Huge green bias

General discussion about StarTools.
patraip1
Posts: 25
Joined: Sat Oct 16, 2021 8:15 am

Re: Huge green bias

Post by patraip1 »

I have a homemade lightshroud which is lightproof (will post a picture of my setup)
I take the darks at my backyard right after the lights with the scope cap on
I then bring the rig inside (6-7 kgs alltogether, focuser locked in place tightly anyway) and take my bias in a dark room with the cap on
I then put the tshirt on and my laptop screen playing a 24hour youtube white screen and figure out the exposure in the same dark room
I finally take the darkflats in the same (pitch black with the shutters down ,door closed) room
I use the camera intervalometer for all.
IMG_20211025_130358.jpg
IMG_20211025_130358.jpg (193.03 KiB) Viewed 3998 times
IMG_20211025_130505.jpg
IMG_20211025_130505.jpg (194.33 KiB) Viewed 3998 times
IMG_20211025_130349.jpg
IMG_20211025_130349.jpg (176.75 KiB) Viewed 3998 times
patraip1
Posts: 25
Joined: Sat Oct 16, 2021 8:15 am

Re: Huge green bias

Post by patraip1 »

Ok so I think the reverse vingetting issue is definetely related to DSS settings.
I tried the same data in sequator and although the result had a lot more noise and huge stacking artifacts at the corners there was no reverse vingetting!!
I tried unchecking the set black point to zero in the RAW settings of DSS, unchecked both white balance options and used average in the first tab, kappa sigma in lights, median kappa sigma in flats and darks and automatic stacking. I used all my calibration frames flats, dark flats, darks and bias and stacked the best 257 subs based on score.
The result was much much better showing only a light pollution gradient and an artificial light source at one corner (street light down the street which I try to block with a board) but no reverse vingetting and no data clipping at the center and m31 showing much more detail!!
I tried the same configuration without darks but it showed signs of the problem. So for the time being I will stick to these settings and try to adjust my workflow in ST. Up to now I only got to autodev in all my stacking attempts until I was happy with the result.
I will keep you posted....

Here is all my data for anyone to try to stack and process

https://drive.google.com/folderview?id= ... X5mGFuX2mp
Last edited by patraip1 on Mon Oct 25, 2021 3:16 pm, edited 1 time in total.
happy-kat
Posts: 372
Joined: Sun Feb 01, 2015 11:31 am

Re: Huge green bias

Post by happy-kat »

as darks are darks then after the lights and flats, you can remove the camera to do bias, darks and dark flats. use camera body cap and could put camera in bag or cover to ensure no light ingress
patraip1
Posts: 25
Joined: Sat Oct 16, 2021 8:15 am

Re: Huge green bias

Post by patraip1 »

Happy kat, thanks for the suggestion....
Had already thought of doing it this way but I prefer to not move the camera at all so that it is ready for my next session (this way my flats and bias should be valid for next few times at the same ISO) ... Besides it is the same room I store my rig and it gets 100% dark when I close the door and shutter
patraip1
Posts: 25
Joined: Sat Oct 16, 2021 8:15 am

Re: Huge green bias

Post by patraip1 »

Unfortunately after processing in ST I cannot find a way to fix the colors once again....

Here are all my data for anyone who would like to take a look...

https://drive.google.com/folderview?id= ... X5mGFuX2mp
Mike in Rancho
Posts: 1141
Joined: Sun Jun 20, 2021 10:05 pm
Location: Alta Loma, CA

Re: Huge green bias

Post by Mike in Rancho »

Ran a few tests but no answers yet. :confusion-shrug:

I even split the lights into two sets, since there was a decent sized field shift in the middle of them, to see if I could stop the gradient/vignetting areas from being smeared around so much. It helped, but still there.

Of course even the total here is only 16 minutes or so? So only so much of M31 is going to be able to be revealed. Still, I didn't really have a major color problem...just that vignetting/gradient and short integration.

Also, though they really should be almost the same except for any thermal and fixed patterns your camera might have, you didn't upload any darks. The files in your darks folder were the same files that were already in the bias folder.

In any event, the big question is...why aren't the flats doing what they should be? :think:
patraip1
Posts: 25
Joined: Sat Oct 16, 2021 8:15 am

Re: Huge green bias

Post by patraip1 »

Thanks Mike for diving in...
I am taking new subs at iso 400 as I have read somewhere that iso 200 with my camera at such short subs tends to have similar issues. I will take new calibration files and update

PS The data I uploaded is 40min.(250 lights X10sec =2500sec/60=41,5 min)
16 min was the old iso800 with the green bias

Here are the darks

https://drive.google.com/folderview?id= ... 9_1cgeSkwI
patraip1
Posts: 25
Joined: Sat Oct 16, 2021 8:15 am

Re: Huge green bias

Post by patraip1 »

I am wondering how vingetting occurs in the first place when capturing lights.....
According to this site
http://www.bbastrodesigns.com/NewtDesig ... l#diagonal

my field diameter is 40mm and my camera sensor is 25 by 16mm.

That is with my truss retracted 50mm to reach prime focus.
I will try with a 1,5 barlow lens element that screws on my focuser adapter and I can reach focus without retracting (I will then have 975mm focal length and f/7.5) which should give my also 40mm field diameter at the focal plane, to see if there is any difference.
My single light frames when imported in ST (I converted one to fts in GIMP with 0 compression) do not show any signs of vingetting in autodev whereas stacked light frames alone do, which is buffling.....
Mike in Rancho
Posts: 1141
Joined: Sun Jun 20, 2021 10:05 pm
Location: Alta Loma, CA

Re: Huge green bias

Post by Mike in Rancho »

Didn't check out your link, but, remember there are two types of vignetting, really.

One is hard, physical vignetting, meaning the full extent of the light cone is being blocked by something away from center.

The other is the more typical optical vignetting that we deal with by using flats -- whereby more light accumulates towards the center of the field of view.

Here is one of your single light NEF's, simply opened in Gimp via Darktable, and then just level-stretched while still in Gimp, then scaled and exported to jpg. Notice how the skyglow is bright towards center, but the corners are darker? Vignetting. Flats should correct that out, if they are working properly.
DSC_0638 gimp stretch.jpg
DSC_0638 gimp stretch.jpg (133.09 KiB) Viewed 3898 times
patraip1
Posts: 25
Joined: Sat Oct 16, 2021 8:15 am

Re: Huge green bias

Post by patraip1 »

Ok guys (Mike actually....)
I came across this
https://stargazerslounge.com/topic/2204 ... newtonian/

He seemed to have the exact issue as me. His flats would not correct vingetting.
He has a newtonian pretty much like mine. It seems my way of taking flats with the light source flush against the OTA, is intoducing some light straight to the sensor along with that coming from the optical path.

I really hope this will get me somewhere....
I took some new subs tonight (of the north anerican nebula this time) and after resolving a minor tracking issue with reducing the weight of my counterweight, at the end of the session I took new calibration frames as I had 15 sec subs at ISO 200 (btw iso 400 just overexposes stars and increases noise without solving the vingetting issue....). This time I took my flats aiming at the laptop screen at 1 meter (in a completely dark room and hitting 3/4 BOC histogram at 1/4 sec. I have to say the histogram looked narrower than before.
Anyway I am hoping for the best and will update as soon as I stack.

PS I will also take some skyflats tomorrow
Post Reply