Using dark and flat frames . . .

[Terry Danks]

Is there a source of info for getting PI to use darks, flats and bias frames?

Terry Danks

[Ioannis Ioannou]

How about the tutorials ? http://www.pixinsight.com/tutorials/master-frames/en.html
Also you will find a lot of usefull information by just searching this forum

[DB]

I've been using PI for while but another program for dark subtraction.  I also can't easily see how the subtraction is performed and find that the tutorial is not helpful as it appears to be more about creating a master dark file than subtracting it from the light images.  I gather that Image Calibration is used but can't seem to get that to work.

Any advice appreciated!

Thanks

DB

[Ioannis Ioannou]

Take a look in Harry's video tutorials, they are a must (eg http://www.harrysastroshed.com/Stack.html , but all are useful).

Also, strongly recommended to search the forum (eg http://www.pixinsight.com/forum/index.php?topic=2564.0 )

My workflow is easy (YMMV) :

master bias
master dark
master flat
(above without any calibration)

calibrate lights with the above (tick calibrate for each, also I use it due to overscan on my camera)

Star Alignment
Image Integration

[Terry Danks]

Not really getting anywhere with this.

My files are all Canon .CR2 RAW files . . .

"Generate master bias. Simply integrate them in Image Integration."

Open Integration tool and select Add Files gives you a window that will ONLY allow .fits files?

I gott'a wonder . . . how many folks are really using this software to apply bias, flat and dark frames?

[Cleon_Wells]

Terry and DB, after you’ve read Vincent’s  directions take a look at the screen shoot of my Calibration example, note the Process Console info. Note how well the Light is scaled with the MasterDark, very close to a 1.0 scaling factor, remember that temperature makes a big difference in PI’s MasterDark scaling process.
You can experiment   with using just one Master -  Bias , Dark and Flat  .fit file, this will allow you to quickly check you Calibration process. Once you get this to work you can Cal and integrate you master Calibration files, check that they work better then the single Calibration files.
Note that in this example I Calibrated a Flat raw sub and three Dark raw subs with different Dark temperatures --- C#. The different Dark temps are caused by my delay in taking the darks, allowing the camera chip to cool down, my bad.
Cleon
PS Terry there are a lot folks that have difficulty with PI's Cal process.

[georg.viehoever]

....
PS Terry there are a lot folks that have difficulty with PI's Cal process.

I am still using DSS because I just got nistete with calibration in PI. While i have no doubt that PI can produce better results, it failed too offen. Still hoping for the Genius that Makes the Procedere easier.
Georg

[astropixel]

http://pixinsight.com/forum/index.php?topic=2570.msg19019#msg19019

Vicent Peris wrote this as a process icon tutorial. I copied it into the DSLR workflow thread. It is generic however, and should answer your questions. The problem with the whole calibration thing in PI, is that there are various ways of obtaining the same result and each has their own preferences. Often, explanations start from an assumed level of knowledge, which is not very helpful when your starting out in PI.

The link above is a good starting point - it covers the basics and is coherent. There are in my view too many fractured explanations on this subject around the forums, when in-fact it is quite an easy process. Difficulties usually arise from poor acquisition techniques of calibration and light frames.

[andyschlei]

I've found PI's calibration routines fairly easy to use, and the tutorials on Darks, Biases, and Flats very helpful.  Perhaps this will help.

I assume you are doing LRGB imaging and your luminance is binned 1x1 and the RGB is binned 2x2.

Your raw imaging frames have artifacts from the imaging train (vignetting, dust, etc.), baseline noise from the CCD, and dark noise from heat and imaging time.  You need to remove these from your images.

Flats correct artifacts from the imaging train, and darks and biases, the other two noise factors.

I assume you've taken flats, darks, and bias frames in addition to your lights.

Combine -- Image Integration -- the darks and bias per the instructions in the tutorial.  These are the master darks and biases.  Then calibrate -- Image Calibration -- the flats images.  Since a good flat is at about 50% average full ADU depth on your CCD, if you aren't taking sky flats, 4 or so flats should be adequate.  If you aren't taking sky flats, calibrate the flats using your master dark and bias and do a simple mean.  Now you have master darks, bias, and flats.

Use image calibration to calibrate your frames, using appropriate flats (one for each filter and rotation), darks and bias frames.  I count on PI to adjust darks for temperature and time.

Then align your luminance frames with StarAlignment.  Use Image integration to combine the aligned luminance frames.  Use appropriate pixel rejection to eliminate outliers.  The help on this is very good.  Save your integration, then use IntegerResample to downscale the luminance to a 2x2 size to serve as the master alignment frame for the RGB.

Align the RGB to the binned luminance image, and I would recommend setting the interpolation on the RGB to nearest neighbor, assuming you have at least 6 color frames.  You should also be dithering your captured image to move any CCD artifacts around the actual image.  Combine the frames using the best pixel rejection scheme.

At that point, you have the ingredients to follow the excellent video tutorial on LRGB combination.

I hope this helps.

Clear skies,

--Andy

[Terry Danks]

Thanks for those who attempted to help but any calibration that starts with my needing to convert all darks, flats and bias frames from .CR2 to fits before even beginning to proceed is a just a total non-starter. I pretty much stopped listening as soon as I opened the Integration tool and found a drop-down window with but a single selection . . . fits or nothing.

I initiated the thread because I was having some problems with DSS on a particular stack.
It was easier to figure out what was wrong with DSS than it was to begin the multi-step and apparently HDD-voracious procedure that PI seems to require.

Seems to me the Integration tool should work with camera raw format just as the Star Alignment tool does, without the need to first convert all files to fits. As long as DSS continues to do what I need done, I just see no benefit in taking on the added complexity of PI calibration.

IMHO, PI really needs work here.

[pfile]

i understand how you feel.

there is a script called "preprocess pipeline" that automates this stuff. it hasnt been maintained in a while but it does work.

http://pixinsight.com/forum/index.php?topic=2861.0

fits is a requirement simply because it supports incremental reading. yes, you could bake CR2->fits directly into the ImageIntegration tool and write the files out temporarily somewhere. but i think PI is trying to be more platform-agnostic as it were.

anyway as others have said the end product of PI calibration and stacking is superior to DSS, so that's why i go through the trouble of doing it with PI. i do have a bunch of process icons so all i need to do is double click them and put in the new files and go.

[astropixel]

Thanks for those who attempted to help but any calibration that starts with my needing to convert all darks, flats and bias frames from .CR2 to fits before even beginning to proceed is a just a total non-starter. I pretty much stopped listening as soon as I opened the Integration tool and found a drop-down window with but a single selection . . . fits or nothing.

Hi Terry. True, bias frames require conversion direct to .fit, and otherwise require no calibration, except to integrate into a master bias.  But it is not necessary to convert to fit first for darks and flats. Once the master bias is generated it is used to calibrate RAW darks and flats, which  are converted to .fit in the process.

Either way, its a two step process for bias, darks and flats. Convert bias to .fit and integrate. Calibrate darks and flats with the master bias and integrate.

I will edit the offending paragraph in the process icon description attached to the DSLR_RAW workflow. Now I am sticking my neck out. In my view it's OK to calibrate RAW files with the .fit bias, because, RAW files are converted to fit during calibration with the ImageCalibration tool. Any other process is just an alternative.

Please correct me if I'm wrong.

Help Juan, I'm about to be eaten alive, but I would like to settle this once and for all.

Cheers,

ap

[Terry Danks]

Many thanks . . . I'll keep at it for a bit.

I just spent a few minutes converting .CR2 Bias frames to fits manually.
Gott'a be a better way of doing it than that.
Is there a tool for converting  bunch of RAW files to fits in one go?

[astropixel]

Terry. Script > Batch Processing > BatchFormatConversion.

BTW. There is no need to convert to .fit, except the bias frames. Just plug your RAW files into ImageCalibration and subtract your newly integrated bias.

[pfile]

be careful though - unless you use the script i posted in the PJSR forum, you will have to make sure to set the DSLR_RAW file handler to bayer, no black point correction, etc. before starting the BatchFormatConversion script.

actually the script i posted has a bad memory leak, so don't use it