amazon tracing LED panel and visually a little bit bluer than neutral white, hence the blues and reds switched places to what you would expect based on the QE curve of the camera. This flat calibrates all the channels just fine so have not bothered looking into a proper panel.

Mine is something similar and has worked fine with the other cameras. The histogram is odd in that although using the ZWO fits viewer it looks normal, but in APP and Affinity is very strange. I am guessing the ZWO program stretches and aligns the 3 channels.

Link to comment
Share on other sites

7 minutes ago, newbie alert said:

If they're not too bright how can it be over correcting

If Nina is correct how can a 80 sec sub for a flat be right?

I think it is overcorrecting some of the gradients - not the image as a whole. In most case it is not as bad as the horrendous image above. But it is still not correcting correctly.

NINA was suggesting an 80 sec flat - I have no idea why. I did not take 80 second flats - I ended up doing them manually to about 1/3 of range.

 

I will try to post some single representative raw frames when I get a chance for people to dissect at their leisure. Thanks for everyone's input so far.

 

Link to comment
Share on other sites

33 minutes ago, Clarkey said:

I think it is overcorrecting some of the gradients - not the image as a whole. In most case it is not as bad as the horrendous image above. But it is still not correcting correctly.

NINA was suggesting an 80 sec flat - I have no idea why. I did not take 80 second flats - I ended up doing them manually to about 1/3 of range.

 

I will try to post some single representative raw frames when I get a chance for people to dissect at their leisure. Thanks for everyone's input so far.

 

I'm not sure that a flat frame will calibrate out a gradient, it will help calibrate out dust and vignette as it shows in your image as overcorrecting.... A few years ago when flats software wasn't available you had nothing to flag up to tell you that your flat panel was too bright or your adu was too high etc, you did it by eye.. what are you using as a light source? maybe that could be causing issues

Is the dust on the sensor, like from the factory ? Presume it has a glass window? Or is it in a filter ?  It's too prominent to be further away

As it happens lots of people have issues with Nina's flats wizard

Edited by newbie alert
Added info
Link to comment
Share on other sites

8 minutes ago, newbie alert said:

I'm not sure that a flat frame will calibrate out a gradient,

That was just my poor wording. I realise that genuine gradients will not calibrate out with flats. (Although there does seem to be a bad area in one corner of the sensor which rotates with the camera / scope. When it is not so cold, I will try to collimate the RC8 as it is a little bit out and it could be the cause). 

9 minutes ago, newbie alert said:

Is the dust on the sensor, like from the factory

Yes. I am a stickler for clean filters / lenses etc. My 3 years old 1600mm and the filters for that show virtually no dust.  I have cleaned everything but still get some prominent dust motes which will not go away. I am pretty certain these are on the inside of the glass. I am just reluctant to send it back to China or open it up. 

Link to comment
Share on other sites

I believe the flat wizard uses whatever you have set the default offset to be. If you have set it to 700, all should be good but if you have left default offset at 768 and just use 700 for imaging sequences, your flats will be at higher offset.

Link to comment
Share on other sites

A bit late, but here are some individual subs. One flat frame and one image frame.

I am seriously considering binning this camera. I tried some more imaging with the same result. What makes it worse is the sensitivity seems so poor. The trouble is, I don't want to sell it if there is something wrong with it (which I suspect), which means I now have a £1000 paper weight. I could sell it and for the money buy another second hand 1600mm pro. At least these work.

@ONIKKINENand @tomato - please feel free to have a look at give me your thoughts.

Image Frame.fits Flat Frame.fits

Link to comment
Share on other sites

1 hour ago, Clarkey said:

A bit late, but here are some individual subs. One flat frame and one image frame.

I am seriously considering binning this camera. I tried some more imaging with the same result. What makes it worse is the sensitivity seems so poor. The trouble is, I don't want to sell it if there is something wrong with it (which I suspect), which means I now have a £1000 paper weight. I could sell it and for the money buy another second hand 1600mm pro. At least these work.

@ONIKKINENand @tomato - please feel free to have a look at give me your thoughts.

Image Frame.fits 49.49 MB · 2 downloads Flat Frame.fits 49.49 MB · 3 downloads

Flat frame looks ok, would expect that to work. Light frame looks terrible, almost as if it was taken under a full Moon or under heavy light pollution. Very low SNR with barely any target visible, but looking at a light pollution map your location should be pretty good. Lightframe looks also exposed long enough with at least 25 electrons of median signal per pixel which is enough to swamp read noise x5 if you used HCG mode.

But these 2 alone wont explain why the flats did not work, can you post the 2 most likely culprits - a dark and a darkflat.

*Divided the lightframe with the flat frame and looks like these are not matching. From different nights, different orientations and looks like different OAG positions? Anyway, these wouldnt work even if the overcorrection were fixed.

Edited by ONIKKINEN
  • Thanks 1
Link to comment
Share on other sites

I did some pixel math for your 2 frames here. I subtracted 500, which is the offset in your fits header, from both the light frame and the flat frame and voila, working flat frame. Of course its not perfect since they are not mechanically matched and there is only one flat frame but the conclusion here is that you 100% have some issues in your darks/darkflats/both.

Synth calibrated frame without the bright corners (but doesn't work as a flat anyway):

2022-12-13T22_18_54.thumb.png.be5714cc01fb431ba09a331390b07034.png

Link to comment
Share on other sites

11 minutes ago, ONIKKINEN said:

dark and a darkflat.

I'll post them tomorrow when I am back at the PC. You have made me think though - the offset on the darks might be different, which will have an effect. (They were taken a while back and I might have altered the offset since)

Don't take too much notice of the light pollution map. It claims I am B4 but 5 to 6 is more realistic. I think this image was when there was some moon, but I am careful to keep the moon as far away as possible. If the poor images are linked to the moon or light pollution, the camera is good as useless to me. I have imaged with a mono camera over a full moon (in RGB using the same scope) without too much problem. Why should this one be so bad?

Link to comment
Share on other sites

2 minutes ago, Clarkey said:

I'll post them tomorrow when I am back at the PC. You have made me think though - the offset on the darks might be different, which will have an effect. (They were taken a while back and I might have altered the offset since)

Don't take too much notice of the light pollution map. It claims I am B4 but 5 to 6 is more realistic. I think this image was when there was some moon, but I am careful to keep the moon as far away as possible. If the poor images are linked to the moon or light pollution, the camera is good as useless to me. I have imaged with a mono camera over a full moon (in RGB using the same scope) without too much problem. Why should this one be so bad?

At bortle 6 skies a single x5 read noise swamped exposure is not really useful as an analysis tool at all. You would need at least a hundred of these to start getting an idea on how the image turns out in the end, and why the 1600MM was better i couldnt tell you. You are comparing a mono camera to an OSC one however, so there is the obvious sensitivity difference there. Even with the higher QE of the IMX571 chip compared to the panasonic something something in the 1600MM the mono one probably still has the edge (if comparing to Lum).

Link to comment
Share on other sites

2 minutes ago, ONIKKINEN said:

At bortle 6 skies a single x5 read noise swamped exposure is not really useful as an analysis tool at all. You would need at least a hundred of these to start getting an idea on how the image turns out in the end, and why the 1600MM was better i couldnt tell you. You are comparing a mono camera to an OSC one however, so there is the obvious sensitivity difference there. Even with the higher QE of the IMX571 chip compared to the panasonic something something in the 1600MM the mono one probably still has the edge (if comparing to Lum).

Actually i peeked into the .FITS header some more. You had an 82%!! illuminated moon at the time of shooting the sub. Cant expect anything out of a single sub like that and this one is definitely not the cameras fault.

Link to comment
Share on other sites

Here are screenshots of a light and flat frame from my RCIMX571c, the light is a 120 sec exposure, gain 100, offset 768, taken under a full moon. The flat is 3.18 secs exposure so not too dissimilar to yours. 

Light

Tomatolight.thumb.jpg.4b1528c9db80f99a8682afe91e88ec14.jpg

Flat

TomatoFlat.thumb.jpg.22c41244d04e4ccea9adf5acc14417d7.jpg

 

Here are your frames:

Light

ImageFrame.thumb.jpg.d27857c645b984ec3fb725dec8be8c9f.jpg

Flat

FlatFrame.thumb.jpg.a3e12249f4582738c00e997cb76269bf.jpg

 

Qualitatively they look very similar to me, both lights have strong LP gradients. However,  the dust motes look strange on your flat, they are rings rather than filled circles, which might suggest that the focus has shifted between the light subs and the flat frames?

I can't ever recall seeing dust motes as ring shadows on a flat frame, but I could be wrong.

It's a shame we are not a bit closer then we could swap cameras, but on the face of it I can't see anything significantly awry with your camera's performance.

Edited by tomato
Moon illumination note added
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.