Butterfly nebula (reprocess)

[Carole]

I have done a reprocess on this following constructive comments back in August.

Original post:
http://forum.orpington-astronomy.org.uk/index.php?topic=8515.0

(Reprocess info):
This has been a real pig to reprocess.  Bias frames were causing terrible horizontal banding and the only way I could process at all was leaving them out.  Then I found the flats weren't subtracting properly (possibly connected to not using the BIAS?) so I had to get rid of the vignetting by doing a linear gradient mask, and removing the two dust bunnies manually.

Following a conversation I had with Mark last night, the horizontal banding problem I have only noticed for the last few months, and I realised I re-did my BIAS frames back in June. I'm going to see if I still have my old BIAS subs, and if so, am going to have a process with them and see whether I still get the banding problem.  

This is what I got if I used the Bias frames (I only did a couple of steps and gave up).


Anyway, this is the final result which I am much happier with - reminder 25 x 600secs 400 ISO usual kit with DSLR:


Link to larger file:
http://gallery.orpington-astronomy.org.uk/albums/userpics/10047/Butt_Sun%2C_Mon%2C_Tues%2C_25_x_600secs_400_ISO_NO_BIAS_Process_3.jpg

[JohnP]

who's a clever girl then... Looks great - Never throw away or discard subs no matter how bad especially after 25 x 10mins...

[Fay]

Great improvement Carole!!

[Carole]

Thanks John and Fay.

I was rushing off to the Outreach event when I posted this thread first thing.  we are now back and I have now had time to look for the old Bias subs from 2011 which although I had deleted them from my desktop, I still had them on the backup drive, so am going to give them a stack to try and ascertain whether my new bias frames are what are causing the problem.  Will report back.

Thanks

Carole

[MarkS]

Your re-process looks really good Carole.  It was definitely worth doing.

As for your bias frames, take a look at them and see if there is any obvious difference.  But my guess is they'll be nothing obvious.  Just use the set that works best.

However, I've given up using bias frames because the pattern noise embedded within them changes each time the camera is used.  I use a synthetic bias with fixed values - for the Canon 350D that is an ADU value of 256.

Mark

[Carole]

Thanks Mark, yes it was worth all the "blood and sweat", and things learnt too.  I have done a re-stack with the old Bias frames, and although I do still get banding, it is no-where near as bad as with the new Bias frames. 

I will just have to take each image now and see what works best.

Thanks

Carole

[RobertM]

That looks super Carole !  It was certainly a very worthwhile reprocess. 

[Carole]

Thank you Robert.

Carole

[Mike]

Is it worth doing Bias frames at all? If you are doing Dark frames then the Bias information is stored in them.

I thought that it was only worth doing Bias frames if you were going to use Darks taken at different temperatures or ISO's.

[Carole]

I keep getting conflicting opinions on this Mike.  Some say no need to do them, some say you should do them, some people say the flats won't be applied properly unless you use Bias, others say it makes no difference at all since flats are completely different to BIAS.  I think I am just going to "suck it and see" in future.  Try with Bias and if it's not helping then just leave them out.  I had a similar problem with the HH nebula I did with my Atik314 a few months ago and had to leave out the Bias frames. 

What I am more concerned about is that DSS doesn't seem to be subtracting the flats properly.

Carole

[Mac]

why dont you do the bias frames when you take the images? its not as if they take any time to do.
The Bias frames are just the readout noise from your CCD, whatever the make.

Mac.

[Carole]

Because I was told it wasn't necessary to re-do them that they will last for ages. 
The ones causing all the bother with the HH were ones done at the time and they completely screwed up the image (as above).

Carole

[MarkS]

From a mathematical point of view, for flats to be applied properly you must either supply bias frames or flat darks.  However if the bias level for the camera is known to be zero then bias frames are not required.  If you are not using flats, then as Mike says, the bias information already in the darks is fine.

If you are using flats but don't supply bias frames or flat darks then it's anybody's guess what the software will do but if the bias on your camera is not zero it will not be possible do do the processing entirely correctly.  However the results might look OK to the casual user.

For a DSLR, the bias is not zero, so bias frames or flat darks are definitely required when using flats.

Because I was told it wasn't necessary to re-do them that they will last for ages.  

It's certainly not true for the Canon I'm using. The pattern noise on the bias frames changes every time I use the camera. That's why I've given up and instead use a synthetic bias with constant pixel values across the whole frame.