Image Calibration Tool

Scooot

Well-known member
Hi all,

In the IC Tool, could someone please explain the difference between:

Putting a cross in Master Bias and selecting a master bias file but not entering a cross in Calibration
or
Putting a cross in Master Bias and selecting a master bias file and entering a cross in Calibration

I understand that calibrating images with the bias is to subtract them but I’m trying to understand what happens if I don’t select calibration, if that makes sense.
 
bias calibration is pretty esoteric... it requires there to be a defined overscan area, which means the sensor actually has to have one and returning that data to the camera driver even though it's not exposed to light. this is sometimes called an "optical black" area.

i think this is something that probably 99% of users don't need to mess with.

btw you can hover over the little tick boxes to get tooltips that describe a little more about what the particular control does.

rob
 
Thanks for the reply Rob,

I’m definitely in the 99% bracket, I wasn’t after anything in depth.

For example, looking at the screenshot from Pixinsight Resources attached. Calibrate isn’t ticked in the Master Bias, so is it still subtracted from the lights?
 

Attachments

  • 22890EBB-BFB9-4ACA-8999-EE4E8D8DE911.jpeg
    22890EBB-BFB9-4ACA-8999-EE4E8D8DE911.jpeg
    372.7 KB · Views: 82
yes - the check boxes along the left side of each section determine whether or not the frame is subtracted from (bias/dark) / divided into (flats) the lights. the check boxes underneath each filename control whether the master frame is calibrated (and/or optimized, for the darks.)

rob
 
Thanks again, I understand that now. I’d always thought the subtraction or division was the actual calibration. In general terms, if possible to explain, what happens in addition to calibrate the images, some sort of standardisation process I guess.
 
calibration of the lights is the subtraction of master bias and/or master dark from the lights and division by the master flat.

calibration of the master frames is a separate thing. you'd need to use a calibrated master dark in order to use dark optimization, for instance. and flats always need to be calibrated or they will "over-flatten" the lights.

the calibration tickboxes for the masters in the ImageCalibration process allow you to also do calibration of the master frames at light frame calibration runtime. this calibration is "temporary" in that the calibrated masters are used in the run but not written out to disk. so you might make an uncalibrated dark master and an uncalibrated flat master and then calibrate them at light calibration runtime. for darks, that's the right thing to do. for flats, people usually calibrate the flat subs separately to make a calibrated master flat.

bernd's sticky thread on calibration is really good; it explains the whole process


rob
 
Brilliant, thanks very much, I’ll have a read of the sticky thread.

So in summary, by ticking both boxes next to the master dark for example, I’d be instructing the tool to first calibrate the master dark and then calibrate the lights by subtracting a just calibrated master dark.
 
Last edited:
well, by ticking the box to the left of "master dark" you're telling the tool to subtract the master dark from whatever is loaded in the file list, which would be your lights when calibrating the lights.

the tickboxes under master dark mean to optionally calibrate the master dark and then optionally scale the master dark before doing the above subtraction.

rob
 
well, by ticking the box to the left of "master dark" you're telling the tool to subtract the master dark from whatever is loaded in the file list, which would be your lights when calibrating the lights.

the tickboxes under master dark mean to optionally calibrate the master dark and then optionally scale the master dark before doing the above subtraction.

rob
“the tickboxes under master dark mean to optionally calibrate the master dark and then optionally scale the master dark before doing the above subtraction.”

As the master dark also contains the bias signal, would I be correct in saying the above subtracts the bias from the master dark before it is subtracted from files on the file list. Otherwise the bias would be removed twice.
 
that is correct - calibrating a master dark means to subtract a master bias. if you loaded (and ticked) an uncalibrated master dark and a master bias and did not tick "calibrate" under the master dark, there would be a double bias subtraction.

as pointed out in bernd's thread it's best to make uncalibrated master darks since subtracting master bias frames from dark subexposures can lead to clipping in the calibrated dark subexposure. it's better to integrate uncalibrated dark subs and then calibrate the dark at light integration time. the likelihood of clipping the master dark with a master bias is pretty low.

rob
 
that is correct - calibrating a master dark means to subtract a master bias. if you loaded (and ticked) an uncalibrated master dark and a master bias and did not tick "calibrate" under the master dark, there would be a double bias subtraction.

as pointed out in bernd's thread it's best to make uncalibrated master darks since subtracting master bias frames from dark subexposures can lead to clipping in the calibrated dark subexposure. it's better to integrate uncalibrated dark subs and then calibrate the dark at light integration time. the likelihood of clipping the master dark with a master bias is pretty low.

rob
Thanks Rob,

I’ve just taken delivery of a new Asi 2600 MC Pro, I haven’t used it yet, so I’ve been trying to get to grips with this to decide the best way of calibration etc.

As the camera has zero amp glow and is cooled, I’ve worked this out in the light of our discussions.

Take Flats, Darks and Flat Darks all at the same temperature.

1. Integrate the Flat Darks.

2. Calibrate the flats by selecting the Master Flat Dark in IC when the flats are loaded. No need to tick calibrate or optimise etc as they’ll match exposure and temperature exactly and there’s no amp glow. The bias will therefore be subtracted from the flat at the same time.

2. Integrate the Flats to create a Master Flat.

3. Integrate the Darks to create a Master Dark.

4. Calibrate the Lights by selecting the Master Dark and the Master Flat, no need to tick calibrate and or optimise either. Again the bias will be subtracted at the same time.

I probably could get away with just using a Master Dark but anyway.

Does this sound like a reasonable approach?
 
this sounds correct to me. i don't use a CMOS camera so i don't have practical experience with them. my understanding is that some of these cameras (not necessarily yours) have issues with bias frames so your "all-dark" approach seems safe.

the main thing to look out for is to make sure the offset and gain of the darks matches the offset and gain of the lights, and likewise the offset and gain of the flat darks and the flats match. i think because you are calibrating the flat subs it would not matter if the gain and offset of the flats/flat darks don't match the gain and offset of the lights/darks. but it won't hurt to keep them the same either. it's just that with bright light sources for the flats, sometimes people need to use a lower gain for the flats to avoid super short flat exposures.

one thing that seems to crop up a lot is that people use one program to capture lights and then a different program to capture the darks. this can present a huge problem because even though the gain/offset might be programmed the same in the interface of the two different programs, if there is a difference between the drivers (ascom vs. native, driver version, etc.) the different drivers program the hardware a little differently and thus the darks are not actually compatible with the lights or flats. the safest thing is to use the same program and same driver for all frames.

rob
 
Hi Richard,

your general workflow is correct.

I probably could get away with just using a Master Dark but anyway.
No, you definitely need either the MasterFlat-Dark that you already have or a MasterBias for the calibration of the flat frames, otherwise (if you don't calibrate the flat frames) the flat field correction will result in undercorrection.

Bernd
 
this sounds correct to me. i don't use a CMOS camera so i don't have practical experience with them. my understanding is that some of these cameras (not necessarily yours) have issues with bias frames so your "all-dark" approach seems safe.

the main thing to look out for is to make sure the offset and gain of the darks matches the offset and gain of the lights, and likewise the offset and gain of the flat darks and the flats match. i think because you are calibrating the flat subs it would not matter if the gain and offset of the flats/flat darks don't match the gain and offset of the lights/darks. but it won't hurt to keep them the same either. it's just that with bright light sources for the flats, sometimes people need to use a lower gain for the flats to avoid super short flat exposures.

one thing that seems to crop up a lot is that people use one program to capture lights and then a different program to capture the darks. this can present a huge problem because even though the gain/offset might be programmed the same in the interface of the two different programs, if there is a difference between the drivers (ascom vs. native, driver version, etc.) the different drivers program the hardware a little differently and thus the darks are not actually compatible with the lights or flats. the safest thing is to use the same program and same driver for all frames.

rob
Ok sorted -:)

Thanks for all your help & time, much appreciate.
 
Hi Richard,

your general workflow is correct.


No, you definitely need either the MasterFlat-Dark that you already have or a MasterBias for the calibration of the flat frames, otherwise (if you don't calibrate the flat frames) the flat field correction will result in undercorrection.

Bernd
Thanks Bernd, I’ll stick to the first plan. -:)
 
Thanks Rob,

I’ve just taken delivery of a new Asi 2600 MC Pro, I haven’t used it yet, so I’ve been trying to get to grips with this to decide the best way of calibration etc.

As the camera has zero amp glow and is cooled, I’ve worked this out in the light of our discussions.

Take Flats, Darks and Flat Darks all at the same temperature.

1. Integrate the Flat Darks.

2. Calibrate the flats by selecting the Master Flat Dark in IC when the flats are loaded. No need to tick calibrate or optimise etc as they’ll match exposure and temperature exactly and there’s no amp glow. The bias will therefore be subtracted from the flat at the same time.

2. Integrate the Flats to create a Master Flat.

3. Integrate the Darks to create a Master Dark.

4. Calibrate the Lights by selecting the Master Dark and the Master Flat, no need to tick calibrate and or optimise either. Again the bias will be subtracted at the same time.

I probably could get away with just using a Master Dark but anyway.

Does this sound like a reasonable approach?


Richard,

I have a QHY268c that uses the same Sony sensor as the ASI2600MC. This sensor has an overscan area and QHY recommend to make use of it. I would therefore assume that this holds also true for your ASI2600MC. Take a look at the following thread: ImageCalibration and Overscan area procedure

/Ralf
 
Richard,

I have a QHY268c that uses the same Sony sensor as the ASI2600MC. This sensor has an overscan area and QHY recommend to make use of it. I would therefore assume that this holds also true for your ASI2600MC. Take a look at the following thread: ImageCalibration and Overscan area procedure

/Ralf
Thanks Ralf,

Thats interesting, I’ll revisit that if I have to.

As you mentioned the QHY I came across some info the other day in a thread on SGL you might be interested in. As you say both use the same sensor but the QHY has a 60,000 FW whereas the Asi has a 50,000 FW. The reason, apparently, is that zero gain on the Asi is equivalent to a gain of 19 on the QHY. In other words the lowest gain you can use on the Asi is the equivalent of 19 on the QHY and ZWO call it zero.
 
As far as I know, ZWO's camera driver doesn't have a setting to show an overscan area. Therefore overscan calibration is not an option with the ASI2600MC.

In principle overscan calibration cannot correct issues with banding either. Overscan calibration is necessary, when the bias offset is not stable in time. However, this doesn't seem to be the case with this camera.

Bernd
 
Back
Top