•

# ADU values pre and post calibration

5 replies to this topic

### #1 dawziecat

dawziecat

Mercury-Atlas

• topic starter
• Posts: 2592
• Joined: 20 Oct 2010
• Loc: Rural Nova Scotia

Posted 14 November 2013 - 01:36 PM

I am trying to use Steve Cannistra's exposure calculator.

The instructions state to use the background ADU values AFTER calibration.

I am surprised to see how different these values are.

For an Ha 15 minute exposure at f/2.8 my minimum ADU values are 1370 pre calibration and 300 after calibration.

For a B band sub, 10 minutes at f/2.8 I get 2200 before and 1650 after.

Does this sound right? Am I making some ghastly blunder?

I am measuring the ADU values with PI using a 15 pixel probe and 16 bit integer size.

Surveyor 1

• Posts: 1523
• Joined: 24 Feb 2007
• Loc: Snohomish, WA

Posted 14 November 2013 - 09:44 PM

Keep in mind what calibration does.

When you calibrate bias, you are basically going through each pixel in the light frame and directly subtracting the value of the same pixel from the bias master. If the background value of your bias frame is around 1000 ADUs, then you have your explanation right there. And it seems reasonable, too, since bias signal is basically added by the electronics in the camera and does not represent any signal coming from the sky.

Dark calibration works a little bit differently. Assuming again that you're using a bias master, the bias value is subtracted from the dark master. This gives the ADU value for dark current on each pixel. If your darks are the same duration as your lights, you can then just subtract the dark current values from the bias-calibrated lights. If your darks are a different duration than the lights, you can determine the ratio of dark duration to light duration and then scale the dark current ADUs, which can then be subtracted from the lights. Some adaptive algorithms use a different formula from just time scaling the dark current, but the result is the same that the dark current is subtracted from the light frames.

The upshot of this is that bias and dark calibration purposely subtract unwanted signal from the lights. It is therefore normal that your calibrated frames have a significantly lower ADU. And about 1000 ADU sounds reasonable to me (but can vary substantially, depending on the offset value used by your camera driver).

Flat field calibration is a different beast. It uses division instead of subtraction. To keep the explanation very short, your flat master is an image of an evenly lit field. In a perfect system, each pixel would then register the same number of ADUs. Because no system is perfect, each pixel will register ADUs that are some ratio of the number of ADUs in the brightest part of the flat master. Flat calibration is therefore division that applies that same ratio to the same pixels in the light frame...if that makes sense.

To sum up, your result sounds fine. I hope that this helps,

Surveyor 1

• Posts: 1523
• Joined: 24 Feb 2007
• Loc: Snohomish, WA

Posted 14 November 2013 - 09:47 PM

Oh, and one more thought.

An exposure calculator would want to work on a calibrated frame because it needs to calculate the amount of ADUs per given time that result from data coming from the sky. Since the bias data is fixed and does not increase with time, you would not want to consider it in the calculation. The same is true of dark current.

### #4 Mike Wiles

Mike Wiles

Viking 1

• Posts: 950
• Joined: 04 Feb 2009
• Loc: Goodyear, AZ

Posted 14 November 2013 - 10:08 PM

For an Ha 15 minute exposure at f/2.8 my minimum ADU values are 1370 pre calibration and 300 after calibration.

For a B band sub, 10 minutes at f/2.8 I get 2200 before and 1650 after.

It seems odd to me that have such a large difference in what gets subtracted between the two frames.

But your B frame only drops about 500 ADUs. Being that they're using the same Bias frame, I would expect that they'd both have very similar numbers in the total amount of ADU that gets subtracted.

Mike

### #5 dawziecat

dawziecat

Mercury-Atlas

• topic starter
• Posts: 2592
• Joined: 20 Oct 2010
• Loc: Rural Nova Scotia

Posted 15 November 2013 - 09:47 AM

Having only a vague appreciation of these processes, I have no explanation as to why the difference pre and post calibration between the Ha and the B band.

### #6 pfile

pfile

Gemini

• Posts: 3340
• Joined: 14 Jun 2009

Posted 15 November 2013 - 07:15 PM

did you do dark subtraction as well? that could account for the difference if the dark was scaled differently between the two images. 15 mins and 10 mins, after all.

anyway with respect to a narrowband vs. broadband filter, the background ADU values should be lower in the narrowband image since so many photons are rejected by the NB filter relative to the RGB filters. the bandwidth of RGB filters is like 60-100nm vs. 3-15nm for narrowband filters…

rob

# Imaging/Sketching Contest

 Cloudy Nights LLC Cloudy Nights Sponsor: Astronomics