Hot pixels in preprocessed image help!!!!

[sctall]

Hi

I have an issue with Hot pixels ( at least that's what I think it is), after processing with BPP.
I used BPP to get all the subs preprocessed, but when I look at the final preprocessed image, I see all these red pixels.

I am not imaging or preprocessing anything different than I normally do.  Not sure where they are coming from. I use a DSLR with a LP filter. I do not see the pixels near as bad in a single light frame, so I think it is coming from the preprocessing.
Would it be bad darks or bias's?

I tried to use the manual way of preprocessing, but during the final II process, I kept getting "no coorelation" issues, and not enough stars issues, that failed the process.  So I went to BPP.  This was successful, but I ended up with this image.

I tried the Cosmetic Correction with the BPP script, but  I got a CFA warning message. (I don't remember exactly what it said)
I made a master dark from the .CR2 files using the II, added that into the CC. I then loaded a light frame, and when I adjusted the sliders, I saw no change in the Hot pixels. So I took a guess at the values by Harry's tutorial, since I saw no diff.
I made a process and added it into the BPP script. When I did the diagnostics, I got a CFA error message, and could not run the script.

Man I hope I did not lose all this time.

Scott

[georg.viehoever]

Since these are DSLR images:
- dont do dark optimization. It does not work for DSLR (discussed elsewhere).
- Make sure your darks have approximately the same exposure time as your lights. That's usually sufficient.
Georg

[MikeOates]

Scott,

One way to reduce the noise with DSLR images, which I can see as vertical banding on your image, is to use a lot of bias frames, how many are you using? I would suggest you need about 150 some do 200. It makes a huge difference.

Mike

[Juan Conejero]

- dont do dark optimization. It does not work for DSLR (discussed elsewhere).

I wouldn't state this as a general rule. I have DSLR data sets where dark optimization works perfectly, giving dark optimization factors around 0.8 - 0.9 consistently.

... use a lot of bias frames, how many are you using? I would suggest you need about 150 some do 200. It makes a huge difference.

This is very true for CCD and DSLR images.

[sctall]

Georg and Mike thanks

I re-ran the BPP with no dark optimization and it came out better.
I still think it is a bit noisier than normal, but I still don't understand how to measure one image from another. I just look at it and can see.
I usually take 4 min exp with ISO 800, but I decided to take 4min at ISO 400, and for some reason the preprocessed image turned out worse.

Being all things the same including my LP, I am pointing to the calibration causing this somehow.
Since I am still digesting this whole concept of calibration, and what can cause what and, understanding PI in general, any ideas of causes are welcome.

Mike I am going to make a master Bias of 100 frames. This one only had 30.  I read that ISO match for bias is important. True? And why?
It is only a very short Dark basically.

Juan, when you say "dark optimization factors around 0.8 - 0.9 consistently", how is that measured, and are you talking about just the "dark" frames?

Also is it possible to put out some doc or ref to explain all the numbers that show in the console during preprocessing steps, to better understand
and or compare to determine if the subs are good.
There are so many, but I am sure there are some pertinent or more important one to concern with.

Being a non-math or statistical individual it is daunting to say the least.
But just knowing what ones to watch for, concerning noise, focus, SNR, etc, would sure be nice.
I have been trying to learn, but just have not got it.
I know you guys are perfectionists, but us lamens are just happy getting it to start with.

Scott

[MikeOates]

Mike I am going to make a master Bias of 100 frames. This one only had 30.  I read that ISO match for bias is important. True? And why?
It is only a very short Dark basically.

I know you guys are perfectionists, but us lamens are just happy getting it to start with.

Scott, Yes, use the same ISO as the main exposures, you need to re-create the same conditions when taking bias frames as when you took the subjects sub frames. This is so that you can get the same noise levels which can then be subtracted from your subs. The same also applies to darks. Remember the bias frame is the shortest exposure you can do with your camera and only use with a cap on the scope to block out any light. It may also help if your doing this in daylight to cover the camera to keep out any stray light especially the DSLR viewfinder.

Bias and darks do not need to be taken every time your out imaging, do them once and re-use for maybe weeks or even months if the conditions don't change, i.e. ambient temperature or ISO settings. But do take flats every time, these are very important to even out the background and remove dust motes etc. You can take flats at a different ISO, in fact using a lower ISO is preferable as this will prevent extra noise being added to the final image from the flats.

I am no expert and consider my self a novice as well, only used PI for about 18 months just after I started taking DSLR then CCD images.

Mike

[sctall]

Thanks Mike

I have read that the exposure time is not critical for the darks.
Is this true and valid with DLSR? It sure would make the darks easier to get.

If I image the lights at 5 min, could I take shorter darks, so long all other conditions are the same as the lights?

And if that works, I understand that the dark optimization must be checked.

Scott

[pfile]

if you go back to basics, darks should be the same length as the lights.

it is possible to scale a calibrated dark, but generally people use longer darks than their lights, and the darks are scaled downward.

rob

[sctall]

OK
But I was referring to this statement in the PI docs

Acquisition Notes

As we'll see in this article, the dark scaling algorithm implemented in the IC tool works better when read noise is negligible in the master frames. IC is optimized to work with calibration libraries: don't worry about differing temperatures and exposure times between dark and light frames. IC will always rescale the dark noise to match every light frame.

Scott

[pfile]

well it's one of those things, some people think scaling darks is a bad idea, others think it's an awesome idea.

rob