M 57 - Luminance Vs Infrared

Contents

  1. Infrared Imaging

  2. Image Acquisition

  3. Subtracting the Dark Noise

  4. Comparison of Luminance and Infrared Light

An earlier post was of a quick M 57 image in Luminance filter (400 nm-700 nm). In my archives i found a three years old data of this nebula which i did not process at the time.. why do i do it with a lot of my data.. is a mystery to me too :)

Infrared Imaging

It was imaged with an Infrared filter.. yes infrared!! Could’t believe i would even attempt it in the first place!

The filter is Astronomik IR Pro 807 which has a bandpass way above 700 nm. So this will show the invisible light only to the CCD chip. Problem with that is the CCD Quantum Efficiency declines in that band. I use SBIG ST9XE CCD and you can see the QE is quite low in that IR region. At 800 nm, only 45 percent remains.

Celestron C14’s big aperture was a hopeful thing i have.. since it will be able to collect more photons which are coming in a rare quantity from M 57 Ring Nebula.

IR Pro transmission.png

Figure 1. Astronomik IR Pro Specturm

ST9 QE.png

Figure 2. SBIG ST9XE QE Graph

Image Acquisition

Now to image in this very faint light, i decided to expose the camera for 10 minutes. For autoguiding the mount, i found a star bright enough to give my mount one correction every 3 seconds. Pressed the Take Image button and waited for 10 minutes. Will any sign of Ring Nebula show up at my screen?

 
RAW Single IR Pro .png
 

Figure 3. 10 Minutes Raw Image with IR filter

This was the image that came up at my screen.. some random stars and lots of dark noise. Do you see anything else there? any hint of Ring Nebula?

Just in the center, i thought something related to a ring nebula is there.. or is it?

Subtracting the Dark Noise

Since there is so much noise there in Raw image.. I should subtract the dark data from this image and that probably will confirm me if there is the Ring Nebula there or not.

There you go folks.. Ring Nebula in Infrared light!!! It is very dim but can be identified in Figure 4.

Calibrated-Single--M57-IR-Pro.png

Figure 4. Calibrated image of Ring Nebula in IR

Comparison of Luminance and Infrared Light

So i thought to compare the IR image (Figure 5 a) with Luminance image (Figure 5 b). Luminance is what your eyes can see.. 400 nm to 700 nm. Infrared image should show more stars.. Following are the inverted images.. sometimes it is useful to see Astronomical images in inverted frames.

How do you compare these images?

M57-IR-Pro-Cropped Final.png

Figure 5 a. M57 in IR band (Above 807 nm )

M57-L-Cropped Final.png

Figure 5 b. M57 (400 nm - 700 nm)

M 57 - Playing with the Images

Contents

  1. Images with Graphs

  2. Sky Background

  3. CCD Camera & Telescope Unwanted Noise/Signal

  4. Read Noise

  5. Dark Frame

  6. Flat Frame

  7. Calibrated Result

Images and Graphs

The previous post was about M 57, the famous Ring Nebula.. Let’s play some more here with the images we have.

For our eyes, it is very difficult to compare or measure the light sources such as stars by just looking at an image. All we see is the black background and a bunch of white dots. some are bright some are dim but no real comparison of how much is the difference. Maxim DL has a wonderful feature where it graphs the light quantity or flux of the image, creating a beautiful visualization in 3D.

Figure 1, is the single 300 seconds raw image directly out of the camera. Raw means it has the light of the stars and the Ring Nebula which we want to image but it also has all kinds of unwanted light which we collectively call noise in the image. More about the noise later.

Figure 2, is the intensity graph of the selected area around the Ring Nebula. The darker colors represent low photon (light particles) counts and the brighter ones are the areas where more light has been detected in the image.

Raw 5 min Sub M57.png

Figure 1. 300 seconds RAW Image (faulty auto guiding/tracking making the stars oval instead of round)

Figure 2. Light intensity 3D Graph (Maxim DL)

The deep blue color at the bottom (Figure 2) is the representation of the lowest light detection in the image, which you might think should be zero because the image is from space and well.. space is all black! But the reality is quite different when it comes to imaging the night sky.

Sky Background

The observatory is in Lahore, a heavily light and dust polluted sky always dominates here. On the day of imaging (May 3, 2020), there was moon in the sky, at 80% illumination which was sending all the unneeded photons in the sky and making it anything but dark. So this all moved the background '‘black’ sky floor to higher levels. Astrophotographers try their best to keep this background as low as possible and there are various gadgets and techniques to do that but this can never be zero or rather i should say, should not be zero because then there is no statistical data in that pixel and it will not behave properly when calibrating (cleaning the image) the image begins. A non zero ADU value is very important to have in your pixel!

If you think this is bad enough, nature has other ways to mess up poor astronomers’ lives!

CCD Camera & Telescope Unwanted Noise/Signal

CCD Camera has two noise sources:

  1. Read/BIAS Noise

    When the chip is read, it adds noise in the measurement which is called the read out noise. So what we do is to record the readout noise and then subtract it from the original raw image.

  2. Dark Noise

    At any temperature, photons are produced in the chip. By cooling the CCD chip, we can reduce the number of photons which gets us closer to the wanted signal of the celestial object.

Read Noise

Figure 3 is the Readout Noise image from the camera and Figure 4 shows the 3D graph of this noise. You can see the noise floor is low, 900 some ADU (Analogue to Digital Unit) value. This will be subtracted from the raw image. Bias pixels show less than 1000 ADU values and this value depends on the camera manufacturer.

Some spikes can be seen in Figure 4. This is because not all pixels are created equal. All pixels give a bit different value but some show much higher values than majority of them. These higher values are corrected in the calibration process.

Figure 3. BIAS frame (a zero second exposure is read from the camera)

Figure 4. 3D graph of the BIAS noise

Figure 5 shows the cursor at a pixel which is located at X=391, Y=358 position on the CCD chip and is showing an ADU value of 2264, which can be read at the bottom in the picture. The average pixel value in the Bias frame is 964.

IMG_8235.JPG

Figure 5. Cursor is at bright Pixel number (391, 358) in Bias frame

Figure 6. Pixel number (391, 358) in Bias frame

Same pixel (391, 358) in raw Image shows maximum value of 65,535 ADU, the highest a 16 bit can record. Hot pixels do that.. they would hit the higher or highest values. We always try to avoid the ADU values close to the maximum number in our images because CCDs are not linear in that range and no reliable measurements can be done around these values.

But hot pixels can be fixed in a number of ways. One of the method to keep these pixels in lower ADUs is to keep the CCD temperature as low as possible.

Figure 7. Pixel number (391, 358) in Raw Image

Here is an actual example of calibrating the image. In Figure 6, the pixel has a value of 2264 and in Figure 7, the same pixel in the Raw Image has the value of 65535. Now if we subtract it, the resultant pixel (Figure 8) shows a reduced value of 63373, a difference of 2162.

Hmm.. shouldn’t it be 2264? because that is the value being subtracted. This happens not with the hot pixel but with other pixels too.. every pixel is about a 100 pixel less than the expected value. There must be an explanation.. I have no idea what that is.

IMG_8236.JPG

Figure 8. BIAS subtracted Raw image

Dark Frame

Figure 9, is the 300 seconds exposure with the camera shutter closed and the CCD at -10 Celsius temperature. Figure 10 is the graph of this dark frame.

Figure 9. Dark frame 300 seconds at -10 Celsius

Figure 10. Dark Frame 300 seconds (-10 C)

Flat Frame

Light coming from the any optical system such as a telescope is not perfectly evenly distributed and there will be dust particles in the way no matter how much we want to clean our equipment. To solve this problem, astrophotographers take Flat Frames by imaging an even white surface (there are many ways to take the flat frames) and then dividing it from the raw frame.

Figure 11. Flat frame with Luminance filter

Figure 12. Flat frame (L filter)

Calibrated Result

That’s the difference between a Raw and a Calibrated image (Figure 13 & 14).. A 3D Ring Nebula.

Figure 13. Graph of 300 seconds exposure raw frame

Figure 14. Graph of calibrated 300 seconds image