SD14 in Astrophotography

Discussion in 'Nature & Wildlife' started by Steaphany, Nov 24, 2008.

  1. Steaphany

    Steaphany Well-Known Member

    Joined:
    Oct 12, 2008
    Messages:
    240
    Likes Received:
    0
    One of the subjects of photography that I enjoy is Astronomy and astrophotography. Prior to purchasing my SD14, I was well aware of the Nikon and Canon dSLRs being the predominant choice, but the capabilities of the Foveon imager, regardless of the reviewers complaints about high ISO noise levels, would still provide the functionality I was looking for. My decision was also based on wanting a general purpose camera flexible enough to handle all my photography needs.

    I know that imager noise in astrophotography can easily be handled by the multitude of image stacking software packages which are commonly used to clean up and combine multiple frames. A common image medium used by amateur astronomers is adapting a low resolution webcam to a telescope and recording and stacking from AVI video files. I was sure a SD14 would out perform any webcam.

    This is the results of my first astrophotography test. I used my Sigma 28 mm f1.8 EX DG Aspherical Macro which provides a 40.57° x 27.69° field of view. My SD14 was on my Velbon DF50 tripod which offers no means to track the sky. The used a manual focus and manual exposure was set to 16 Seconds with an f1.8. I set the ISO sensitivity to 100, to keep the noise down and to see just how well the SD14 would perform. I chose 16 Second exposure since it was the trade off point between maximum light collection while keeping the stars looking like point sources and not streaks across the sky. The White Balance was set to Overcast since my earlier attempts to photo a Moon rise turned out very inconsistent results with auto. To trigger the shutter, I used my RF remote control that I purchased through a vendor on ebay (I know I need to update the thread on Gear-Talk/Accessories).

    I chose to aim the SD14 to Cassiopeia since it was on the meridian and nicely high in the sky, well away from the effects of the atmosphere near the horizon. Cassiopeia also has the Milky Way passing through it along with several star clusters and faint nebulae. It was also a small enough constellation to completely fit in the field of view.

    I never bothered to let my eyes adapt completely to the dark, so I found it difficult to focus the 28mm lens on the bright stars of Cassiopeia since they were too faint. I knew they were in the center of the field of view, but not clear enough to properly focus. I knew that the 28mm lens, at f1.8, had a hyperfocal distance of 24.23 meters, so I temporarily aimed the SD14 towards the lights of a cellphone tower roughly 3 miles away, which would have the stars, effectively at infinity, in perfect focus.

    This is what the SD14 recorded:

    SDIM0245.jpg

    This is a decent star field, low noise, with good focus. The bright stars of Cassiopeia are easily visible as is some of the color variations between the stars.

    I then continued to shoot a total of 24 light frames while leaving the tripod and SD14 exactly in the same position. I also placed the lens cap over the lens and shot 2 dark frames. This took me roughly 15 minutes from start to finish. Keep in mind that the tripod has no means to track the sky as time passed by. Each frame had the stars in a slightly different position when compared to the other frames. The image stacking software combines the light frames, subtracts the effects of the imaging system recorded in the dark frames, and aligns each field to produce the final image.

    I used IrfanView, You must be logged in to see this link., to bulk convert the 26 X3F files into PNG format and I used DeepSkyStacker, You must be logged in to see this link., to align, stack, and adjust the final image. An advantage of DeepSkyStacker is that it is able to identify a star from the surrounding sky making alignment automatic.:

    Cassiopeia_DSSs.jpg

    I was really surprised to see just how many stars where captured. DeepSkyStacker discarded several low quality frames and built this up from just 17 of the original 24 light frames. Remember this was at ISO 100 and the total exposure time was only 4 Minutes and 45 Seconds.

    I did trying using RegiStax, You must be logged in to see this link., but found it taking a lot longer and the alignment algorithm brought the stars tagged by multi-point references into proper alignment but the surrounding faint stars were dragged into arcs facing the nearest reference point, caused by segmentation of the image. By contrast, DeepSkyStacker uses an alignment algorithm which positions and rotates each frame as a whole. Both have their place as I found Registax will accurately align and combine terrestrial and non-star field photos.

    To help those here who are not astronomically inclined, I created a composite of the DeepSkyStacker image with a screen capture from my Sky map program. The alignment is not 100% but close enough for each of Cassiopeia's features to be identified:

    Mapped.jpg
     
  2. foveonfan

    foveonfan Well-Known Member

    Joined:
    May 25, 2007
    Messages:
    322
    Likes Received:
    0

    Hi, Steaphaney!

    That is a great post and of immense value to me.

    I am possibly missing something though (pardon my ignorance in advance), but if all frames are exposed equally, why were the results of several unsuccessful?

    Keep up the great work!

    Sincere regards, Jim R
     
  3. Guest .

    Guest . Banned

    Joined:
    Feb 29, 2008
    Messages:
    528
    Likes Received:
    0
    Hi Steaphany,

    I am deeply impressed, too ... although absolutely ignorant in this issue.

    I once did experiments with Image Stacker.

    You must be logged in to see this link.

    It is a very nice tool for artificial long time exposures. Now, that the SD14's BULB-mode became extended from 30 seconds up to 120 seconds .... this tool is hardly necessary any more. ..... practically, the former 30s were absolutely enough for my purposes.

    However,

    thanks a lot for this very interesting post. :z02_respekt:

    See you with nice stars

    Klaus
     
  4. Steaphany

    Steaphany Well-Known Member

    Joined:
    Oct 12, 2008
    Messages:
    240
    Likes Received:
    0
    Both DeepSkyStacker and RegiStax employ a means to evaluate the quality of each frame. As the frames are processed, they are assigned a Score value which is based upon characteristics of image content. For example, if camera vibration causes the stars to be in a non-round shape, the quality is lower and the score is not as high as a frame with perfectly round stars. By default with both packages, any image with a quality below 80% is discarded from inclusion in the stacking process.

    From the DeepSkyStacker's FAQ:

    What is the score, and what is its meaning?
    The score is a measure of the picture quality. To put it simply, the higher the score, the more round and not too big stars were found.

    Is the score a measure of the absolute quality of a picture?
    No. The score is a relative measure that is only used to sort the pictures of roughly the same area in order to keep only the best pictures for stacking.
    From the RegiStax pdf manual:

    At the same time as alignment is going on, RegiStax is also estimating the quality of each frame, and sorting the frame set into order of quality. There are five different methods by which this can be done, and you can choose the method used before clicking the Align button. Of the five however, the best results are usually achieved with the "Gradient" method. It is also the simplest to use as no additional settings are required. You can choose the method used by selecting it from the drop-down list in the Quality Estimator section of the Alignment options tab.
     
  5. foveonfan

    foveonfan Well-Known Member

    Joined:
    May 25, 2007
    Messages:
    322
    Likes Received:
    0
    Thanks, Steaphaney.

    Sincere regards, JR
     
  6. tc95

    tc95 Well-Known Member

    Joined:
    Sep 21, 2008
    Messages:
    684
    Likes Received:
    0
    Thank you...

    Steaphany,

    Thank you....I have been slowly aqiring several pieces to be able to do night time shots....I will have to play a little now...

    Tony C.
     
  7. Brazile

    Brazile New Member

    Joined:
    Nov 26, 2008
    Messages:
    4
    Likes Received:
    0
    Very interesting!

    Thanks very much for the thorough description of your work. It was very interesting to me, because I have been thinking about trying this myself.

    I have a couple questions, if you don't mind:

    1. Why the dark frames? I was under the impression that the SD14 automatically subtracts a dark frame from long exposures, is that incorrect, or insufficient?

    2. Have you yet looked into shooting with a telescope? What method of connection are you considering?

    Robert
     
  8. Steaphany

    Steaphany Well-Known Member

    Joined:
    Oct 12, 2008
    Messages:
    240
    Likes Received:
    0
    Robert,

    To your first question, I have to say that I have not experimented to see the distinction that the additional dark frames would make. At the time when I did the Cassiopeia photo, I saw how my SD14 would take the exposure, close the shutter, and sit for a period before writing the data to the CF card. It was only later when I learned what it was doing during the closed shutter wait.

    In some ways, I find it problematic that I can not turn off the SD14's internal dark frames. I enjoy photographing lightning and the internal dark frame means that every exposure carries a corresponding period where the storm activity is missed. I also want to shoot time lapse star videos and static camera star trail images, to achieve this with out having the stars blinking or the star trails becoming a series of dotted lines, the internal dark frames mean I need to use short exposures. I also question how raw the X3F files are when the SD14 has already performed it's own dark frame compensations. I'm aware of most photographers not being experienced with the common tools to process astronomical images, but for those that are, I would prefer that the SD14 had an option where the photographer can choose to shoot included dark frames, to shoot and store dark frames separately, and control how they are used with the light frames. Leaving this flexibility to the photographer may also increase the capability for the SD14's high ISO's which are inherently noisy.

    As the evenings get warmer, I plan on answering some of these questions.

    To your second question, Yes and I've had information in three threads on the subject using a SD14 with a Telescope:

    You must be logged in to see this link.

    You must be logged in to see this link.

    You must be logged in to see this link.

    I hope this helps you out in your efforts.
     
  9. Steaphany

    Steaphany Well-Known Member

    Joined:
    Oct 12, 2008
    Messages:
    240
    Likes Received:
    0
    A valid URL to buy a T-Mount adapter for Sigma cameras can be found on this Gear > Accessories thread:

    You must be logged in to see this link.
     
  10. laurence2

    laurence2 Well-Known Member

    Joined:
    Jan 15, 2009
    Messages:
    129
    Likes Received:
    0
    Amazing and fascinating! Although I probably won't find myself delving into astrophotography, I am an itinerant "armchair astronomer", and the physics and astronomy books are what I usually buy or check-out from the libraries.
     
  11. Steaphany

    Steaphany Well-Known Member

    Joined:
    Oct 12, 2008
    Messages:
    240
    Likes Received:
    0
    I found Gold !!!

    Curious about the photographer's work behind the You must be logged in to see this link. on NASA's Astronomy Photo Of the Day web site, I checked out his web site, the link is on the APOD page below the image and I'll repeat it here:

    You must be logged in to see this link.

    On Drew Sullivan's Technical page, I became curious about the software that he used to make the manipulations correcting for internal optical reflections for his Pleiades image. I've noted problems with my stacked star field images shot through my SD14, where I get an imager induced artifact, the bright triangular blooming phenomenon associated with the brighter stars, and I'm sure his technique would eliminate this.

    I emailed Drew and he just sent me a reply, pointing me to a software package published by William-Bell, Inc.

    You must be logged in to see this link.

    As described by Dennis Di Cicco, Senior Editor of Sky and Telescope magazine, "It's the finest book to date covering the entire gamut of digital astrophotography. It should be mandatory reading for those shooting digital pictures of the Moon, planets, or deep-sky objects regardless of what camera or software they use."

    William-Bell, Inc. specializes in books on astronomy and I have their text "Astronomical Algorithms" by Jean Meeus.

    Two aspects that really impressed me was the list of AIP4WIN features spanning 5 pages:

    You must be logged in to see this link.

    and the list of supported Camera formats spanning 3 pages:

    You must be logged in to see this link.

    which includes all three Sigma Cameras, the SD9, SD10, and SD14 !
     
  12. sendhil76

    sendhil76 Active Member

    Joined:
    Dec 8, 2008
    Messages:
    29
    Likes Received:
    0
    Very informational thread! I have read quite a few image processing texts...will look this book up also with its focus on Astronomy and photography related to it...something that i dont understand at the moment :) though it fascinates me a lot as i look up to the skies whenever i wish to experience magic for free!

    Thanks for sharing!
     
  13. Steaphany

    Steaphany Well-Known Member

    Joined:
    Oct 12, 2008
    Messages:
    240
    Likes Received:
    0
    The Gold is a bit lacking

    After downloading and installing Version 2.3.0 of AIP4WIN and receiving my copy of the book "Handbook of Astronomical Image Processing" which contains the serial number necessary to register the software license, I was disappointed to learn that, despite all three Sigma cameras being listed as being supported, I just could not load a X3F file into AIP4WIN.

    I sent an email to the support address and was contacted by both authors, Richard Berry and James Burnell. The bottom line is a third party provided their RAW camera support and until I provided them with a sample X3F file, they never had one to work with prior. I have since sent them my collection of 24 Cassiopeia X3F files as well as 16 bit / channel TIF files that I created from the X3F files with SPP. Since I only have a SD14, I would appreciate anyone with a SD9 or SD10 interested in aiding this effort by providing files to ensure support of those cameras as well.

    -=-=-

    I have been corresponding with Drew Sullivan in regard to flaws in the stacked Cassiopeia image posted earlier in this thread and I have made progress in the area of Sigma Astrophotography.

    First, if your astronomical image manipulation software can not handle the X3F format, use SPP to create 16 bit / channel TIF files. This simple change from my original process flow made all the difference when stacking and SPP gleaned some really nice results on individual frames.

    Here is what SPP did with a single frame:

    SDIM0258_s.jpg

    There are still flaws in this image, but they are limited to only the 5 brightest stars exceeding 98% of the image dynamic range. This means the exposure through my Sigma 28 mm f1.8 EX DG Aspherical Macro of 16 Seconds, f1.8, ISO 100, with a white balance of 6500K (Overcast), was marginally too long and the photosites receiving light from these stars were close to being saturated. For this image I set the final image white balance to the star πCAS which has a magnitude of 4.95 and was nicely exposed. I did boost the saturation to bring the stellar spectral colors into better view. Later, I reviewed the faintest individually discernible stars and found the limiting magnitude, the point beyond which a star blends into the background, at 10.0. For a point of reference, the limiting magnitude for unaided human vision, as the best of conditions is 6.

    Here is an image with these points identified:

    SDIM0258_Labeled_Text_s.jpg

    Remember, this is the result of an individual 16 second exposure which was processed in SPP
     
  14. jasonh

    jasonh Well-Known Member

    Joined:
    Mar 24, 2009
    Messages:
    55
    Likes Received:
    0
    Hi Steaphany

    I'm curious if you've done any more astrophotography with your SD14 lately.

    I've only had my SD14 for a little while, but after seeing yours and some other examples, this seems like it would be a fun thing to do.

    I just made a DIY cable release last night so I'd be able to do this even easier. I've also seen some designs for making a simple barn-door tracker that would be good enough for stacking images.

    Unfortunately I'm using a Mac, and I have yet to find any of these astrophoto stacking programs made for Mac :(

    One of these nights soon when it's clear I'm going to have to go out and see what I can get with just a tripod and an 18-200 lens.
     
  15. Steaphany

    Steaphany Well-Known Member

    Joined:
    Oct 12, 2008
    Messages:
    240
    Likes Received:
    0
    Jason,

    My latest SD14 astrophoto was an experiment that I did a couple weeks ago.

    I tried to photograph Orion and the surrounding area with the internal IR blocking filter removed to see if I could get better imaging of the various nebula. My results were not what I expected and it turns out that the sensitivity of the Foveon imager caused even short exposures to be saturated by the brightness of the stars. Instead of being small points of light, stars like Betelgeuse, a variable red giant varying from 0.4 mag to 1.3 mag, just ended up being, comparatively, a large round red blob.

    I have searched and found several US patents which Foveon received on their technology. It turns out that the channel sensitivity curves have quite a bit of over lap and I'm suspecting that much of the compensation must be occurring in the SD14. When the IR blocking filter is removed, the resulting image is almost devoid of any Green, Blue is weak, and Red dominates. I will have to plan a series of experiments to better characterize how the SD14 performs for infrared and broad spectrum photography and astrophotography.

    The weather or bright Moon has been in the way of further attempts.

    I have two recommendations for your astro image processing. First is a shareware image stacker which runs on the MAC:

    You must be logged in to see this link.

    Since all my computer facilities are either Sun Microsystems UNIX/Solaris or Intel Windows XP based, I have no hands on experience with Keith's Image Stacker.

    The second is a book with DVD containing the tutorial images from Willmann-Bell:

    You must be logged in to see this link.

    Since Photoshop runs on both the PC and MAC, this text will suite both environments. I have this book in my astrophoto collection.

    One thing you need to keep in mind is having your shutter open long enough to collect a good exposure while also being short enough for a SD14 on a static tripod to still show stars as points and not streaks. It is also amazing how quickly the light from a star can saturate the Foveon photosites at even ISO 100. I recommend that you choose a static White Balance, I normally use Overcast which corresponds to 6500K.

    I hope this helps and I'll be looking forward to seeing what you can accomplish.
     
  16. jasonh

    jasonh Well-Known Member

    Joined:
    Mar 24, 2009
    Messages:
    55
    Likes Received:
    0
    Steaphany,

    I'm glad you threw in that little tidbit about the effects of IR on the other channels of the Foveon chip. I had actually noticed this effect while trying to do some visible+ir shots soon after getting the SD14. The only thing the green channel contained was the very brightest of highlights, and I wasn't sure why. Looks like you cleared it up for me :)

    Thanks for your suggestions. I had thought of just using Photoshop, but I figured it would get very tedious with more than just a few exposures.

    Keith's image stacker looks like it'll probably be my best bet.

    I found a few other programs as well, some look like they have potential.

    Hopefully I'll be able to get out there soon. Been having weird snowy weather lately, so it's not quite prime conditions right now.

    Thanks
     
  17. jasonh

    jasonh Well-Known Member

    Joined:
    Mar 24, 2009
    Messages:
    55
    Likes Received:
    0
    Also, I had another thought about the IR thing....how short were your exposures? I'm curious if you could stop the aperture down and shorten the shutter time enough to get a decent exposure(s)...then stack those and use them purely for Luminosity, and also take some shots with the IR filter on for RGB...just a thought.
     
  18. Steaphany

    Steaphany Well-Known Member

    Joined:
    Oct 12, 2008
    Messages:
    240
    Likes Received:
    0
    I used a range of shutter speeds.

    All had the SD14 set to Manual Exposure and ISO 100. I used my Sigma 28mm DG EX Macro lens set to manual focus, and the internal IR blocking filter was removed.

    The best result for the evening was with a 10 Second Exposure through a You must be logged in to see this link. that I have to easily switch between Infrared/Broad Spectrum and Visible light shooting with out having to remove the lens from the SD14 in the field. (Not something I want to do when there is a risk of getting dust on the imager)

    SDIM1426s_10S.jpg

    As noted here, the brightest stars severely saturated the imager:

    SDIM1426s_10S_Labeled.jpg

    As you can see, it can be difficult to balance such an extreme range in intensities.

    Here is another image where the only change was a 1 Second Shutter:

    SDIM1430s_1S.jpg

    My attempt at Visible Red + Infrared exposure was a short 3.1 Seconds through a 25A filter:

    SDIM1391s.jpg

    I was intending on shooting more and getting a better range of exposures, but clouds were moving in ruining the view.
     
  19. jasonh

    jasonh Well-Known Member

    Joined:
    Mar 24, 2009
    Messages:
    55
    Likes Received:
    0
    Wow, that is amazing just how much IR is collected from those bigger/closer stars.

    It would be interesting if there were an easy way to use a few very long IR exposures to pick up the faint stars and combine that with several visible frames to take care of the bright ones and the colors.

    Since I don't have the filters for the front of the lens (boy are 72mm filters expensive) I think I'll just not bother with removing the internal filter any time soon :)

    Hopefully I get a clear night soon to try some of this myself.
     
  20. Steaphany

    Steaphany Well-Known Member

    Joined:
    Oct 12, 2008
    Messages:
    240
    Likes Received:
    0
    Jason,

    Check out the thread You must be logged in to see this link., I just checked and as I write this, the Moon is presently 53% illuminated. An easy subject to get some experience with, while also being quite challenging at the same time.

    The best time to photograph the Moon is when the terminator, the boundary between the Sun illuminated and dark sides meet, is visible. The mountain tops, crater rims, valleys, and crater floors provide a nice sense of 3D and spherical dimension to the Moon.

    During the full and near full phase of the Moon, you'll loose these depth cues and the Moon will look flat.

    A good web site to see what's going on in the sky is You must be logged in to see this link. and a good free open source computer planetarium is You must be logged in to see this link.. It runs on Windows, MAC, and LINUX showing a realistic sky in 3D, just like what you see with the naked eye, binoculars or a telescope which is perfect for planning astrophoto sessions and for identification of objects recorded in your photos.

    I do recommend that you use a hand held GPS receiver. Knowing Latitude, Longitude, and Elevation are essential to set your location on the Heavens Above web site and in Stellarium. Plus, the time displayed on a GPS receiver comes from the atomic clocks carried on the GPS satellites which can be very useful when you need to know the exact time for your observations and photography.
     

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice