Bias question

[MineralMike]

I think I know the answer, but I'd like the experts to weigh-in.

In the past, with an ST-8 CCD camera, I would subtract a bias frame from light subs. But my ASI1600MC Pro OSC camera has very low read noise, and subtracting a bias from a narrowband sub results in a totally black image. Here is what PI displays for pixel values in dark regions:

15-minute Ha sub: 0.0134
15-minute master dark: 0.0120
Bias: 0.0119

The sum or the bias and dark is 0.0254. If I subtract this from the sub's 0.0134, the result is negative, so I assume PI substitutes 0.

Question: Is my assumption correct, and should I not subtract a bias frame?

Thanks.

--- Mike

[Niall Saunders]

Hi Mike,

Well - you are already taking Darks of the same exposure time as your (Ha) Lights, so why do you need Biases at all?

Further, if the statistical data for your (Ha) Lights shows ADU values that are at the same level as your Darks, this would suggest that you are just not capturing any photons through your Ha filter. If you can't improve on the SNR (e.g. by exposing for linger), then perhaps you need to start from there.

Failing that, you need to explore more than just these 'numbers' - if you perform a simple interation of your un-calibrated Ha Lights, and then very carefully apply an STF (automatically at first, then manually - by very small amounts - after that), what do you get? Do you see a meaningful image?

[MineralMike]

Hi Mike,

Well - you are already taking Darks of the same exposure time as your (Ha) Lights, so why do you need Biases at all?

Precisely what I was wondering! I had been processing my OSC subs with no bias, but I was curious if including a bias would improve them. Guess not!

Further, if the statistical data for your (Ha) Lights shows ADU values that are at the same level as your Darks, this would suggest that you are just not capturing any photons through your Ha filter. If you can't improve on the SNR (e.g. by exposing for linger), then perhaps you need to start from there.

The dim areas of my Ha images are close to the same level as darks, but actual image values are significantly higher. I'm definitely capturing emission nebula photons through the filter.

Failing that, you need to explore more than just these 'numbers' - if you perform a simple integration of your un-calibrated Ha Lights, and then very carefully apply an STF (automatically at first, then manually - by very small amounts - after that), what do you get? Do you see a meaningful image?

I have stretched some NB dark frames. They look pretty good -- mostly dark with some amp glow around the edges. Fortunately, dark-subtraction removes this same amp glow from light frames.

--- Mike

[GregW]

I see a lot of beautiful images on here from this camera. I tend to shoot at gain zero whenever possible. May I ask what you folks are using? I was from the school of thought to get maximum dynamic range and anti blooming properties from the sensor. However I see many people using gain 76 and even 200. I use the Astrodon5nm HAand 3nm o3 at f/5.5 or so. Right now I'm using 76 for oxygen 3 and 0 for the hydrogen Alpha. I'm having trouble conforming to anything else.

[GregW]

Just take bias it's easy you only have to do it once

[bulrichl]

GregW,

as Mike already noted, the usage of a master dark is important to get rid of the amp glow. Taking darks with the same exposure time as the light frames is fine. Then there is no need for scaling the darks, and subtracting the master dark from the light frames is all you need to do. You MUST NOT subtract both master dark and bias (or master bias), because the bias offset is already in the master dark. You don't need bias frames at all, so the bottom line is: DON'T take them.

If you are going to use flat frames, you should go on analogously for the calibration of the flat frames: take additional dark frames with the same exposure as the flat frames (i.e. flat-darks), make a master flat-dark, subtract the master flat-dark from each flat frame and integrate the calibrated flat frames to a master flat.

Regarding the appropriate gain and exposure time, Jon Rista gave good advice on the cloudynights forum "CCD/CMOS Astro Camera Imaging & Processing" for the ASI1600. It depends on the FWHM of the used filter, the aperture ratio of the scope and the extent of light pollution at your site.

Bernd

[GregW]

interesting. My data sets consists of Lights, Flats and dark Flats with Master Flats and darks at given gains and temp. Following the Warren Keller book, I have found use for the darks and bias and flats, but I haven't found what to do with my dark Flats(yet). I am moving over from Astro pixel processor.

[GregW]

perhaps the flat darks are used when creating a master flat. I have been using Master Flats I created elsewhere up to now yes. hmm