Flat Fielding

Partial Success

On June 7th puffy clouds passed by on a windy day. There were large gaps between clouds, and the temperature was 80 degrees. Although solar features were not particularly spectacular this day, I was eager to try two new procedures to improve my solar imaging.

The first new procedure was a change to settings within the FireCapture software that controls my camera. I reduced the frame rate during video capture from 30 to 10 frames per second, and I increased the "save thread priority" in the performance settings. Both these changes were designed to increase the video capture rate and prevent capture from freezing while videos are recorded. The changes seemed to work! 400-frame videos were completed in 6.67 seconds! I was very pleased to have fast continuous video capture throughout my entire observing session.

The second new procedure was to obtain flat field images to remove annoying interference fringes from my pictures. A flat field is the image produced by a uniform light source. In a perfect world the flat field image would be featureless, with uniform brightness across the entire image. But if there are specks of dust on glass surfaces through which light passes, if the optical system causes the camera's field of view to be unevenly illuminated, or if there are interference fringes present caused by multiple reflections within the optics, then the image of a uniform light source will not be uniform. The flat field will record deviations from uniformity, and it can be used during image processing to remove the imperfections.

Expert solar observers have used plastic bags stretched over the telescope's objective lens to produce a uniform light input for flat fields. When the telescope is pointed at the center of the Sun's disc, the plastic bag, apparently, diffuses sunlight and smears surface features making light reaching the camera uniform except for optical imperfections. Some observers have used white plastic supermarket bags. I initially tried a white bag, but it blocked too much light. When I tried a clear plastic bag, it worked! The next image shows the plastic bag attached to the front of my telescope.

The next image shows the flat field image recorded with this setup when a 3X Barlow lens was used with my camera.

Notice the alternating dark/light vertical bands on the right side of the image. Also present are a set of very subtle oval rings, the dreaded "Newton's Rings". Next is an image made through the 3X Barlow lens without applying the flat field correction.

Notice the alternating dark/light vertical bands on the right side. Now look at the next image where the flat field correction has been applied.

Hooray! Flat fielding eliminated the vertical bands! Both of the previous two images have been identically processed except for the added flat field correction in the second image. In this example it's hard to see if flat fielding also eliminated the oval "Newton's Rings" because the rings were so subtle in the uncorrected image.

Removal of the vertical bands finally allowed construction of an unblemished mosaic with the 3X Barlow. I combined 9 individual images to produce the following picture of the eastern side of the Sun. (Click on the image for a larger view.)

After finishing with the 3X Barlow I installed a 5X Barlow and proceeded to record images for the same view presented above. The flat field recorded with the 5X Barlow is the following picture.

The 5X Barlow flat has the same vertical bands as the 3X Barlow flat, but the 5X flat also has much more pronounced "Newton's Rings".

Next is an image made with the 5X Barlow without applying the flat field correction.

Vertical bands are visible on the right side, and the oval pattern of "Newton's Rings" is also present with more intensity than the relatively subtle pattern seen in the uncorrected 3X Barlow image. The next image shows the result of applying the flat field correction.

Flat fielding eliminated the vertical bands, but the oval fringe pattern became worse! The previous two images were identically processed except for the application of the flat field to the second image. Geez! Why can't life be simple?

After considering this surprising effect for some time I finally realized the oval fringe pattern recorded in the flat had somehow picked up a 180 degree phase shift relative to the vertical bands recorded in the flat! I tested this hypothesis by applying an inverted flat to the original image. (In the inverted flat positive and negative are reversed.) The next picture is the result of this action.

Just as I suspected! Now the oval fringe pattern has been removed, but the vertical bands have been made worse! So, the process of creating the flat field for the 5X Barlow somehow created a 180 degree phase shift between the vertical fringes and the oval fringes. Light that produced a dark vertical fringe somehow produced a bright oval interference fringe instead of a dark oval fringe.

In spite of the failure to completely remove the interference patterns for the 5X Barlow I was able to salvage two decent images. The next picture is a 10-panel mosaic of the Sun's eastern side similar to the 3X Barlow mosaic image above. This 5X image is imperfect only in the upper left corner where some oval fringe pattern is most visible.

A cropped portion of the previous image shows good detail with no fringes. (Click on the images for a larger view.)

In the future I'll try to remove the troublesome "Newton's Rings". Perhaps a camera tilt in addition to the flat field will work.

Still Testing New Camera

Using a 3X Barlow Lens

May 23rd was the last mild observing day before scorching, humid, summer conditions began. It was 65 dry degrees with a light breeze. The Sun was very quiet with only a few small sunspots, but the excellent, comfortable conditions made getting familiar with my new camera enjoyable.

On this day I used a tilt adjuster between the camera and telescope to attempt removal of annoying dark interference fringes which appeared on previous images taken with a 3X Barlow lens. The very first video showed stray light leaking onto the camera chip from an uncovered gap in the tilt adjuster. I fixed this with black electrical tape. Since the interference fringes are not visible in the live laptop screen preview, I processed the first videos immediately to see if the fringes were removed by the tilt adjuster. I didn't see any fringes in the individual images, so I thought the tilt adjuster had worked. I then proceeded to record videos for a mosaic of the entire Sun. Eventually I constructed the 17-image imperfect mosaic shown in the first picture below.

The mosaic shows lots of prominences and filaments, but few sunspots. I like the resolution of details with the 3X Barlow. (Click on the image for a larger view.) Unfortunately, the tilt adjuster did not entirely remove the vertical interference fringes. They became more subtle, but they are still present as you can see on the image below. This is disappointing because they ruin an otherwise nice picture.

The inverted disc image looked fairly nice with the fringes barely apparent.

No fringes are visible in this image of the prominences alone:

Next is a 3-image mosaic of a particular region entirely within the solar disc. The picture has been processed to increase the visibility and contrast of features. (Click on the image for a larger view.) . The fringes are barely visible and detail is quite good.

Lots of trial and error remain before operation of my new camera becomes routine. Perhaps I can remove the interference fringes with a flat field. I'll try this next.

Testing New Camera

Interference Fringes

May 13th was a lovely day with 72 degree temperature, scattered clouds, light breeze, and, frequently, good seeing. It seemed like a good day to try different Barlow lenses with my new ZWO ASI174MM camera.

The first picture below, a 5-image mosaic, was made with a 3X Barlow lens. A large magnificent prominence adorns the northeastern limb and a string of sunspots are arranged on a diagonal line running down to the right from the prominence. (Click on the image for a larger view.)

The 3X Barlow seems to give better results than a 2X Barlow. I'm quite satisfied by the resolution of details in this image. Unfortunately, the picture is ruined by dark vertical bands. These are interference fringes caused by multiple reflections of monochromatic hydrogen alpha light within the optics of my system. I've encountered interference fringes before with my higher power Barlow lenses, but those fringes, often called Newton's Rings, were circular, indicating multiple reflections between a curved surface and a flat surface. The vertical bands above are almost perfectly straight, more like interference fringes caused by two nonparallel flat pieces of glass. These fringes were not apparent on preview images in the FireCapture camera operating software, so I wasn't aware of their presence until I later processed the images.

Sometimes interference fringes can be removed by slightly tilting the camera. This has worked for me in the past, but now the effect will not be immediately visible on my laptop. I'll have to process future images to see if tilting actually removes the fringes.

I calculated an image scale of 0.53 arc seconds per pixel for the 3X Barlow lens with my Lunt 100 mm telescope. The 3X angular field of view is 17.2 arc minutes by 10.8 arc minutes. The image above was reduced from full size to two thirds size because this seemed to give the most pleasing resolution.

Next, I tried a 5X Barlow lens. The following picture is a 15-image mosaic of the Sun's equatorial region reduced to half size from the original.

Once again, annoying vertical interference fringes ruin the picture, but the details and resolution are very nice. (Click on the image for a larger view.) 

I calculated an image scale of 0.32 arc seconds per pixel for the 5X Barlow lens. The 5X angular field of view is 10.4 arc minutes by 6.52 arc minutes.

Sunspot groups 2345 on the left and 2339 on the right show in good detail in the next 4-image mosaic made with the 5X Barlow. The image has been reduced to two thirds size. Click on it for best detail.

Finally, check out this detailed 5X Barlow single image of sunspot group 2339 which has been reduced to two thirds size. Interference fringes seem almost completely absent. White flares are present near center. (Click on the image for the best view.)

The previous image is among the best I've ever achieved for detail. I might use the 3X Barlow for my workhorse lens in the future if I can remove the interference fringes. 

Video file sizes from the ZWO camera are much bigger than files I get from my DMK41 camera. A 400-frame AVI ZWO video produces a 0.92GB file! A 1,000-frame AVI video produces a 2.3GB file! My image processing software, Registax6, would not handle the large 2.3GB file, so I will be limited to fewer video frames if I continue with Registax6. This really isn't much of a problem because 400-frame or 500-frame videos produce very nice images.

New ZWO ASI174MM Camera

Testing New Equipment

It was warm, even borderline hot on May 5th, but I withstood the heat because I was anxious to try my new ZWO ASI174MM monochrome camera. The camera is the small red cylindrical device attached at prime focus on the back of my telescope in the images below.

The new camera is theoretically superior to my old DMK41 camera in three ways. First, the chip size is significantly larger. The new ZWO camera's chip has 5.86 micron square pixels in a 1936 by 1216 array compared to the DMK's 4.65 micron square pixels in a 1280 by 960 array. This means the ZWO chip is 11.34 mm by 7.13 mm compared to the 5.95 mm by 4.46 mm DMK chip. The ZWO chip has three times the imaging area of the DMK chip. 

Second, the new camera has the ability to record a greater brightness range. The DMK camera is an 8 bit device capable of 256 brightness levels. The ZWO camera can function as an 8 bit, 10 bit, or 12 bit device. In 10 bit mode it records 1024 brightness levels. In 12 bit mode it records 4096 brightness levels. 

Third, the new camera has faster download speed. The ZWO camera uses USB 3.0 communication while the DMK uses USB 2.0. This speed increase might be the most useful improvement for my personal observing requirements. For reasons I've never been able to figure out the DMK camera usually freezes temporarily during capture downloads. For example, the capture of a 400-frame video takes 27 seconds when the DMK is behaving properly. But proper behavior seems to come and go at random. More times than not the DMK takes 54 seconds or more to download a 400-frame video. In 8 bit mode the ZWO took only 6.67 seconds to download a 400-frame video, and in 12 bit mode it took only 12.76 seconds! This is a very pleasant change from the DMK! Unfortunately, I still experience temporary download freezes with the new camera. The freezing isn't constant. It comes and goes as if the laptop is handling competing tasks. I just don't know what these competing tasks could be. It would be wonderful to discover some computer setting that would eliminate this problem.

The new camera's large chip size means a full disc solar image can be captured at prime focus with my Lunt 100 mm telescope. The picture below has not been cropped. It shows exactly how the full disc fits on the imaging surface. The field of view is 51.7 arc minutes wide by 32.5 arc minutes high. Resolution here is not pleasing, but the full disc image could be obtained from one video without making a mosaic. Unfortunately, illumination is slightly uneven from left to right, perhaps because the disc was not precisely centered horizontally on the chip.

I usually use a 2X Barlow lens with the DMK camera because it gives a more pleasing resolution than prime focus. So I used a 2X Barlow with the new ZWO camera to see how images would look. The following 3-image mosaic, made with the 2X Barlow, shows large sunspot group 2335 below center, 2338 near center, and newly visible sunspot 2339 near the limb in the upper left. (Click on the image for a larger view.) There is a slight uneven illumination, particularly in the bottom right corner. The resolution here is much better than the prime focus image, but still not quite as nice as what I get with the DMK plus 2X Barlow.

The next picture, a 2-image mosaic, shows filaments stretching across the southern hemisphere. It was processed to show disc features, so prominences are not visible.

The next image is the same mosaic as the previous one, but disc features are overexposed in order to reveal the filament to prominence transition on the right limb. (Click on the image to see the transition in more detail.)

The best picture obtained with the new camera so far is the following single image showing prominences, sunspot group 2335, and nearby filaments.

Resolution is not quite as good as I'd like. Click on the image to view it full size, then compare the area around sunspot 2335 with the resolution of the next image. You will probably agree that the DMK41 image below is a little nicer.

I may need to use a 3X Barlow with the ZWO camera in order to achieve the resolution I'm looking for. It will take a while to try different exposure times, Barlow lenses, and settings in the FireCapture software that runs the camera. But I'm encouraged by these initial results and I'm looking forward to using the new camera again soon.

New Hinode Solar Guider

Sun on the Dashed Line

Nearly perfect conditions prevailed before noon on May 4th: clear sky,  temperature in the upper 60's to low 70's, periods of good seeing, and almost no breeze.

Large sunspot group 2335 was positioned left of center above a horizontal dashed filament line stretching across the Sun's southern hemisphere. The following 19-image mosaic, made with a 2X Barlow lens, shows features above the dashed line. (Click on the image for a larger view.)

Sunspot group 2335 included several small umbras amid complex structure with a nice curving filament to its right.

Today I used my new Hinode Solar Guider for the first time. The guider is the white rectangular device mounted near the top of the telescope tube in the image below. It is designed to correct the telescope mount if the mount is not tracking the Sun accurately.

In the next image below you see two small circular apertures in front of the guider facing toward the Sun. Each circle contains a narrow slit. The two slits are mutually perpendicular. Sunlight enters through these mutually perpendicular slits and the device thereby detects the position of the solar disc. The guider repeatedly detects the position of the disc, notes any changes in position, and sends signals to the telescope mount to keep the Sun centered. The small glowing yellow circle at the rear of the guider shows the solar finder in action. When the bright yellow image of the Sun is approximately centered within the darker yellow target area, the guider is correctly pointing at the Sun.

How well does the guider work? I've used it twice so far. Each time the calibration procedure was not immediately successful. In the first session I had to run the calibration procedure twice before the device would guide correctly. In the second session I had to calibrate four times before success. This wasn't a major difficulty, just a puzzling inconvenience. In the first session it didn't calibrate with a 17-second test run, but did calibrate with a 25-second test run. In the second session it didn't calibrate with a 25-second test run, but did calibrate with a 17-second test run. Maybe I'll get more efficient with more practice.

Once the guider was properly calibrated I found the most aggressive correction setting to be unsatisfactory. At this setting the image jerked around more rapidly than it would due to bad seeing. The medium aggression setting seemed to be most effective. I monitored a live laptop image from a DMK41 camera with a 2X Barlow lens in order to observe the guiding. On the medium aggressive guide setting guiding was smooth and helpful, though not absolutely perfect. During my first session the guider kept image drift within a 40 by 40 pixel box (equivalent to a 20 by 20 arc second box) for 15 minutes. During my second session the guider did better, keeping image drift within a 16 by 16 pixel box (equivalent to a 7.8 by 7.8 arc second box) for 17 minutes. So the image was not absolutely rock steady, but the guider did eliminate accumulating unidirectional drift. It successfully kept a full disc prime focus image of the Sun centered in the field of view of my new ZWO ASI174 camera for many minutes. The guider, of course, cannot eliminate the random swirling dance caused by ordinary atmospheric turbulence.

I intend to use the guider when I make time lapse movies. It probably doesn't make sense to use it for normal solar imaging since my cameras take only a few seconds to capture enough video for individual still images. 

Enormous Prominence

Prominences Galore!

On April 23rd a dramatic giant prominence rotated into view on the Sun's eastern limb. Although only two days had passed since my last observing session, I went out again to image the behemoth. It was another windy spring day with gaps between clouds.

The following 5-image mosaic, made with a 2X Barlow lens, shows a huge portion of solar limb occupied by the monumental prominence.

A magnified view shows the mighty prominence itself.

Prominences also graced the Sun's western limb as shown in the following 5-image mosaic. (Click on the images for a larger view.)

Disc features are overexposed in each of the previous images in order to reveal the dim prominences.

I spent a good deal of time this day trying out a new camera and wondering why it wasn't functioning as expected. It finally dawned on me that I inadvertently bought a color camera instead of a monochrome camera! I thought I had purchased a monochrome camera and expected monochrome camera behavior, so I was constantly puzzled by the results I was getting. It makes no sense to use a color camera for hydrogen-alpha solar imaging because solar filters pass only very, very monochromatic red light. I hope to have the correct monochrome camera soon. I'll write about the camera's performance in a future post.

Dodging Clouds

Windy Day

It tends to be windy at my observing site in spring. On April 21st it was 67 breezy degrees with fair weather clouds drifting by. I managed to image a few sunspots and one interesting prominence through gaps in the clouds.

I captured only ten videos this day because I explored possible solutions to communication problems between my camera, telescope mount, and laptop. Hoping to reduce USB traffic, I removed the USB receiver for my wireless trackball and controlled the laptop only through its touchpad. Next, I connected my DMK41 camera directly to its own dedicated USB port instead of feeding camera data through the same cable used by the telescope mount. With this setup the camera and the telescope mount each have their own dedicated USB ports. I was pleased to find the camera downloading captured video at full speed. There was also no communication break between the mount and computer. I wondered if this setup would work during longer observing sessions. (It didn't. Camera downloads were frequently slow in subsequent sessions.)

The following 3-image mosaic, made with a 2X Barlow lens, shows prominent features in the Sun's western hemisphere on April 21st. Double umbra sunspot 2325 is at lower left. Sunspot group 2324 is above 2325. Sunspot 2326 is on the right among filaments. A loop prominence is faintly seen lower to the right of 2326.

Nice prominences along the western limb can be seen in the next 2-image mosaic where disc features are overexposed in order to reveal the dimmer prominences, particularly the bottom one.

A magnified view of the lower prominence, associated with departing sunspot 2322, reveals a loop structure.

Sunspot 2327 sat in isolation near center on the solar disc with a small magnetic dipole to its right.

It would have been nice to make a time lapse movie of the loop prominence, but there were too many clouds drifting past the Sun, and it was too late in the afternoon.