This tutorial describes my current workflow for processing L-RGB images with Iris and Photoshop.

The examples in this tutorial are based on a set of test L-RGB images of the Ring Nebula which I took with my Atik 16HR.

Raw files are available for download and can be used for this tuturial.

(last updated: 3.2.2008)

Luminance images are greyscale images which represent only the brightness levels of an object.

You should capture the luminance data at the best possible conditions because it is the most important quality defining factor for the final image. Sometimes I use nights with bad seeing to capture only RGB data of different objects and use the truely good nights for capturing luminance only.

The picture shows a single, unprocessed, 5 minute exposure of M57 (The Ring Nebula) with just an IR/UV blocking filter.

Darkframes are recorded without exposing the chip to any source of light. (=telescope cap closed)

They are very important because you can use them to identify hotpixels, noise, amplifier glow and other unwanted things. They should be recorded at the exact the same temperature as the actuall exposures with the same exposure time.

I usually take at least 3-5 darkframes for each set of exposures, more if there is any significant change in temperature.

You have to record a set of darkframes everytime you modify the exposure time, the binning or if there is significant change in temperature meaning that you will also need a different set of darkframes for the RGB channels!

A flatfield is an exposure of an very even illuminated object. e.g. a white wall or the sky at daytime.

Falt fields are used to correct the vignetting of the optics or artefacts like dust on the sensor. The picture on the right shows you how a flatfield looks. You can see some dark spots in it which are caused by dust on the chip.

To create flatfields I simply use a sheet white paper which I put in front of the telescope while capturing the sky at daylight.

Now that Iris is able to access the images you can try to display one of them with the load command.

To do do that you have to open the command window by clicking the command button in the toolbar. (marked red in the picture) Once opened you can enter commands in it.

If you have downloaded my example raw files you can try to load a file with:

load m57_l_5min_1

Hint: Play a little around with the controls in the threshold window to change the way Iris is displaying the image.
Don't worry, this is just for visualization, the image data will not be modified by it.

Let us begin the real work and start to create the master darkframe for the luminance channel.

The command smedian will be used to create the master darkframe. This command has 2 parameters: filename base and number of files.

If you have 2 darkframes named m57_dark_5m_1.fit and m57_dark_5m_2.fit, the command would be:

smedian m57_dark_5m_ 2

After you have entered the smedian command the final master-dark-frame is computed but not yet written to disk. Use the save command to save it to disc.

save master_dark_lum

If you like you can press the auto button in the threshold window and take a closer look at your master dark.

Now that the masterdark is saved to the disc we can proceed in creating a list of all hotpixels with the find_hot command. First we have to do some math to evaluate the correct threshold for this command.

Enter stat in the command window to get the statistics of the darkframe. Iris will now open yet another window with the output of the stat command. The imporant values for us are Sigma and Mean.

Multiply sigma by 16 and add the value of mean to it. In our case (Sigma*16+Mean) would be 24.4*16+283 = 673.4 This is the correct threshold parameter for the find_hot function which we will now execute:

find_hot cosme_lum 673.4

With this command Iris calculated a list of all hotpixels in the darkframe and saved it as cosme_lum.lst to disc. We will use this file later to automaticly map out all hotpixels. If you are curious you can view this file with a text editor.

Since I consider offset files not very helpfull for CCD imaging, we just create a dummy offset file with the commands:

fill 0
save dummy-offset

This will create a dummy image filled with binary zeros at the same resolution as the previous loaded image.

In this tutorial we will also create a dummy flat field because the example images are captured with a Pentax 75 and there was no dust on the sensor. (The Pentax offers an almost perfect flat image, especially with small chips)

We just use the same command like before but we cant to fill the file with ones instead of zeros:

fill 1
save dummy-flat

In any other case you will probably want to create a real flatfield. Here is how it works: (comming soon)

Now that we have prepared a master dark file, a cosmetic file (to map out hotpixels), a dummy flatfield, a dummy offset file we are ready to stat the automatic preprocessing of the luminance images.

To do that go to Digital Photo in Iris menubar and select Preprocessing.

A new window will open where you can enter the filenames.

In this tutorial we will use:
m57_l_5min_ as input generic name,
dummy-offset as offset,
master_dark_lum as dark,
dummy-flat as flatfield and
cosme_lum as cosmetic file.

You can choose any output name you like. I use usually use p_lum_

The number represents the number of input (and output files). In our case we have 5 input images (m57_l_5min_1 to m57_l_5min_5)

As soon as you press the OK button Iris will start processing the input images and save the final images as p_lum_1 to p_lum_5.

Now that all the luminance images are hopefully free of any hotpixels, have reduced noise and a corrected field, it is time to register them before we can stack them.

You can use the command coregister2 to do that:

coregister2 p_lum_ r_lum_ 5

The first option is the basename of the input files, the second the output files and the last option is the number of files.

Sometimes coregister2 fails to register all images correctly. This is especially the case when you have allot of stars in the image (like milkyway widefields).
If this happens you can try the command coregister4:

coregister4 p_lum_ r_lum_ 512 5

The 3rd parameter is the size of the fields iris will use to detect stars. 512 or 256 are usually good values but it depends on the size of your image.

After we have correctly registered the images we are able to stack them. Iris provides serveral ways of stacking. I usually use the composit stack because it is possible to filter out things like cosmic rays, satellites, airplanes, etc.

Let us try to do a simple composit stack of the images with the composit command:

composit r_lum_ 3 1 1 5

The options for the composit commands are:
composit basename Kappa Iterations Normalize Number

If you use my example raw files, you will probably have noticed that we have some trails in the image caused probably by a satellite. If this happens, you can decrease Kappa to 2 or even 1. In our case only Kappa = 1 works corretly:

composit r_lum_ 2 1 1 5

Sometimes you want to obtain the highest possible signal to noise ratio.
In this case you can use the add2 command to just add the images together:

add2 r_lum_ 5

After you have stacked the images you should save them before you continue to process them:

save lum_stack

Before we export the luminance channel to photoshop we need to verify and correct the offset level of the image. This will help us to set the blackpoint more easily in photoshop.

Use the stat command to check the current offset.

In our example the Mini. value is -324.0.

To correct the offset to 0 we use the commmand:

offset 324

Now we can export the luminance image to photoshop:

savepsd2 lum_stack