« previous next »
Pages: [1] 2   Go Down

testing eclipse filters using zebra or histogram

  • 10 Replies
  • 1050 Views
*

Offline axman

  • ***
  • 143
testing eclipse filters using zebra or histogram
« on: 11 / August / 2017, 19:39:46 »
Advertisements
Hello, all

I'm trying to use a crippled ixus160 as an eclipse-filter tester, before deploying said filter to more expensive cameras..  Filter is #14 Welder's shade.

Any ideas or comments are welcome.

Idea1:  Set CHDK exposure override to absolute minimum, take an image, then review the image with zebra active.  Gradually increase the exposure time until it's close to overexposed, and call that good enough.

Idea2:  Using histogram - I guess Blend Y would be best choice, but this way leaves no margin for error - if the welder's shade is poor quality (eg not really a #14 shade) I will have valid experimental results, as well as a fried image sensor.

Idea3: Is it possible to use chdkptp 'lvdumpimg' cmd to capture the zebra which is overlaying the CHDK osd?  And/Or, is it possible to have zebra on at all times, not just during half-press?  I saw some posts on how to use lua cmds to invoke half-press, I'll try to search & read more about this possibility.

Experimenting is needed because I have no idea what exposure settings would be best in the two use cases: 

1) full sun until totality behind welder's shade
2) no sun during totality, and no welder's shade

I also have the sense that histogram and zebra values from ixus160 will be different than my other Canon's..  Not sure how much different, so experimenting with crippled cam (to get "safe exposure zone" is the first pass;  )  If success, (no fried image sensor) then I'll reduce exposure time by 50% and try it on the sx530hs..

Planning to use my SX530HS and M10 (with the long lens) for taking the actual keeper images.  If I get this right, pics will be &*^%$#ing awesome! 

Thanks in advance.

ps Saw @wwingz mention of script for eclipse image snarfing.  Hoping to experiment with that too, but as fall back, my existing chdkptp stuff will do.
Logged
*

Offline waterwingz

  • ******
  • 11608
    • Waterwingz
Re: testing eclipse filters using zebra or histogram
« Reply #1 on: 11 / August / 2017, 19:46:16 »
Quote from: axman on 11 / August / 2017, 19:39:46
ps Saw @wwingz mention of script for eclipse image snarfing.  Hoping to experiment with that too, but as fall back, my existing chdkptp stuff will do.
FWIW, my various posts have all be specific answers to questions that start with "How do I ..". 

I have no idea how best to photograph an eclipse - although I'm collecting some generally great ideas from the questions posted.

YMMV.
Logged
Ported :   A1200    SD940   G10    Powershot N    G16
*

Offline reyalp

  • ******
  • 11546
Re: testing eclipse filters using zebra or histogram
« Reply #2 on: 11 / August / 2017, 23:23:44 »
YMMV, but many of us have taken multiple timelapses where the sun passed through the field of view without apparent harm. So I'd expect a welders shade to be fairly safe for the camera.

Attached are images I took of an eclipse and the transit of Venus through a welders shade (unknown value, sorry) with an A540.
Logged
Don't forget what the H stands for.
*

Offline axman

  • ***
  • 143
Re: testing eclipse filters using zebra or histogram
« Reply #3 on: 12 / August / 2017, 00:36:22 »
Thanks for the feedback, guys.

Preliminary tests did not obviously ruin the ixus160.  But gotta check for sure, there are some "stars" appearing in the same place, in all the test images.  Could be dirt on the lens, or..  Stuff I read and pictures I saw re: cameras getting fried by laser seemed to have an entire row of pix ruined, or a row+column ruined, or complete garbage images where pixels used to be..

Reviewing test images using CHDK histogram showed a peak in green channel (did i say that right?) that was maybe overexposed.  Zebra review of images with "overexposure sensitivity" set at 32 did not cause zebra to activate.  But,  welder's shades, I think, are going to be ok.  Encouraged by reyalp's pics!  The ones I took are too big for attachments.  Will post a link to file drop if any interest to see them.

One thing I noticed:  if I zoom in when viewing images on PC, I can see details of the corona, and stars in the background.  Awesomely spooky physics going on..

Re: photographing the eclipse: 

I found this link  http://www.clarkvision.com/articles/photograph-the-sun/  by Dr Roger Clark, retired NASA scientist, and photo/digital imaging expert.  Going to take me several readings to digest that page, but lots of good info.  Check it out.

edit: correct Dr. Clark's name
« Last Edit: 24 / August / 2017, 18:51:01 by axman »
Logged
*

Offline reyalp

  • ******
  • 11546
Re: testing eclipse filters using zebra or histogram
« Reply #4 on: 12 / August / 2017, 03:29:14 »
Quote from: axman on 12 / August / 2017, 00:36:22
Preliminary tests did not obviously ruin the ixus160.  But gotta check for sure, there are some "stars" appearing in the same place, in all the test images.
If they are always at the same pixel position in the image, they are most likely hot pixels.


Quote
One thing I noticed:  if I zoom in when viewing images on PC, I can see details of the corona, and stars in the background.  Awesomely spooky physics going on..
If you are taking pictures of the sun through welding glass (or the un-eclipsed sun without), you definitely aren't seeing stars or significant corona. More likely you are seeing the poor optical properties of the welding glass.
Logged
Don't forget what the H stands for.
*

Offline axman

  • ***
  • 143
Re: testing eclipse filters using zebra or histogram
« Reply #5 on: 15 / August / 2017, 14:48:42 »
Test using SX530HS.

The DNG of the attached jpg shows what looks like corona when zoomed way in.  I guess the fuzzy stuff around the sphere is due to scattering (absorption?) or internal reflection of the light as it's coming thru welder's shade.  I think that's what you meant by poor quality optics. 

Trying to find spec sheet.  I agree that a #14 welder's shade != Thousand Oaks solar filter, but I'm curious to know how much different, and what effects would be expected when using one or the other.

Image:  f/5.6  1/260s  ISO 100  with lens zoomed out (66.59mm reported by rawtherapee)
Logged
*

Online blackhole

  • *****
  • 615
  • A590IS 101b
    • Planetary astrophotography
Re: testing eclipse filters using zebra or histogram
« Reply #6 on: 16 / August / 2017, 12:47:32 »
Quote
Trying to find spec sheet.  I agree that a #14 welder's shade != Thousand Oaks solar filter, but I'm curious to know how much different, and what effects would be expected when using one or the other.
The solar filter will give you a sharper image.
Baader AstroSolar ND5 filter is used for this image, the camera is sx410IS.
The image is taken from the hand, the magnification is 40x (optical) + 20x (ZoomPlus).
« Last Edit: 16 / August / 2017, 12:55:37 by blackhole »
Logged
*

Offline axman

  • ***
  • 143
Re: testing eclipse filters using zebra or histogram
« Reply #7 on: 17 / August / 2017, 15:41:23 »
@blackhole, thanks for info.  That is a very fine image.

Good filters or bad, in any case, we will have fun.  :-)  I would love to capture nice images.

But the most important thing is to see this eclipse with my own eyes.  It will be something to remember.
Logged
*

Online blackhole

  • *****
  • 615
  • A590IS 101b
    • Planetary astrophotography
Re: testing eclipse filters using zebra or histogram
« Reply #8 on: 17 / August / 2017, 16:54:33 »
Quote from: axman on 17 / August / 2017, 15:41:23
@blackhole, thanks for info.  That is a very fine image.

Good filters or bad, in any case, we will have fun.  :-)  I would love to capture nice images.

But the most important thing is to see this eclipse with my own eyes.  It will be something to remember.

It will certainly be an unforgettable experience.
Welding glass will serve well.
My friend goes to Jefferson City, Missouri, he will hold a lecture about astrophotography.
Have fun, and good luck with shooting!
Logged
*

Offline axman

  • ***
  • 143
Re: testing eclipse filters using zebra or histogram
« Reply #9 on: 29 / August / 2017, 13:05:15 »
Destination: ~15 mi (20km) west of Lime, Oregon.  Camped on BLM land just off of Mormon Basin road, along with a few hundred other folks.  44.424, -117.567.

Camp was on top of a smaller ridge in between surrounding taller ones.  Estimate elevation at 6000' (1800m) above sea level.  Extreme dry conditions meant no camp fire, no charcoal, or open flames.  Sandwiches!  And lots of water.

CHDK scripting was mostly fail, my fault.  Nothing wrong with scripts or CHDK.  Poor timing and power save on the LCD, and other operator errors.  I got a couple of quick shots in manually, one of them is pretty good.

Also had un-attended video cam facing SW/W that recorded totality coming and going.  It goes out of focus pretty bad.  Links below to a 36 sec movie w/no sound.  One jpg is extracted every 15 secs of video; playback encoding shows each extracted frame for 2 sec.  Firefox played both videos ok.

mp4 version https://drive.google.com/open?id=0B1mUSgMF9DmfRnVjcmQxSC1sLUE

webm version https://drive.google.com/open?id=0B1mUSgMF9DmfZTZic0trSlo2bDQ

Trip was amazing.  Worth it in every regard, with lasting and good memories.  Here's my best pic, taken on an sx530hs.

Thanks to all who develop and support CHDK.
Logged
Pages: [1] 2   Go Up
« previous next »
 

Related Topics