#130 - Thursday, April 19, 2018 – Eye of the Needle (and Utility of the Refrigerator)

The forecasts were looking very promising for Thursday night, so I loaded up my camera gear and went out to the observatory to image!  It’s been about a month since I was last out there.  It’s “Galaxy Season,” the time of year when the winter Milky Way and all the nebulae around Orion are setting early, but when the summertime Milky Way isn’t rising until late at night.  This means we are seeing above the plane of the galaxy in the first half of the night, a great time to see the plethora of galaxies that are visible in amateur scopes, including dozens in Coma Berenices.  The Needle Galaxy is a rather bright and rather large member of that group, making it a worthy target for the astronomy club’s 5-inch refractor.  I’ve only imaged it once before – at last year’s Texas Star Party, through my 11-inch Schmidt-Cassegrain, after seeing it in a fellow club member’s Dobsonian scope and being surprised at how big and bright it was for not being a Messier object.  How did Charles Messier miss this gem??

Finally, a good selfie with this scope.

Now that spring is finally shoving winter back into its deep dark hole, night is coming later now – sunset wasn’t until about 8:15 PM.  Since it was a weeknight, I could only be out until about midnight or so, but I decided to image with the ZWO ASI1600MM Pro anyway.  I decided to go ahead and try to get a full set of LRGB data, since I’m not going to be able to get out to image again mostly likely until after the Texas Star Party.  So I planned on taking an hour’s worth of luminance data, and an hour’s worth of color data, which is 20 minutes on each color channel.  I stuck with my standby exposure time of 5 minutes, although next time I think I will try doing many shorter exposures, which I’ve read online is really much more in this camera’s wheelhouse. 

I also wanted to try using AstroTortilla, which is an application that does plate solving and will slew your scope to center the object you want for you.  I had set it up earlier that day, and during installation, you have to set the range of your fields-of-view, largest and smallest, to cover all of your gear setups.  My smallest FOV is 21 arcmin x 16 arcmin, which is my ZWO attached to my 11-inch SCT without a focal reducer (what I might do for planetary nebulae and small galaxies, and my largest is my Nikon D5300 attached to my Borg, which is about 2.6x1.8 degrees (the largest FOV on a mount that AstroTortilla could control, anyway – so this excludes my Vixen Polarie, of course).  During setup, you select the index files that not only cover this range, but on the narrow end includes 20% of your smallest FOV, which for me is the 2.8-4 arcmin index file.  The smaller the FOV, the larger the index file for plate solving is (the index file is the reference images of the night sky to compare your camera’s images to), so I had to download several gigabytes of data.  But it finished just before I left the house!

After getting all attached and powered up, a process that went smoothly, it was still light out, so I slewed to the moon.  I had to do this manually – the catalogs in Gemini’s computer-control app contain a lot of catalogs of both bright and dim objects, but it does not contain a catalog of solar system objects!  I opened up SharpCap and got all the settings set how I wanted them, and I had a pretty decent frame rate going of about 15 fps, but once I hit the Capture button, it fell to like 0.5 fps and it was dropping more frames than it was saving.  I eventually found the culprit: my tablet’s hard drive was nearly full!  I had forgotten to move some darks I had been taking recently over to my desktop, and the 16-bit nearly-full-frame images I was saving off were quickly filling up the leftover space.  I went through and deleted as much extraneous stuff as I could (forgetting all about the 21 GB of movies I had on it for airline trips).  I deleted previous Windows installations and some other things, and managed to create just enough space to record a few captures on the moon, and later Jupiter, but using only 8-bit images instead of 16 bit.  I got much better frame rates after that – nearly 50 fps when I used a smaller region-of-interest for Jupiter at the end of the night.  I was worried I wouldn’t have enough space for the rest of the imaging I was planning to do that night, but I was taking long exposures of the Needle Galaxy, so I didn’t fill up as much memory.  Needless to say, I went onto Amazon and bought a 64GB micro SD card that very night to put into my tablet!

Anyway, I got some nice images of the moon!

Date: 19 April 2018

Object: Moon

Camera: ZWO ASI 1600MM Pro

Telescope: Vixen 140mm neo-achromatic refractor

Accessories: Astronomik L filter, Type 2c, 1.25"

Mount: Losmandy Gemini II

Short exposure (top): 156 frames

Long exposure (bottom): 436 frames

Exposure: 10ms (top)

   300 ms (bottom)

ISO/Gain: 0 (top)

 300 (bottom)

Stacking program: RegiStax 6

Once it was dark, I slewed to Arcturus for focusing, and the diffraction spikes from my Bahtinov focusing mask already looked pretty close since I had just focused on the moon, so I slewed over to the Needle Galaxy.  AstroTortilla did appear to start working, although I had to close out of SharpCap so that it could talk to the camera, but it wasn’t able to plate-solve the image.  With how little time I had that evening, I decided I’d try that again on a different night, and since the Needle Galaxy is bright, I was able to see it readily in Sequence Generator Pro with 3-second frames when I turned on auto-stretching the histogram.  I got it centered, and started my luminance frames.

Those went well, although they looked out-of-focus in the raw frames.  Raw FITS frames, which is what comes off of astro cameras, look grainy and terrible, and sometimes also out of focus.  But I had just focused and didn’t move the scope very far, so I felt pretty certain they were actually in focus.  Guess I’d have to find out!  

Single 5-minute luminance frame on the Needle Galaxy.  It's not as blown out as it looks - FITS files have adjustable luminance curves.

After an hour of 5-minute luminance frames, I rotated my filter wheel to red for 20 minutes, then green, then blue.  All went smoothly, and guiding looked great – I was getting only about 1.5 arcsecs of error, which is pretty decent. 

After midnight, when I was done capturing the Needle Galaxy, Jupiter was up, and I couldn’t resist!  However, it is very tiny inside that 140mm aperture, 800mm focal length refractor, so I was pretty sure it was going to come out poorly.  I took the data anyway, and upon further review, yeah it was pretty crappy.  I’ll just have to set up my 8-inch in my front yard soon!  Luckily, you can image planets out to Saturn easily under even a full moon because of how short of exposures you need, so I can do that next week if we have some clear nights that aren’t freezing.

I didn’t get a chance to process the Needle Galaxy image for a few days, but I stayed up late the following Saturday night and rush-processed them before heading off to a work conference so I could write the blog post :D  I didn’t think it was going to turn out that well anyway, since I only got four frames per color channel, which gives poor signal-to-noise ratio for the color channels.  I also didn’t have a set of bias frames for the temperature I ran the camera at (-25C) yet, but I decided to stack without them anyway.  I ran into a snag early on though – a weird issue with the new version of DeepSkyStacker that just came out, 4.1.0.  I was really excited about it because they finally had a 64-bit version, which means it should be able to use a lot more RAM.  I stacked the luminance, red, green, and blue frames, saved them out as 32-bit TIFFs, and imported them into Photoshop.  I zoomed in to see how well the detail on the luminance channel came out and saw that the image was full of holes!  Missing pixels everywhere.  So I went back to DSS and turned off hot pixel removal, just in case (I hadn’t had troubles with that before), but got the same result. 

                

I thought next that maybe it had to do with the lack of bias frames, and since I can control the temperature of my camera and am therefore not at the mercy of the exact outdoor temperature, I decided to throw the camera out on the back porch like I always do to take darks and just grab a set of bias frames real quick.  Bias frames are the same gain and temperature, but the shortest available exposure time, as your light frames, so they’re quick to acquire.  However, it was kind of warm outside, and my dryer was running, blowing hot air right onto my dark capture spot!  So the only alternative was to put my camera in the fridge!

The door still closed easily over the cables, so I let it cool, grabbed the frames, and was back in business!

So I re-stacked just the luminance frames with the biases, aaaaaand still had holes in my picture.  So there were two more options: either my newly-acquired flat frames from a few weeks ago were bad, or something was wrong with the new version of DeepSkyStacker.  To check, I launched the older version of DSS and processed the images – and they came out fine, no holes!  Now I need to get onto the DSS forum and report the problem.  And see if anyone else is having this problem, probably over on Cloudy Nights.

I stacked all the channels in the old version of DSS, stretched them in Photoshop, and saw that my flats didn’t flatten very well, so I had to make synthetic flats.  (I’ll write a blog post on that soon, I promise!)  I combined the LRGB all together, adjusted levels and curves to get the color balance and contrast I wanted balanced against the high noise that’s a result of the low number of frames, ran the blue halo removal routine in Noel Carboni’s astronomy tools for Photoshop toolkit (which works well on my DSLR images, but less well on my astro camera images, it seems), and here it is!

Date: 19 April 2018

Object: NGC 4565 Needle Galaxy

Camera: ZWO ASI 1600MM Pro

Telescope: Vixen 140mm neo-achromatic refractor

Accessories: Astronomik LRGB Type 2c filters

Mount: Losamndy Gemini II

Guide scope: Celestron 102mm

Guide camera: QHY5

Subframes: L: 12x300s (1h), 1x1

   R: 4x300s (20m), 1x1

   G: 4x300s (20m), 1x1

   B: 4x300s (20m), 1x1

   2h total

Gain/ISO: 139 (unity)

Stacking program: DeepSkyStacker

Stacking method (lights): Auto-Adaptive Weighted Average

Darks: 20

Biases: 30

Flats: 30

Temperature: -25C (chip), 45F (ambient)

Kind of noisy, got some weird dark halos around the stars as well that let me not be able to blacken the background as much as I would have liked, but not bad coming from my light-polluted area on an achromatic refractor!  I’ll most likely image this again at the Texas Star Party with this camera on my 11-inch SCT to try and get better detail.  This time, I’ll try many shorter exposures though to see how that works, since this is a brighter target that could probably support that method.

Astrophotography is always an adventure!

Next post: #131 - Friday, April 20, 2018 – Astronomy Night with the Brownies

Table of Contents

#129 - Wednesday, April 11, 2018 - Library Telescope Program

Last year, my astronomy club furnished two small telescopes to one of the local libraries.  More and more astronomy clubs and libraries are teaming up to bring astronomy to the public this way, and it's fantastic!  Library patrons can check out a telescope to bring home for a week and do some astronomy from their backyard!  Several other libraries in the area thought they were awesome, so they bought their own.  One of the club members modified the scopes to make them more user-friendly and less likely to lose parts - the eyepiece gets glued in, strings attach the objective cover and eyepiece cover to the mount, a plastic plate is put over the end so the primary mirror doesn't get messed with, the button-cell battery in the red dot finder is replaced with two much more easily-accessible AAA batteries, etc.  Instructions and star charts are also included and are clipped onto the mount and attached with string.

The telescopes our local libraries have, and I think many other libraries as using as well, are Orion Starblast 4.5's, which are 4.5-inch Newtonian telescopes on tabletop alt-az mounts.

The first program I did with these was actually last August, just before the solar eclipse, as part of a "Millennials" series one of the libraries was putting on.  (Some decidedly non-milennials joined as well!)  I talked about the upcoming eclipse, how to watch it, and then also about the library telescopes, including a demo and looking at the some various objects.  The same library invited me back again on March 22nd, and I showed a group of about 15 people on a chilly but clear night at a local park how to use the telescope, and then we looked at a few things, such as the Pleiades cluster and the Mizar-Alcor double star.  I was actually rather impressed with what we were able to see with this scope!  And it was very simple to use.  I was glad for this, since as an imager, I have to shamefully admit I only have a rough idea of where the things I image are.  (The Rosette Nebula is somewhere off Orion's right shoulder...M81 and M82 are in the north somewhere...)  But I was able to find things pretty readily (I took some notes on where to find things using my SkySafari app, and since that scope has a fairly wide field-of-view, I just sort of hunted till I found it...we're mostly looking at bright things anyway!)

I liked this program enough that I got in touch with the four other libraries in the area who had purchased their own, and I have done this at two of them so far - two to go!  One of them we had to hold indoors, however, due to cloud cover and rain, so I'm not counting that toward my log.  (And I forgot to log the first one!)

For the indoors one on April 10th, I showed patrons (we had a group of nearly 30, including kids, which was amazing for that small community and the fact that it was cloudy!) how to use the scopes, and had them look through the eyepiece just at the wall to get a feel for it.  Then I pulled up Stellarium on my laptop and showed them how to find things like the Double Cluster, Orion Nebula, Polaris, and Mizar, and pointed them to apps and websites where they could find star charts and interactive sky maps.  

On April 11th, I did the program with a group of about 8 at another local park.  It was chilly, but not too bad.  We did have some fairly significant cloud cover roll in, however.  Fortunately, Sirius poked its face out through a thin section of clouds, so the patrons got to practice using the scope on an object that was very easy to find.  They seemed to get the hang of it, and I think they feel much more comfortable with the scopes now (several had checked them out previously, but had trouble figuring out how to use them).  There were also two young boys (maybe 5th grade) who asked some really smart questions on astronomy and astrophysics, like what is dark matter and what does it do, what will happen to the sun when it dies, things that I was more than happy to dive into and go way over time!  

I've got another program next week, and then one in July sometime!  And I'll probably repeat some of these with the libraries again in the fall.  I love doing outreach!

Next post: #130 - Thursday, April 19, 2018 – Eye of the Needle (and Utility of the Refrigerator)

Table of Contents

Combating Noise in Astro Images

When it comes to measuring anything, whether that be recording sounds, taking images, or scientific measurements in a laboratory, noise is an inescapable part of that measurement.  While nature adheres quite closely to the laws of physics, nature is also messy and chock full of uncertainty.

It is impossible to measure something with infinite accuracy - the exact point will always be a little bit spread out.  In addition, some level of background noise is always there.  In daytime photography, your signal will usually vastly overwhelm your noise, so the noise is not noticeable - think about the background sound of a a ticking watch while in a crowded cafeteria.  But in astrophotography, our signal is very weak, oftentimes just barely above the noise, especially if you are shooting from a light polluted location or with a camera operating at ambient temperature.

I have often had people show me some attempts they've made at astrophotography, and how disappointed they have felt when their single images are noisy and dark.  The key is, you can't stop at a single image - you've got to get statistics to work for you rather than against you.  Imagine drawing conclusions about how wind effects gas mileage when you have only driven one mile!  You have to increase your sample size in order to increase what's known as the signal-to-noise ratio, or SNR for short - how far above the noise your signal is, or how much more prominent the deep sky object (DSO) is over the background light and noise of your image.  When it comes to a good astrophoto, it's all about SNR.

Sources of Noise

There are several sources of noise:
Dark current - The basic function of a camera sensor is to convert photons into electrons.  The electrons generated over the length of a single exposure are held in each pixel, and then the accumulated charge is converted to a digital value of intensity, which is stored in the image file.  Cameras are not only sensitive to visible light, however - ambient heat and heat from generated in the battery and circuitry can also create electrons in the pixel, generating a false signal.  This is known as dark current.  If you put the lenscap on your camera, you can record that dark current.  It is random, however, since who can say whether a given pixel will be the one to record a bit of that heat?  In general, the intensity of the dark current will increase with the exposure time (so doubling the exposure time will double the intensity of the recorded dark current), and it will also double for about every 13 degrees Fahrenheit (6 degrees Celsius). Some pixels will also be more sensitive to this noise than others.  Below is an example dark frame, a 5-minute exposure on my Nikon D5300 at ISO-3200 and 72 degrees F.  The top image is the raw frame, but it can be hard to see (especially when it's converted to jpeg as it is here), so on the bottom is a brightened version.

 An example dark frame from my DSLR (top), and a brightened version of the same frame (bottom).

Read noise -  Wherever there is electronic circuitry, there is electronic noise.  Some of it shows up as a random pattern, since noise is an inherently random process.  Some of it will show up as a fixed pattern, since variations in the manufacturing process will cause some pixels to run "hotter" than others, or have some low level of baseline charge (known as offset or bias).  Some of it may show up as horizontal or vertical lines in your image, especially if you are using a CCD camera, since those are read off row-by-row instead of pixel-by-pixel.

Quantization error - The charge that is held in each pixel is an analog signal.  In order to create a digital image, however, it must be digitized and turned into a number - it must be measured.  This is done using an analog-to-digital converter (ADC).  Different cameras record images at different bit depths.  My new ZWO ASI1600MM Pro will record images at 12 bits.  This means that 0 is black (no signal was collected), and 2^12, or 4,096, is white (the pixel was saturated, meaning it held as much charge as it can store).  This means that the images saved out from the camera can have 4,096 shades of intensity.  Now, my camera has a well depth of 20,000 electrons, meaning it can hold 20,000 electrons before it saturates and can't hold anymore.  This means that every level of intensity is a difference of 5 electrons.  If we pretend the camera has perfect quantum efficiency (how well it converts photons to electrons), this means you can not differentiate between parts of an image that differ by fewer than 5 photons, since they will appear the same brightness.  This loses you that subtle detail in knots and Bok globules in a nebula, for instance.  Now, you could decrease the gain (or ISO) on your camera until the pixel could only hold 4,096 electrons, but then you lose dynamic range, since the separation between the brightest and darkest parts of your image is much smaller.  So it's a tradeoff.

Shot noise - Shot noise has less to do with the camera and more to do with the uncertainty that exists at the most fundamental levels of physics.  Due to how dim and distant the things we image as astrophotographers are, the photon flux, or rate at which photons arrive at your camera is quite low.  Imagine you are standing in a rainstorm; the raindrops are coming frequently enough that you can't tell the difference in how many raindrops are hitting you per second.  But if it's just beginning to rain, you have no idea when the next raindrop is going to arrive - a half second later, five seconds later, etc.  The same goes for photons - at these low rates, you don't know when they're going to arrive.  In one frame, 5 photons might hit your camera; in the next, it could be 12, or 3.  It may surprise you to know that the shot noise actually increases with intensity - but as the square root.  If you have 10 photons get absorbed by your camera in one frame, and then 100 in another, your signal has increased 10 times, but your noise has only increased by sqrt(10), or 3.2.  So even though the noise is higher, the signal is much higher.

Quantum and Transmission Efficiency - Quantum efficiency is essentially how good the sensor is at converting photons to electrons.  There is no guarantee that just because a photon strikes the detector that it will get converted to an electron.  If it doesn't, then it is lost.  The fewer signal photons you collect, the harder it is to distinguish your signal from the noise.  Transmission efficiency is how much light makes it through all of the optics and filters between your camera and the sky.  Generally, telescope optics are high quality and have high transmission.  Color filters, however, can lessen your signal.  This is especially true in DSLRs, where the Bayer matrix (the array of red, green, and blue filters laid over the top of the sensor so that you can image all three colors at once) can have relatively low transmission.  I borrowed the chart below from one of Craig Stark's presentations on astrophotography (found here, and it is a great resource!).

The top graph compares a monochrome CCD camera, the QSI 540, to is color version, the 540c.  The bottom chart compares the monochrome QSI 540 to a Canon 40D/50D DSLR.  As far as the filters themselves go, independent from the camera sensor (the images above include the camera's responsiveness to those wavelengths), they can have nearly 100% transmission, such as with the Astronomik LRGB Type 2c filters I have (and love).

Again, fewer photons means it's harder to distinguish the signal from the noise.

Light pollution - Light pollution isn't quite the same as the other sources of noise, but it can certainly result in a lot of problems.  Light from a nearby town reflects of particles in the atmosphere, and the camera will capture those as well.  Think about trying to see a dim cell phone screen out in full daylight versus in a dark room - it's a lot harder to get much contrast in the galaxy you're trying to image when the background light is just as bright!  In order to get more signal, people will usually turn up the gain or ISO on their cameras, which increases the camera's sensitivity not only to light, but also to heat and read noise.  In addition, remember shot noise?  Light pollution is also a source of shot noise, except it doesn't contribute to your signal (the object you're imaging), so it's only adding additional noise.
Take a look at the two images below - the top one was taken in a suburban/rural transition sky (5 or yellow on the Bortle Scale), and the bottom in the dark skies of west Texas.  Note that the top image used a light pollution filter - the useful thing to look at here is the contrast between the object and the background.

5-minute frame, ISO-1600, from a light-polluted location

 6-minute frame, ISO-1600, from a much darker location

Okay so the bottom frame is one minute longer exposure time, but you get the picture.

How can we ever hear the beautiful music over all of this noise??

It sounds like a daunting task!  Don't fret, however - we have the power of digital processing at our fingertips.  Stacking, calibration, and post-processing are extraordinarily powerful tools that let us turn noisy messes into beautiful recreations of the universe's many wonders.

I've got oodles of examples of subframes that look noisy and terrible, and processed frames that look way more awesome.  As far as distinguishing the target from light pollution, though, this one takes the cake.

I have wondered for some time now, is it possible for me to quantify how much better the images get with stacking and processing?  The answer is yes, of course, if you don't mind a little math!

Stacking

The whole point of stacking (see this post for a tutorial on how to stack astro images) is to increase the certainty that the light in a given pixel is "real" and not noise or light pollution.  If you have one picture, and you look at a given pixel, you may not be able to tell.  But if you have 20 pictures, and in 19 of them the pixel is just about the same color and brightness, you can be pretty sure that that is the real value of the pixel.  Stacking is a statistical process that increases that certainty, and the bottom line is you get an increase in the SNR.  In general, your SNR increases by the square root of the number of frames you, but you can see much larger gains by applying calibration frames and doing some post-processing, as you well see evidence for in a moment.

Calibration

There are two kinds of calibration images that will help you reduce noise - darks and biases.  (Flats reduce vingetting, or the darkening of the corners of your image, so I'm not including those here).  Darks record your dark current, and biases record your read noise and fixed pattern noise.  It is important to note that your dark frames will also have read/fixed pattern noise, but apps like DeepSkyStacker handle the subtraction so that the biases don't get subtracted twice.  For more on how to capture calibration frames, see this post.  Now, remember how I said noise is random - you can't take a dark and a bias frame and just subtract them.  The distribution of noise moves around like static on an empty analog broadcast TV channel.  Here again we get some help from statistics.  You take several dark and bias frames, and then DeepSkyStacker or your other favorite stacking app will average them and subtract a "master" from your stacked image (where the noise has also been averaged).  In Gaussian statistics, which most noise sources in nature are (the class bell-shaped curve), the average value approached the truth.  If you have a noisy camera (like a DSLR), you're going to want more dark frames.  I find that DeepSkyStacker struggles with more than about 60 (and that's if you've undergone the process of expanding its RAM-using capability from 2 GB to 4 GB - see this website for how to do it (it does require Microsoft Visual Studio, but the Community Edition (the free one) will do it)).  (Wow, nested parentheses!)  I usually use 20.  I wouldn't go less than 10.  (Again the square root law applies here, the mainstay of Gaussian statistics).

Other considerations

Having a cooled camera sensor makes a world of difference, as I have already begun to see in the first images from my ZWO ASI1600MM Pro.  Your dark current diminishes dramatically, which is a huge source of noise.  The combination of a cooled sensor, longer subframes, more subframes, a lower-read-noise chip, and much higher quantum efficiency and transmission efficiency combined to drastically decrease the noise between these two images of the Orion Nebula.  

Nikon D5300, 12x60s frames at ISO-1600, taken on an Orion ST-80 ahchromatic refractor, ambient temperature = 36 F (2 C)

ZWO ASI1600MM using Astronomik LRGB filters, total 113x60s frames, gain=unity (139), taken on a 140mm Vixen neo-achromat refractor, sensor temperature = -30C (-22F)

Light pollution

Darker skies will give you greater contrast, making it much easier to pick up dim details (see the Whirlpool Galaxy images above) and enabling you to distinguish very dim signal from the noise and background light.  They will also decrease the extra shot noise added by light pollution.

All right, show me the numbers!

For my experiment, I chose a dataset where I actually had enough subframes to measure the difference in SNR in stacking greater numbers of subframes - usually I am rather impatient and don't gather more than about 25 subframes.  I picked M8-M20 #2, an image of both the Lagoon and Trifid Nebulae (M8 and M20) captured on the back side of Casper Mountain on August 17th, 2017 while I was there for the solar eclipse.  It was taken with my Nikon D5300 attached to my Borg 76mm apochromatic refractor, using a Hotech SCA field flattener.  The telescope was attached to my Celestron NexStar SE mount (chosen for this trip so I wouldn't have to polar align it before dawn the morning of the eclipse, since this mount is an alt-az mount).  The subframes are 30 seconds long (the NexStar has some serious periodic tracking error) and ISO-1600.  The temperature was between 53-55F over the course of the acquisition of that dataset.  All stacks were done in DeepSkyStacker, and with the exception of the image I stacked without calibration frames to measure the difference in SNR, I used 20 darks, 20 biases, and no flats (I didn't have any for that scope yet, it was fairly new to me).  I stacked 10 frames, 44, and 88, and then stacked 88 without the calibration (dark and bias) frames (all using the auto-adaptive weighted average stacking option, my preference as of late).  I saved out the raw 16-bit TIFFs with changes embedded, not applied, and didn't do any adjustments in DSS.  In Photoshop, I stretched the histograms of the stacked images, and the 88-frame one the calibration files I used my post-processed, completed image.  I didn't make any additional adjustments to the 10 or 44 stacks, or the 88 stack without calibration.

The method of calculating SNR is quite simple: take a sample of the image over a flat area (on your DSO, since that's what you care about the most), so either in a nebulous region that isn't changing brightness much and doesn't include any stars, or between spirals of a galaxy, or something like that.  Grab the mean and standard deviation of that area.  SNR = mean / standard deviation - that's it!  It's a unitless value, although if you are into radio or other kinds of signal transmission and love your decibels, you can convert SNR to decibels using dB = 20*log10(SNR).  (log10 = log base 10, if that wasn't clear).  

Now, you need all of the images your comparing to be positioned the same way so that you are sure to grab the exact same area of each image, since the SNR over the region you are sampling will change depending on where you are sampling it.  I did this by opening up each of the images I compared in Photoshop, cropped them to be the same size, and then copied them over one of the images all as layers.

The Layers pane in Photoshop (if you don't already see it in the lower right corner: Window -> Layers)

The order these are in does not matter.  I turned off viewing all of the layers except for the base layer (my finished image) and each image at a time by clicking the eyeball box to the left of the layer thumbnail, clicked on the image I wanted to align (in the screencap above, the "single frame" layer), and turned the opacity down to about 50-60% (the "opacity" selector box).  Then I hit Ctrl + A for Select All, and clicked on the Move Tool on the left panel.  Then I used the arrow keys to nudge the image until it was in line with the base image by looking at the stars.  Get it as close as you can, knowing that there is probably some sub-pixel shift and you won't be able to get it exactly.  

"Single frame" not aligned with "Complete"

 The two images are now aligned (or are at least close)

Repeat for any other versions of the image you want to compare - different numbers of stacked frames, calibrated vs not calibrated, using different stacking methods, using a different number of darks and biases, etc.  Hit Ctrl + D when done to deselect.

Next, zoom in on your DSO, and try to find a flat area without stars or much change in brightness.  The larger your sample, the better, but if it's too large you'll get true variation in the DSO, which will skew your measurement; I used a 25x25 pixel box.  This is 625 pixels total; the square root of that is 25, which is 7.7% of 625, so I'll have an inherent 7.7% error in my SNR measurements (less than 10% is safe, by rule of thumb). You can set the size of the selection box by clicking the Rectangular Marquee Tool (the dashed-box-shaped icon on the left panel), change the "Style" to "Fixed Size," and set your width and height.  Click the area you want to put the selector box there.  I use brightest image of my comparison set so I can know what I'm looking at to choose the area - in this case, the completed image (again by clicking the eyeballs to toggle viewing the layers).

If you go to Window -> Info, you can have a panel open that shows your coordinates (in inches by default, although you can click the options button (the three lines in the corner of the panel) and change it to pixels, centimeters, what have you).  I'd record these wherever you are writing down your measurements for future reference, and I use the upper left corner of the selected area.  

Finally, in the Histogram panel, click the options button (three lines) and select Expanded View and Show Statistics.  This will show you the statistics you need.

In Source, select "Selected Layer" so that you're only measuring the image layer you have selected.  In Channel, select Luminosity.  This is so you aren't taking color variances into account.  The human eye notices differences in luminance far more than chominance anyway.

Now we are ready to roll!

First, here is a raw, single frame (converted to jpeg of course).

It is quite dark.  If you zoom in, you will see the noise.

Single raw frame.  You will also see how good the tracking is on my NexStar mount. (sarcasm)

I recorded the mean of that square to be 20.15, and the standard deviation to be 5.07.  Dividing those two, we get a SNR of 3.97.  This means that your signal doesn't rise very far above your noise, a fact that is easy to see in the image above.

Next, I selected the layer that is my stack of 10 frames, and recorded a mean of 43.56 and standard deviation of 2.72, which yields SNR = 16.01.  Just by stacking 10 frames and doing dark and bias subtraction, we have quadrupled our signal to noise ratio!

Stack of 10 frames, dark and bias subtracted.

Now, statistics says a stack of 10 will only get us a SNR increase of sqrt(10) = 3.16, and 3.97 x 3.16 = 12.5.  But again, dark and bias subtraction help us out.

Next, I measured a stack of 44 frames (half of the maximum), and recorded a mean of 35.99 and standard deviation of 1.58.  This means a SNR of 22.77.  This is a 1.4x increase.  You can see our diminishing returns happening already, but it's still a solid increase.

Stack of 44x30s frames, dark and bias subtracted.

Finally, I measured my stack of 88 that I post-processed - stretched the histogram, adjusted the light curves, adjusted the color balance (not important for noise), clipped the left end of the histogram (important for background and dim noise reduction, although you also lose real signal that's lost in the mix), and denoising blurring algorithm.  I recorded a mean of 99.07 and standard deviation of 2.45, which gives us a SNR of 40.44.  Fantastic!

By stacking, calibrating, and post-processing, we have increased the signal-to-noise ratio by 10 times!  And it shows - compare the single raw frame to the final product.  

Like all good research, let's summarize the results in a nice Excel table.

"Increase from single sub" here is the factor of the increase of the SNR - SNR of the frame / SNR of the single frame

I am going to look into some image quality metric tools as well, but I wanted to do this as a warm-up.  What a fun exercise!

Bottom line: Signal-to-noise ratio matters a lot!  Cameras are noisy, but we can beat that down with the power of statistics and get some really nice-looking space images, even though the odds can seem to be against us.  

Whew!  That was a long post.

Next tutorial: Image Processing Method 2.0

Table of Contents

#128 - Thursday, March 15, 2018 - The Clear Skies Call, and I Must Go

With springtime due to arrive any day now, the weather goes through its usual rough transition as winter grasps each day and refuses to let go.  Cloudy days can turn into sunny days at the drop of a hat, and vice versa, so when I saw it was sunny on Thursday, I waited until the last minute to decide whether to make the trip out to the observatory or not.  The forecasts made it look like there would be occasional clouds rolling through, but nope, it was clear as a whistle the whole evening!  In fact, I saw more stars than I've seen in a while from the observatory, meaning that transparency and seeing were much better than predicted.

I got out there a little after 7 with the hopes of taking flat frames before the sun set, but alas, the sky was darkening too quickly to get flat frames that were all the same exposure time, once I figured out how to use SharpCap's flats tool.  I'll just have to deal with vignetting in my images a little while longer.  I did mark the orientation of my camera to the filter wheel, and the filter wheel to the topmost screw on the club's refractor, with a silver sharpie so that I can align my camera the same way with reference to the telescope every time, so I can keep using the same set of flat frames.

Despite the beautiful skies (or maybe because of them), knowing that luck can only extend so far, I faced a number of other challenges this evening.  After I got the camera attached and marked, I slewed to Dubhe, the star on the top outermost edge of the ladle of the Big Dipper, so focus the scope.  I had a hard time finding the right camera settings that would let me easily see the diffraction spikes from the Bahtinov mask focusing aid (see my little blurb on this here), so I switched to the FWHM focusing mode in SharpCap (SharpCap has some really helpful focusing tools, by the way!) and used that to achieve focus by minimizing the size of the star.  This took a while, especially because even though I finally bought a USB 3.0 extension cable to connect my USB hub to my tablet, which now means I have USB 3.0 all the way through from my camera to the computer, I was still only getting less than a frame per second viewing speed.  I'll have to work on figuring that one out.

My next blockade came with calibrating guiding.  I slewed over to Regulus to calibrate PHD, since it says it will have the least amount of guiding errors if you calibrate on a star between -20 degrees and +20 degrees in declination, and I had such trouble guiding on my last trip.  I had it auto-select a bright star in the region (I slewed a little away from Regulus so that it wouldn't swamp the sensor), and it started calibrating.  But the star didn't move!  Calibration involves PHD moving the mount west, east, north, and south by different amounts to see how the star moves so that it knows how much and in which direction to pulse the mount so get the guide star back into place while it's guiding.  (See this post for more on guiding).  So you will normally see the star move during the calibration process.  I turned on my imaging camera to 30s exposures, and yes, the mount was indeed moving!  So I put the cover over the guide scope, aaaaand what I thought were stars were actually hot pixels!  Usually they're a little easier to differentiate from the stars.  I didn't have a set of dark frames loaded into PHD for my guide camera because I've had issues with it doing the dark subtraction wrong and the image turning out all gray, but I went ahead and took another set of dark frames anyway.  About halfway through that process, I accidentally hit the spacebar on my tablet, which hit the "stop" button and cancelled the whole thing!  So I had to start over...It worked this time, and all but one of the hot pixels disappeared.  So I calibrated (which went very quickly since I was looking so far south), and then slewed back over to M81 & M82, my target for the evening.  My plan was to get the rest of the RGB data I didn't get to collect last Saturday.

I set SharpCap to take repeating exposures so I could center M81 & M82 the way I wanted, but I couldn't see them at all, so I turned up the gain all the way and did 30s exposures to center them.  This took forever.  Finally I got them where I wanted them.  (I really need to learn how to use AstroTortilla, a plate-solving program where you tell it where you want the FOV, and it uses star patterns in the image with shorter exposures to center your telescope!)  I started acquiring the red images and waited for the first one to come in so I could see how it looked.  Well, it looked out of focus - there's always the risk of something moving ever so slightly when you slew the telescope a fair distance across the sky.

Out-of-focus red frame

In-focus red frame

So I took a screenshot of the coordinates where I had the scope set to (I control it from the computer - way easier than moving through menus on the hand controller!) and slewed to Dubhe to go through the whole rigmarole I went through earlier in the evening to focus.  Finally, I slewed back to the coordinates I had saved, which got me close but not exact, so I had to adjust it a little.  But this time, I used Sequence Generator Pro's Frame & Focus routine, and SGP has a tick box called "stretch the histogram," which will stretch the histogram of the preview image on the screen so that you can actually see things.  This let me take only 4s exposures, so centering the image went waaaaay faster.  Remind me to do it in SGP from now on. Finally, an hour and a half after I got started, I was ready to image.

I acquired 7 more red frames (since I had gotten 8 last Saturday and I wanted 15 total for each channel), and when I got back to swap filters, I realized that I didn't move the dome enough earlier, and part of the scope's view was cut off.  So I lost a few red frames and had to re-take them.  But it turns out I lost more than I thought, so I only wound up with three red frames.

The dome cut off part of the scope's field-of-view - notice the diffraction spikes around the brighter stars.

After that, I changed to the green filter, and those went off without a hitch.  But after the greens were done, it was time to do a meridian flip, since M81 & M82 had crossed the meridian that divides east and west.  Equatorial mounts have to flip around for each part of the sky in order to keep the telescope from hitting the mount.  Doing this always means adding a lot of extra time - usually I have to re-focus again afterwards, and it doesn't always go back to the exact same spot in the sky, so I had to re-center the targets again.  But first, I had to deal with another problem: the guide camera fell out of the guidescope while the mount was flipping around.  This meant that I had to re-calibrate, since it would be impossible to get it back into the exact same orientation it was when I calibrated earlier.  I didn't want to slew all the way back over to somewhere south, so I just calibrated where I was at.  It worked just fine.  In fact, the seeing was good enough that guiding was more accurate than usual - I could see the pattern of the periodic error in the mount, and I was getting nearly only 1 arcsec of error, which is great for this area.

Not the best guide graph there ever was, but not bad...and good enough for how large the FOV of my camera + telescope combination is (1.3x1 degrees)

The blue frames went off without a hitch, and then I finally got to pack up and go home around 1:15 AM (I had only planned to stay until midnight, but I reaaaallllyyy wanted to get the rest of the frames I needed to make a complete image).  I got to sleep around 2 AM, and still had to go to work the next morning!  I didn't feel all that tired on Friday, but I had a very difficult time typing, either on my phone or my computer - I kept making lots of mistakes.  But I powered through!

I've gotten good at stacking LRGB images in DeepSkyStacker (post on that coming soon), but post-processing in Photoshop proved to be a challenge.  Color balance was difficult.  But the synthetic flats procedure I recently learned worked really well (post on that soon too).  I'll eventually write a post on the post-processing process, but here's a quick rundown:

These are the raw FITS files for each color channel.  They look blown out, but they're not - the FITS viewer I use (AvisFV) stretches the histogram a bunch, or else you wouldn't really be able to see them.  Blue is reversed from the others because I had to do a meridian flip.

After stacking in DeepSkyStacker, saved as 32-bit, and imported into Photoshop.  Notice that DSS aligns them (I'll explain how to do this in a future post)

Stretched & flattened with synthetic flats

RGB images combined in Photoshop

Luminance image applied as Luminosity layer

And...drum roll please...after additional work in Photoshop...

Details:

Date: 10 & 15 March 2018

Object: M81 & M82

Camera: ZWO ASI1600MM Pro

Telescope: Vixen 140mm neo-achromat

Accessories: Astronomik LRGB Type 2c 1.25" filters

Mount: Losmandy Gemini II

Guide scope: Celestron 102mm

Guide camera: QHY5

Subframes: L: 18x180s

   R: 11x180s

   G: 18x180s

   B: 14x180s

   Total time: 3h3m

Gain/ISO: Unity (139)

Stacking program: DeepSkyStacker

Stacking method (lights): Auto-Adaptive Weighted Average (x5)

Darks: 20

Biases: 20

Flats: 0

Temperature: -30C (chip), 29F & 35F (ambient)

See on AstroBin

I still didn't manage to get the red hydrogen gas being ejected from M82, but hopefully someday...I'll need to take much longer exposures.  Or use a hydrogen-alpha filter :)

Pretty excited about this camera and my increasing skill level!!

Next post: #129 - Wednesday, April 11, 2018 - Library Telescope Program

Table of Contents

#127 - Saturday, March 10th, 2018 - Daylight Saving Time, and Other Tales of Woe

The forecast coming up on Saturday night was uncertain - cloudy, clear, clouds at 6 PM gone by 8 PM, clear until 11 PM and then clouds...none of the forecasts agreed, and when that happens, there's sure to be at least a few thin clouds floating around.  There were, but they mostly stayed toward the horizon, so a dedicated few of us from the club layered up to brave the cold and do some astronomy, as well as prepping for next Saturday's Messier Marathon.  (The Messier Marathon is a clear night in mid-to-late March when a dedicated observer can power through all 110 Messier objects in a single glorious night).

Due to getting sidetracked spending the whole afternoon tearing my house apart in search of my Raspberry Pi 2 that I need for a workshop I'm attending this week, I didn't leave the house until sunset, so it was dark by the time I finished setting up.  I had hoped to go out early so I could take some much-needed flat frames (see this post about what those are good for), but I forgot my silver sharpie to make index markings of where I took them anyway.  (All of your gear needs to be the same orientation every time if you want to use the same flats).  I need to get like a LED tracing tablet or something.

I spent the drive out to the observatory contemplating what target I wanted to shoot, and I finally settled on the Cone Nebula, which I have not yet imaged.  It was a risky decision, since I wasn't sure how well it was going to show up or how exactly to orient the image, since the view I see in SkySafari (the astronomy app I use to know where things are) can't tell me which way my camera is oriented with respect to the object.  It was also already across the meridian, meaning it was headed quickly towards the most light-polluted area of sky to the west of the observatory.  But I figured I'd give it a shot anyway.

After spending a while making sure I had good focus and then getting it centered exactly how I wanted it, I was finally ready to start imaging.  I had to take 1-minute-long subframes in order to see it, so centering the image took a while.  (And even then, I had to blow out the image just so I could kind of see the Cone).  I calibrated guiding, started the run for 5-minute exposures, and went inside.

Well, before I went inside, I took a few peeks through the club's 20-inch Dobsonian that the other club members present had pulled out - we looked at M51 Whirlpool Galaxy, M65 & M66 (part of the Leo Triplet), and M42 Orion Nebula, and they looked at a few other things while I was setting up.  M51 showed the two cores quite nicely, and if you used your imagination you could just see the spirals.  M65 & M66 had fuzzy shape to them and looked nice together.  M42 had its usual greenish cast, with the Trapezium nicely resolved in the center, with the giant curved dust lane of M43 just visible against the star that lights up that region.

I came back out 30 minutes and a cup of hot cocoa from the Keurig later to find that guiding was a mess - all of my images had some motion in them.  I stopped and re-started guiding, but was still having troubles, so I finally gave up and decided to image M81 & M82 instead, a photogenic pair of bright galaxies up north in Ursa Major.  After taking much care to get them centered exactly where I wanted them, I went to re-calibrate guiding, and PHD gave me a warning about attempting to calibrate north of +20 degrees declination, since when you're pointed north using an equatorial mount, there is not much motion in the declination axis, which can cause problems with calibration.  But I had already spent so much time centering the image that I went ahead and calibrated anyway.  It worked all right.  I took 20x3-minute subframes in luminance, and then 8 in red (I didn't want to be up too late), and then I started 8 in green and went inside, but when I came back out, the mount was hitting the pier!  The Losmandy Gemini mount that the club's memorial telescope has can go quite a ways past the meridian, so I was hoping to finish before it hit a hard stop, but that didn't work out.  So I did the meridian flip, re-centered the image, had to rotate the camera since it had gotten bumped, re-focused as a result of that, and then finally re-started the green frames, only to have guiding not work (you can flip the calibration without having to re-calibrate).  It was weird - the graph that shows the corrections that PHD is sending to the mount was totally flat, which would normally indicate that maybe the image coming from the guide camera wasn't actually updating, but when I bumped the mount, it sent a correction command (albeit a tiny one).  But I had previously been getting upwards of 1.5 arcsecs of RMS error, a measure of how much PHD was having to correct the mount, and there was no way it was even possible to be flat like that(it would have require perfect atmosphere conditions and no mount tracking errors).  At that point, it was already nearly 1:30 AM, and I had to get up and go to church the next morning, especially since it was Girl Scout Sunday and two of the girls in my troop were receiving a religion award they had earned.  So I packed everything up and headed for home.

To add to this underwhelming night, I left the observatory at 1:45 AM and got home at 3:15 AM - thank you Daylight Saving Time!  Ugh.  I only got 4 hours of sleep, and that was with skipping a shower.  I may or may not have drank coffee during Mass...

If nothing else, I did get a usable stack of L images, although I don't have flat frames for them.  I cropped the image to remove at least some of the vignetting.  It looks promising, though.

Date: 10 March 2018

Object: M81 & M82

Camera: ZWO ASI1600MM Pro

Telescope: Vixen 140mm neo-achromat

Accessories: Astronomik LRGB Type 2c 1.25" filters

Mount: Losmandy Gemini II

Guide scope: Celestron 102mm

Guide camera: QHY5

Subframes: 18x180s (54m), L only

Stacking program: DeepSkyStacker

Stacking method (lights): Auto-adaptive weighted average, x5

Darks: 20

Biases: 21

Flats: 0

Temperature: -30C (chip), 29F (ambient)

Nights like this are a good reminder to all that astrophotography is hard!  There are a lot of things that have to come together in order to successfully make an image.  But if you persevere, great images will result eventually!

Next post: #128 - Thursday, March 15, 2018 - The Clear Skies Call, and I Must Go

Table of Contents

#126 - Friday, March 2, 2018 - Curing Cabin Fever

It has certainly been a cold, cloudy winter this year!  I've only been out eight times since winter began, and one of those was an outreach event.  The last time I was out imaging was over a month ago, so I'm glad to be back!

And I'm especially glad because I have a new camera!!  After saving up for the past year or so, I have finally saved enough to buy the ZWO ASI1600MM Pro for $1,280 (that's the low end for astrophotography cameras!)  It arrived on Wednesday.  Now, the rule in astronomy is that once you get a cool new telescope or a fancy new camera or eyepiece or something you're really excited about, you are bound to have clouds for at least a week or two.  However, I was strategic in my timing: it has been cloudy here for the past few weeks with lots of rain, and now that it was finally cleared a bit, the moon is full.  But!  The full moon gives me a chance to go test the camera to make sure everything works and get some preliminary results on it.  I just had to choose a bright target.

The ZWO ASI1600MM Pro astrophotography camera.

So, of course, I went with the classic first-light target: the Great Nebula of Orion, Messier #42.
I'll get into the details of this camera in a separate post soon.

The weather was cold, hovering around freezing, but there was no wind, so it was tolerable with a sufficient number of layers, as well as my new snowboots paired with my thermal and wool socks.  The Asian beetles inside the memorial dome appear to not have reanimated in the recent spout of warm weather we've had, although there were a few moving flies inside.  The carpet squares on the floor were soaked from the rain, which must have leaked in from somewhere, but the electronics all seemed fine.

Setup went smoothly, thanks in part to some at-home setup - before I left, I scrambled to get the drivers updated and tested.  The ZWO English software site was down, but I managed to find the Japanese version of the software page, where I was able to locate both the latest camera driver and the ASCOM driver, which were necessary to connect this new camera model to my computer.  (I had older versions of both drivers that I had used last May at the Texas Star Party to run the ZWO guide camera I borrowed from my minion Miqaela).  Crisis averted!

My new ZWO ASI1600MM Pro attached to the Vixen 140mm refractor I borrow from my astronomy club.

The atmosphere wasn't great - I was getting around 2.8 for the FWHM of the star I was focusing my guide camera on, and the guiding was bouncing all over the place with a RMS error bigger than 2 arcescs, which lead to unstable guiding (even with 3-second exposures) that squished my stars this way and that way on all of the subframes, but being under a full moon anyway, I was less concerned with that because the main goal was to test out the camera.

I first tried the Horsehead & Flame Nebulae, but they were not bright enough to shine through the moonlit sky, even with a 5-minute exposure.  So next door to the Orion Nebula I went!  I decided to take 60-second subframes since I didn't want too much background light from the moon.  I also took a series of 5-second luminance subframes to composite with the 60s ones so show detail in the core  of the nebula, which was waaaay blown out in such a sensitive camera at 60 seconds.

60-second luminance frame, stacked (39x60s), stretched 

5-second luminance frame, stacked (45x5s), stretched

I tried a technique from this video by David Rankin to blend the two luminance frames in order to capture the dimmer areas and finer detail of the 60s image with the detail in the brightest parts of the nebula of the 5s image.  It worked quite well, and was pretty easy!

After stacking and stretching the red, green, and blue channels, I combined them into one image in Photoshop (I will eventually write a tutorial on this once I have more practice), added the composited luminance frame on top and set the layer type to Luminosity, and set to work tweaking the Levels, Curves, and Hue/Saturation to my liking.  The colors came together amazingly well - I used my new Astronomik LRGB Type 2c filters, which say they are designed to be color-balanced in sRGB space, the standard color space.  That definitely proved to be true - I had to do very little adjustment of the colors, really just reducing their presence in the background (from the moonlight and light pollution). 

All right, here it is!

Date: 2 March 2018
Object: M42 Orion Nebula
Camera: ZWO ASI1600MM Pro
Telescope: Vixen 140mm neo-achromat
Accessories: Astronomik LRGB filters
Mount: Losmandy Gemini II
Guide scope: Celestron 102mm
Guide camera: QHY5
Subframes: L: 39x60s (39m)
                                       L: 45x5s (3m45s)                   
                                    R: 26x60s (26m)                  
                                  G: 26x60s (26m)                 
                 B: 22x60s (22m)
Total time: 1h57m
Gain: 139 (unity)
Stacking program: DeepSkyStacker
Stacking method (lights): Auto-Adaptive Weighted Average (x5)
Darks: 25 (60s)        
 30 (5s)
Biases: 21
Flats: 0
Temperature: -30C (chip), 29F (ambient)

The fine details in the nebula clouds are so amazing!  No longer are they lost due to noise.  I didn't have to do any noise reduction on this image, even with the short exposures.  I feel like I can reach out and touch it!  I also saw that if I crank up the exposure a ton (which makes it look kind of terrible), I also caught a lot of the very dim background of this nebula, so maybe I'll do some much longer exposure images of it as well and composite those together to make a very high dynamic range image.  Also, the giant blue stars are a result of using an achromatic refractor-type telescope, so at some point I'll need to image this again with my apochromatic Borg to get some better-looking stars.  But anyway, despite a full moon, I am super happy with this image, and can't wait to see what else I can do with this camera!  Time to take things to the next level!

Next post: #127 - Saturday, March 10th, 2018 - Daylight Saving Time, and Other Tales of Woe

Table of Contents