Are these flats working? Sorry to have to ask.......

[swag72]

I'm sorry, I know this gets asked a million times, but I am getting rather near the end of my tether with this now.

I have finally decided to take some flats, and as I use Maxim for stacking I thought it would be nice and easy to add them. I go through the calibration wizard, add the flats in and create a master flat - So I know that bit is right at least. Then I stack the images, and tick auto calibrate, so that my calibration master flat should work with the final stack.

Well, I don't think it does. So take a look here and tell me if you agree. The first image is the master flat in Maxim. The second is the stacked and stretched image (with flats added) and then the resultant gradient that DBE removed in PI. Surely there should be no gradient to remove if the flats have done their job? Or am I just expecting too much from a set of flats?

[dph1nm]

Looks very much like the DBE gradient is following M33 itself. I have found DBE rather difficult to get to work when there are large objects in the frame - needs a lot of manual adjusting (this was using Pixinsight LE).

NigelM

[swag72]

So you are saying that I am removing something of a natural gradient and not one caused by vignetting?

Here's my puzzle - Here's 2 stacked images with a quick PI stretch done on them. Can you see a difference? I can't hence I don't think the flats have worked.

With flats

Without flats

[Coco]

Sara if your using Artemis you can tell if your flats are correctly exposed by checking the ADU value on the lower frame of the capture window, if you move your mouse about over one of your flat images after it has appeared in the capture window you should be looking for an average value between 20,000 & 25,000 dont assume that becasue you can see see bunnies in the image in maxim that the ADU is enough.

by increasing and decreasing the exposure length you can fine tune the exposure value, dont bother doing this at 'sun up or sun down' as you will chase the value and end up not getting anything.. far easier to use an Ipad or lap top screen in the dark as a light panel, even better buy a posh light panel

[swag72]

These were taken using a light panel and Artemis. They exposed at approx 30k ADU, which I believed was OK.

Can you see a difference inbetween the stacked images? I would expect to see something different if the flats are working.

[davew]

I can see and measure a difference.

I took both into PS and used the info tool set at 5x5. The calibrated image is much flatter.

Dave.

Did you use the DBE tool on auto ? I agree with Nigel, it's trying to remove the Galaxy.

[ollypenrice]

I haven't measured it but clearly 'with' looks far better with more contrast between faint galaxy and background sky. (This is one of the plusses with flats.)

In Artemis you don't need to scroll over the image, I don't think. Just look at the black and the white points. The white should be around 20 to 30K and the black somewhat lower. The range (which will be a few thousand ADU) is the erronious variation that the flats will correct.

You can get a gradient in flats, though. I do them in decent darkness even with a panel. How the gradient gets in by day I simply don't know but it usually does! If I don't get a clear darkening in all the corners I know they are not going to be right.

Olly

[A320Flyer]

Sara,

You dont say which camera you used, but, if you have used the OSC camera you need to tick the option in the Set Calibration window to use a Boxcar filter on the flats. This is needed to destroy the bayer matrix on OSC cameras, otherwise you get all sorts of funny results. Just a thought!

Bill

[swag72]

Cheers Dave and Olly - I am pleased that you can see a difference, I still can't!! I am going to retake them tonight, in the dark and with a slightly lower ADU of around 20-25k. Restack them and see if I can see a difference.

@Bill - It's a mono, so no Boxcar ticked!

[Spikey]

I can see a difference in the pic that has had the calibration. M33 stands out more from the background.

[ollypenrice]

I can see a difference in the pic that has had the calibration. M33 stands out more from the background.

Exactly so.

Olly

[swag72]

Cheers Olly, it is your fault that I am doing these damn things!!!! They had better make a difference!!

[Spikey]

Don't forget to take flats for each filter you use to make the final image. I'll duck for cover now

[harry page]

Hi

It looks like you are leaving samples on m33 , the samples in DBE should be on the free background if possible

Harry

[swag72]

It was just a quick DBE as I wanted to see if any gradient was left after using flats. It was the only way (flawed I now know) I thought to test if the flats has worked. Unfortunately, my belief that flats worked like magic and that after seeing 2 frames (one with and one without calibration frames) I'd be blown away by the improvement was rather wrong. Maybe I need to be just a little more realistic in my expectations!

[harry page]

Hi

Flats will only remove illumation errors caused by your scope and imaging train , I have not seen a broadband image that can not be improved by the correct use of DBE , even olly's dark site images

Harry

[ollypenrice]

Hi

Flats will only remove illumation errors caused by your scope and imaging train , I have not seen a broadband image that can not be improved by the correct use of DBE , even olly's dark site images

Harry

Quite true, Harry, and I use DBE on most images. However, that is very different from using DBE on an image which has not been flat fielded. Here you are asking it to do the impossible - like remove a dust bunny half embedded in faint nebulosity. In fact at the moment all images from Yve's instrument are entirely flattened in DBE because we haven't yet got the flats to work. I don't regard this as remotely satisfactory, though.

I regard both flats and DBE as essential, certainly for RGB however good your flats are. Sometimes my Tak images in L and Ha don't go through DBE but the RGB always does.

Olly

[harry page]

Hi

I agree , I did not mean DBE will replace flats , Its just that DBE is still needed in most cases after flats

Harry