So what am i doing wrong now?

[martin_h]

30 subs in DSS - 28 out of the 30 stack ok, but when I add dark,flats and bias frames only 1 out of the 30 is used - 'cant find stars' I've set the threshold down to its minimum 2% and it only finds 1 star!

Any ideas people?

[Psychobilly]

let us have a look at a single sub please...

[martin_h]

let us have a look at a single sub please...

There you go! Its a JPEG version

[Psychobilly]

Ok the lights are fine and not showing to many hot pixels... hmmm

I'd be inclined to have a look at the Dark , Bias and FLat frames next - the Master ones (tiffs)

Sometimes the "calibration" frames can do wierd things as you have found out...

What camera are you using and what exposures etc...

Peter...

[martin_h]

The camera is a QHY8 and they were 3 min subs, darks the same, but the flats were only 0.01 second, I need to dim down the EL panel a bit, I've got 2 sheets of opal perspex but need some A3 paper now..Its probably the flats that are messing things up

[Psychobilly]

I tend to take a step backwards and stack the lights with the calibration files one type at a time until i find the culprit...

Mis-matched darks can be one of the biggest causes of "problems"...

[martin_h]

Its deffinatly not the darks - I took a new set last night - so they are matched to the lights........what should a master flat look like, just took a look at mine and its nearly black all over?

[Psychobilly]

The ones i took with the 50mm the other night and the camera on auto are about 70%grey level...

How are you setting the exposure for the flats...?

[martin_h]

Taking the mean pixel reading in Nebulosity 2....around 2200

1st pic is a single flat

2nd pic is the master....the mean pixel reading comes down to 30!!

[Psychobilly]

There has been a considerable debate ont he best "level" for flats on here and other forums .. I use DSLR's rather than CCD's so I just switch to Av mode and shoot away...

Something wierd going on there though and almost certainly the cause of your problem

[martin_h]

with a flat like that its no wonder DSS cant find any stars!

[martin_h]

I've just done a master flat in Nebulosity 2 and its totaly different to the DSS flat!

[SteveL]

Taking the mean pixel reading in Nebulosity 2....around 2200

1st pic is a single flat

2nd pic is the master....the mean pixel reading comes down to 30!!

Thats very very low, you should be looking at 20,000-30,000 mean pixel count for your flats, or somewhere pretty close to it.

[Darth Takahashi]

Yes that's right, I have read it should be half of the full well value, for a 16bit camera that's approx 32000 ADU'S.

For a 14bit camera its 8192 ADU's. Not sure how many bits your camera uses to record the image data!

[ollypenrice]

I don't use a DSLR but that flat is still showing the Bayer matrix, which it certainly should not in the workflow I use. What I do (in AstroArt) for OSC CCD is go to Image-Shift and set the x and y axes to 0.5

This removes the Bayer, then I add a light Gaussian blur to reduce noise.

For OSC CCD flats I don't use a dark in the normal sense. Following advice from Harry Page I use a master bias (about 50 frames combined) where I would normally place a master dark. That is just for the flats, I stress.)

As for flats ADU, I go to one third of the full depth. In my case about 22,000 ADU - but I can't say that a bit either side makes any difference that I have seen.

Olly

[martin_h]

Ok so I'm a plonker, I've been reading 2200 as 22000 !!!!!! Well it was 4 AM and I have a raging tooth ache. So I'm going to ditch those flats and do another set tonight.

I'll still have the tooth ache but hopefully I'll be able to process some images better tomorrow.

Does DSS take care of the Bayer matrix when making a master flat?