Same flats process, now not working!

[Rustang]

What are the most common reason's for flats not to work?. I have been using the LED light pad method for a short while with good results. Today I took some Flats using the same method and with the camera, light pollution filter and scope in exactly the same position of focus from last nights session but they haven't worked to remove any artifacts as they are showing on the stack. I'm back to using DSS until I can afford APP.

Edited October 10, 2020 by Rustang

[bottletopburly]

Not bright enough possibly ,what is your setup,  camera  ,scope etc what length was exposure  what was histogram showing 

[Rustang]

Scope is SW80ED DS Pro at 600mm FL with none reducing FF. Canon 600D with Optolong L-pro LP filter. Exposure was at a shutter speed of 200 to hit the histogram at approx one third 👍

Edited October 10, 2020 by Rustang

[andrew s]

Have you tried comparing an old working flat with the new failing one. They should not be that different just the odd dust mote. Images might help us see what might be happening.

Regards Andrew 

[alacant]

4 hours ago, Rustang said:

they haven't worked to remove any artifacts

Hi

Post -links to- an example flat frame which works and an example of the latest, which doesn't.

Cheers

 

[Rustang]

I will post tomorrow, a flat frame taken today and the stack stretched to show the artifact and one that has worked as well 👍

[Rustang]

49 minutes ago, andrew s said:

Have you tried comparing an old working flat with the new failing one. They should not be that different just the odd dust mote. Images might help us see what might be happening.

Regards Andrew 

Will post up some info/images tomorrow 👍

[Rustang]

Below are 3 images, first is a single raw flat that didnt work, second shows the stack over stretched and b&w and slightly cropped, to see firstly the sensor dust wasn't removed and also the lines (could be my normal banding issue) and gradient that seems difficult to get rid of. There shouldnt really be much gradient as the target was practically at its highest point in the sky!. Lastly the flat taken that worked before.

Edited October 11, 2020 by Rustang

[Rustang]

Also to add a photo of the camera screen of a single flat from the batch that didn't work. It sits dead on a 3rd of the histogram. 

[alacant]

Hi

These seem to be .jpg files. Please send links to: 

- a single raw flat file that worked

- a single raw flat file which didn't work 

One observation: you need to reduce the light from the LED panel so your exposures are at least 1s.

Cheers

[Rustang]

Ok il post the raws later. I thought shutter speeds didn't matter as long as the histogram sits between 1 3rd and half way!? I never heard of the shutter needing to be around 1 second!? From what I've read previously, if you stick the camera in AV mode it will choose the correct exposure, to be fair I normally do it manually. 

 

Edited October 11, 2020 by Rustang

[ollypenrice]

Flats won't remove gradients from the sky and, even at my site (SQM reaches 22), I have residual gradients on almost all broadband images, which I fix with DBE in Pixinsight.  Even on the zenith you're pretty well bound to get them. Nor will flats correct for dust actually on the chip because the affected pixels receive no light.

For all that, I have, like you, had occasions when my flats would suddenly stop working and I never did get to the bottom of it. The problem was unique to a rig which was capturing in Nebulosity, which makes me think it was to do with the capture process. I shot a set of flats capturing in AstroArt and these worked perfectly for some time - and then didn't.  Mystery.

Olly

Edited October 11, 2020 by ollypenrice

[alacant]

1 hour ago, Rustang said:

the shutter needing to be around 1 second!

The light panel refreshes many times per second. Any capture shorter than the refresh rate and you'll get only part of the light; uneven. 

Remember that this is astrophotography. One can never know, predict or explain. Anything. It's magic!

Cheers

[Rustang]

1 hour ago, ollypenrice said:

Flats won't remove gradients from the sky and, even at my site (SQM reaches 22), I have residual gradients on almost all broadband images, which I fix with DBE in Pixinsight.  Even on the zenith you're pretty well bound to get them. Nor will flats correct for dust actually on the chip because the affected pixels receive no light.

For all that, I have, like you, had occasions when my flats would suddenly stop working and I never did get to the bottom of it. The problem was unique to a rig which was capturing in Nebulosity, which makes me think it was to do with the capture process. I shot a set of flats capturing in AstroArt and these worked perfectly for some time - and then didn't.  Mystery.

Olly

I thought that one of the reasons for using flats was to remove the sensor dust marks!?, it has appeared to do this every other time!!?

Edited October 11, 2020 by Rustang

[Rustang]

First is non working flat, secound is the from when it worked!

IMG_7303.CR2 IMG_7056.CR2

[ollypenrice]

39 minutes ago, Rustang said:

I thought that one of the reasons for using flats was to remove the sensor dust marks!?, it has appeared to do this every other time!!?

Normally the dust with which flats deal successfully is on some surface forward of the sensor in the light path. Chip window, filter, flattener, etc. These produce doughnut-like shadows which flats correct. A contaminant (unless very tiny indeed) on the chip itself will block light from reaching a pixel and, since the flats won't have any signal there either, a division of light by flat values will be worthless. Flats only work where the pixels receive some signal. The small dark circle you point to with your arrow will not be from dust on the sensor. It covers many pixels, is not completely dark and is round. It will be on a surface above the sensor but, being small, it will be close to the sensor. (There's a formula somewhere for calculating the distance of the contaminant from the chip. I dare say it would Google but the answer probably wouldn't be much help. The link I had on file is no longer active.)

I'm wondering if the unsuccessful correction comes from the possibility that the contaminant is slightly reflective and responds differently to the bright light of the panel than it does to the dark night sky. It might be worth dimming the panel with typing paper and upping the exposure time as Alacant suggested for other reasons.

Olly

 

 

[alacant]

Hi

You can see that the optical train has changed between flat frames being taken. The first shows the shadow of the camera's flip mirror. The second, doesn't.

You can also see evidence of too short a shutter speed -uneven lighting- on the bottom quarter of image 1.

HTH

 

 

 

Edited October 11, 2020 by alacant

[Rustang]

Ok thanks guys, il have another go at some flats today and look for any artifacts on different areas of the optical train but it is exactly where there's a dust spot is on the sensor which is strange. Cheers 👍

[Rustang]

5 minutes ago, alacant said:

Hi

You can see that the optical train has changed between flat frames being taken. The first shows the shadow of the camera's flip mirror. The second, doesn't.

You can also see evidence of too short a shutter speed -uneven lighting- on the bottom quarter of image 1.

HTH

 

 

Ok, I will look to improve the flats today. The optical train /focus point will be different between the two sets of flats as different field flatteners were used for the two different imaging sessions. Thanks for your help 👍

Edited October 11, 2020 by Rustang

[alacant]

4 minutes ago, Rustang said:

The optical train /focus point will be different between the two sets of flats as different field flatteners were used

Ah;)

[Rustang]

1 hour ago, alacant said:

Ah;)

Ive been having a play around with trying to do some new flats, it seems that with the lowest setting on the LED panel which already has one layer of paper across it, the slowest shutter speed I can go to is 1/50th, is that going to make any real difference? Shall I just do a few a see if it stacks any better? If i cant get it to work I guess I'm going to have to yet again have another look at a different way to do them.

[alacant]

9 minutes ago, Rustang said:

has one layer of paper

Dim the panel and/or put more sheets of paper until you get at least one second. Do it in a dark room to ensure no extraneous light enters the camera.

 

Edited October 11, 2020 by alacant

[Rustang]

10 minutes ago, alacant said:

Dim the panel and/or put more sheets of paper until you get at least one second. Do it in a dark room to ensure no extraneous light enters the camera.

 

I did just test it with the 1/50th flats and the strectched stack now appears with out the spot to the top left but will do as you say aswell 👍

Edited October 11, 2020 by Rustang

[Rustang]

34 minutes ago, alacant said:

Dim the panel and/or put more sheets of paper until you get at least one second. Do it in a dark room to ensure no extraneous light enters the camera.

 

I have attached a new flat taken at 1.3 sec's, any chance you could stretch it out like how you did above to see if there are any issues!? 🙂

IMG_7375.CR2

Edited October 11, 2020 by Rustang

[alacant]

24 minutes ago, Rustang said:

like how you did abov

Much better. Now it looks like a proper flat frame:)