Try increasing gamma if dark sections aren't distinguished

Try increasing gamma if dark sections aren't distinguished

Sunday, December 10, 2017

2017 eclipse color composite

since the human visual system is more accurate for luminance (black and white) than color, one can combine a low resolution color image with a high resolution black and white image without any perceived loss of detail.  this phenomenon is often exploited in image printing, processing, recording, display etc.  wikipedia link.  

applying this concept to eclipse imaging, i had hoped to combine a lower resolution color digital camera image (DSLR) with a sharp black and white image from the telescopic video.  as mentioned previously, i blew the focus on my DSLR shot.  fortunately, my eclipse neighbor, James Pothier, sent me his DSLR image from the same location and time.  Jim and wife Ellen hail from Massachusetts.  7 for 7 eclipse viewers, they set up right next to me at the Residence inn, Casper, Wyoming.  I combined the color from his DSLR shot with my black and white image, aligning on the colored prominences and voila:

here's the color applied to my previous composite combining multiple exposures:
click for full size

for reference here's my DSLR shot with focus blown (many thanks to the Pothiers):

Image details:
-totality wider field:
Nikon D-60, AF-S DX NIKKOR 55-300mm f/4.5-5.6G ED VR, f/5.6 300 mm 1/20" iso 400
-Pothier image
Nikon D300, f/6.3 ISO 200 1/640 -1step exposure bias 380 mm.
-telescopic view:
DMK 51 web cam, Takahashi FS-60C, 60 mm aperture at f/4.2 with a reducer, 252mm. Fotga IR/UV cut filter.  The full field of view is approximately 96x72 arc minutes with a resolution of 3.6"/px.  video capture at 12 fps, aligned in autostakkert, wavelets in registax.
Casper Wyoming
42,50.9694N 106,15.5688W

Sunday, December 3, 2017

2017 eclipse video and imaging techniques 2

so i had 2 minutes and twenty seconds to video the eclipse at various exposures (determining level on the fly), take DSLR shots for color, and of course a few cell phone shots, not to mention actually look at the eclipse, sky, stars, horizon and generally woop it up
while there were some flubs (like blowing focus for the color DSLR images), it was generally a successful outing.

before the video, here are a few more of the interesting still frames

dark ring, with some lunar surface detail visible on the left edge (click on image for full size):

ending diamond ring with multiple blooms, prominences still visible (click on image for full size):

filter back on for crescent shot as the brilliant sunlight returns:

Here's the video of totality with time variably compressed from 3x for starting and ending diamond rings to 12x for longer images of the dark ring, and outer corona. what you'll see:
1. filtered left crescent of the sun
2. filter off 20 seconds before totality, brilliant light overwhelms the image and illuminates passing clouds :(
3. the light fades to the diamond ring, the vertical lines are due to the bright light overwhelming the digital camera, spilling signal into adjacent pixels
4. totality, the dark ring with the moon moving slowly from right to left, progressively covering the chromosphere on the left side while exposing more dramatic prominences on the right, arcing off the solar edge.
5. change in exposure to get the outer corona, then back and forth to the dark ring
6. final shot of chromosphere on right side with prominences
7. right sided diamond ring with more blooms
8. filter back on for right side crescent 30 seconds after totality ends

be sure to adjust the quality to high (gear lower right) and increase the image size (lower right)

in addition to 2 full dry runs, i ran at least 10 practice sessions of 2 minutes and 20 seconds of totality with a carefully scripted list of what i was going to do during totality, including an observation plan.  i rehearsed all the steps including pretending to look at the eclipsed sun with binoculars etc.  worked out minor issues (taking off all hats and lens caps before totality), to complex issues (a software glitch that would have ruined the capture--can't change exposure in firecapture with the autocapture window open).  i even practiced an imaging fail, giving myself 60 seconds to sort out a software issue, then forgetting about it and proceeding with my viewing plan.  An app called Solar Eclipse Timer was very helpful for the practice runs and imaging session.  

i elected to watch the first diamond ring phase on my computer monitor (so i didn't blow out my vision), quickly adjust the exposure at the start of totality, then watch the first half of totality.  at max eclipse (halfway though totality), i went back to imaging with a tightly scheduled plan, then viewed the final diamond ring.

i wound up with 4 pages of hand written notes.  here's the page for totality, choreographing the event.
note the shadow reflecting completion of the list
Max is the halfway point

Here's the full video in real time with minimal editing (calibrated and removed shaking frames during filter transition) if anyone's interested.

calibration was a problem, i finally figured out that PIPP (software program) only allows you to calibrate video with another video, rather than a single frame.  fortunately, i'd saved my calibration video runs.

-video imaging seemed to give a sharper image than a still shot, as any plantery or solar imager will tell you.
-did not detect any motion in the fine structure of the corona at this image scale thru clouds
-disadvantages of video: lower dynamic range, consider moderately slow change of exposure length during imaging to compensate for this (or greater dynamic range)
-CCD chip bloomed on diamond ring phase, CMOS preferable
-the larger DSLR field of view seemed optimal for the corona:
4.5 decrees by 3 degrees or 4.180 arcsec/px
-focused my DSLR on clouds rather than on sun with solar filter which did not work.  should pre-focus with a dedicated solar filter rather than clouds.

Image details:
DMK 51 web cam, Takahashi FS-60C, 60 mm aperture at f/4.2 with a reducer.  Baader solar film, Fotga IR/UV cut filter.  The full field of view is approximately 96x72 arc minutes with a resolution of 3.6"/px.  20 second video capture at 12 fps, aligned in autostakkert, wavelets in registax.
Casper Wyoming
42,50.9694N 106,15.5688W

Monday, November 27, 2017

2017 eclipse warm up images and techniques 1

almost all publications on eclipse imaging say something like: "for your first total eclipse forget about taking photos, just watch".  sound advice, but i've been honing my solar imaging techniques for several years...

i posed the dilemma to my astroimaging group
veteran eclipse chaser Dave Kodama got a puzzled look on his face and simply said, "if there's no picture, it didn't happen".

though far from a proponent of organization for organization's sake, i realized this was going to take some  preparation. 

There are many accounts of photographers missing the total eclipse by forgetting to take solar filters off, or forgetting to replace the filter, leading to fried equipment.  Mindful of this i ran two complete dry runs of the imaging session from my imaging location at the same time of day. 

Here's an animation of one of the sessions.  note the slight rotation of the sunspots across the solar surface even over the course of a few hours:

Sun 8/18/2017
Here's the image from the day before, excitement growing as a second sunspot grouping rotates into view:
Sun 8/20/17

ironically, my best full disk white light image of the sun was taken on 8/21/17, only to be overshadowed by eclipse images :)
Sun 8/21/2017

note two separate sunspot groupings and well defined surface granulation.


in an excellent ebook on photographing the eclipse, Alan Dyer gave a comprehensive list of imaging techniques ordered by level of complexity.  my approach (webcam, laptop, close up telescope, tracking mount) seemed to be over the top, making preparation a must.  

Starting in January i began working out an imaging rig, field of view, and processing techniques for the sun and moon in white light. 

next i tested filters, finally concluding that green + an IR/UV block gave the best detail on sunspots and surface granulation.  however, at the 11th hour i decided to dispense with the green filter in order to minimize the fiddle factor, as imaging totality with a green filter would block the red light of the solar prominences and switching would complicate the process. 

Then to the mount.  after numerous tests with a simple manual alt-azm mount were moderately successful, Alan Smallbone gave a mount review, and talked some sense into me. so i switched to a simple polar aligned single axis tracking mount.  streamlining the imaging process.  
single axis tracking mount, mylar laptop cover bottom
focus during solar imaging has always been problematic for me.  the problem is that bright sunlight makes it almost impossible to view the image on a reflective laptop screen.  i'd typically bake with my head and laptop under a blanket while trying to focus, reaching blindly for the focus knob.  2 solutions:
-a laptop "tent" or lapdome complete with a shade to block light reflecting up from the ground (also wear a dark shirt)
Lapdome, note shade between keyboard and screen blocking reflected light from the ground, secondary mylar cover in background
-a mylar drop cloth.  most mylar "space blankets" are too thin to block all the light.  i found a thicker version that did the trick, light weight, easy to pack.  placing the blanket over my head and laptop blocked out stray light, and reflected radiant energy keeping things cooler.  

tuned laptop: new battery so it would last thru the eclipse and "custom water cooling" to keep it from overheating during a long imaging session in hot weather.
Laptop intake fan, bottom of laptop
filter box elevating laptop, improving airflow:
custom water cooling:
frozen water bottle
condensing unit:
towel containing condensation

laptop ran cooler like this in hot weather than it did in an air conditioned airport sitting flat on a table.  

Image details:
DMK 51 web cam, Takahashi FS-60C, 60 mm aperture at f/4.2 with a reducer.  Baader solar film, Fotga IR/UV cut filter.  The full field of view is approximately 96x72 arc minutes with a resolution of 3.6"/px.  video capture at 12 fps, aligned in autostakkert, wavelets in registax.
-cell phone Droid Turbo ;)
Casper Wyoming
42,50.9694N 106,15.5688W