Problems with Bias and a single cyan band on DSLR

[Photosbykev]

Good morning folks,

I was shooting M31 on Tuesday evening just to test some dithering settings with my setup (stock Canon 5D MkIV and WO Zenithstar 61 with flattener on a SkyGuider Pro mount shot through an IDAS-D2 filter) Not a perfect evening with the 85% moon coming up during the short test but good enough for what I wanted to do.

The image below is from 20 x 240s lights, 20 flats, 20 darks and 50 bias frames batch preprocessed in PixInsight and then a simple stretch transfer function to pull up the detail. The image is uncropped, full frame and just resized down to 2048px wide.

There is clearly an issue with the bias removal but more significantly there is a horizontal cyan band along the bottom which I've never come across before.

I've also included a short video (if it plays) showing the Ra dithering movement over the twenty lights which I suspect is too much movement but I can sort that.

Any help with the bias and cyan band would be appreciated.

best regards

Kev

Edited September 19, 2019 by Photosbykev

[rockinrome]

Could it be an internal reflection? Long shot I know, but do you cover the camera eyepiece with the rubber *thingy* that should be on the camera strap....?

All the best.
MJ
 

[david_taurus83]

I have to say that's an impressive flat field for such a small scope and a full size sensor!

The dither movement looks fine. You need quiet a sizeable movement with a DSLR. Try stacking only the lights, bias and flats. Dithering can sometimes do away with the need for darks.

 

Also, try cropping out that bottom stacking artefact, run DBE and then try the Canon banding script.

[Photosbykev]

1 hour ago, rockinrome said:

Could it be an internal reflection? Long shot I know, but do you cover the camera eyepiece with the rubber *thingy* that should be on the camera strap....?

All the best.
MJ
 

I didn't MJ, it's not something I've ever done and I was in a dark location so not much to reflect tbh

[Photosbykev]

1 hour ago, david_taurus83 said:

I have to say that's an impressive flat field for such a small scope and a full size sensor!

The dither movement looks fine. You need quiet a sizeable movement with a DSLR. Try stacking only the lights, bias and flats. Dithering can sometimes do away with the need for darks.

 

Also, try cropping out that bottom stacking artefact, run DBE and then try the Canon banding script.

I'm very very impressed with the 5D4 and WOZ61 and flattener, it seems to be a well matched combination.

I'll play with the restacking options and then DBE and the banding script but I would like to remove the source of the banding rather than remove it in post

Kev

Edited September 19, 2019 by Photosbykev

[Photosbykev]

2 hours ago, david_taurus83 said:

I have to say that's an impressive flat field for such a small scope and a full size sensor!

The dither movement looks fine. You need quiet a sizeable movement with a DSLR. Try stacking only the lights, bias and flats. Dithering can sometimes do away with the need for darks.

 

Also, try cropping out that bottom stacking artefact, run DBE and then try the Canon banding script.

Following up on this comment. I stacked and integrated light/flats/bias frames (excluding darks) in APP. In PI I cropped the bottom to remove the nasty bias edge, I guess I'm doing something wrong there, maybe removing the bias twice? I need to read up on this but not using the darks doesn't appear to have a significant impact at all, I guess dithering does work

I ran DBE to remove the minor vignetting and then the Canon banding removal script followed by Photometric colour correction. The image below is a result of just a simple stretch transfer function resized down to 2048 pixels wide.

Kev

[michael8554]

The band looks electronic in source. 

Go through all your sources looking for one with a fault. 

Michael 

[Photosbykev]

1 hour ago, michael8554 said:

The band looks electronic in source. 

Go through all your sources looking for one with a fault. 

Michael 

I'm separating out all my camera leads tonight lol

[alacant]

12 hours ago, Photosbykev said:

nasty bias edge

Hi

Not sure what this is. Is it the bias that is causing the banding?

Are you sure it's not simply artefacts left after aligning the frames during stacking? Leaving edge aretefacts -particularly as you dithered- before stretching may produce the type of localised banding along one edge.

HTH.

Edited September 20, 2019 by alacant

[Photosbykev]

2 hours ago, alacant said:

Hi

Not sure what this is. Is it the bias that is causing the banding?

Are you sure it's not simply artefacts left after aligning the frames during stacking? Leaving edge aretefacts -particularly as you dithered- before stretching may produce the type of localised banding along one edge.

HTH.

Im pretty sure the banding is related to the camera, maybe cabling interference as I can just make out the banding on each 4 minute sub without aligning/stacking. I'll have a session tonight with all the cables repositioned and see if it's still visible. 

The poor bias subtraction along the bottom edge is, I think, an error in calibrating the images but I mnot sure where I'm introducing the error yet. 

 

Kev

[Photosbykev]

I finally sorted out the bias issue, it was being subtracted twice! and below is the final image from 20 x lights/flats/darks and 50 bias images at iso 800 4 minute exposures. More exposure time, subs and a new moon required to really pull the detail out.

Image calibration details were:
Center (RA, Dec):    (10.519, 41.372)
Center (RA, hms):    00h 42m 04.521s
Center (Dec, dms):    +41° 22' 17.713"
Size:    5.43 x 3.63 deg
Radius:    3.265 deg
Pixel scale:    9.54 arcsec/pixel

The final image has a minor crop of around 50-100 pixels around the border to remove the dithering artifacts but I'm really pleased with the corner stars on a full frame camera.

[teoria_del_big_bang]

A very nice image. Thats a massive improvement, so pleasing when something eventually goes to plan 🙂 

Steve

[michael8554]

23 hours ago, Photosbykev said:

I'm separating out all my camera leads tonight lol

Glad you got it sorted. 

Just for reference, by sources I meant stacking sources - lights, flats, Bias etc, not Leads. 

Michael 

[Photosbykev]

1 hour ago, michael8554 said:

Glad you got it sorted. 

Just for reference, by sources I meant stacking sources - lights, flats, Bias etc, not Leads. 

Michael 

Well I seriously misread that piece of advice lol.

I use blink in PI to scan every image and then remove any with satellites/planes in them and also any image with distorted stars (touch wood I don't get many of those). I shoot flats for every session so I don't think the issue is there. I think I'm having finger trouble in selecting options during calibration/registration of the data which I need to figure out. 

Currently shooting 3 hours of subs on the Veil with all the cabling reconfigured so I'll have some fresh data to test with. 

Kev

Edited September 20, 2019 by Photosbykev

[wxsatuser]

9 hours ago, Photosbykev said:

Well I seriously misread that piece of advice lol.I use blink in PI to scan every image and then remove any with satellites/planes in them

Kev

You don't need to remove these unless the plane ones are really close.
Correct stacking and more than 12 subs will get rid of any satellites and most planes.

[Photosbykev]

3 hours ago, wxsatuser said:

You don't need to remove these unless the plane ones are really close.
Correct stacking and more than 12 subs will get rid of any satellites and most planes.

Thank you FWIW I shot the Veil last night 40x4min subs and the cyan banding was still present in all the subs. Plenty of detail for a stock Canon 5D4 but the banding is strange

Kev

[wxsatuser]

I suppose it's not the mirror when its up, it might just get in the light path.

I had a similar thing on the 6D with a Astronomik lp filter in place it just left a band
 but if I remeber right it was'nt coloured.

[Scooot]

16 hours ago, Photosbykev said:

Well I seriously misread that piece of advice lol.

I use blink in PI to scan every image and then remove any with satellites/planes in them and also any image with distorted stars (touch wood I don't get many of those). I shoot flats for every session so I don't think the issue is there. I think I'm having finger trouble in selecting options during calibration/registration of the data which I need to figure out. 

Currently shooting 3 hours of subs on the Veil with all the cabling reconfigured so I'll have some fresh data to test with. 

Kev

Don’t know if this might help at all but I saved some screen shots as a guide I use to calibration/registering. I don’t use the batch processing.

So integrate the bias & darks with these.

Calibrate Flats with these settings.

Integrate Flats with these. (Usually winsorised clipping though depending on number of frames)

Calibrate Lights with these settings.

[Photosbykev]

6 hours ago, wxsatuser said:

I suppose it's not the mirror when its up, it might just get in the light path.

I had a similar thing on the 6D with a Astronomik lp filter in place it just left a band
 but if I remeber right it was'nt coloured.

I would be very surprised if it was. The IDAS-D2 filter is inside the Flat61A flattener so in theory the mirror should be well clear of the optical path. It certainly isn't visible in any flat images