The Sun today - 1 image attached

3 views
Skip to first unread message

William Bottaci

unread,
Feb 12, 2024, 2:58:57 PMFeb 12
to croydo...@googlegroups.com
Today's sun.
  Altitude      25º
  Diameter   32º 24'
  Distance    0.987 AU (we are closest in winter, January 3rd)

The large group could be seen by a camera from Mars and it turned our way over a week ago, but with such rotten weather I was wondering if we'd ever see it. It was larger several days ago when an easy naked eye object; I saw it only when low down just over the rooftops and then only for minutes.

Canon R7 with RF 100-500mm L IS lens (set to 500mm, f/8, 1/200 sec, ISO 100), hand-held stabilised. Baader solar film filter.

A provisionally processed single image. My workflow for this image and the Sun in general (is not good enough to be a secret :)-

First expose so that the viewfinder histogram curve is not quite against the right/white edge (maximum brightness) use the camera histogram, even better if it features an over/under exposure clipping indicator when previewing.

 - Rotate to upright: I use the Solar Dynamics Observatory image from: https://www.spaceweather.com/
 - Crop closely;
 - Levels: for each colour of RGB adjust the black and white sliders to fit in only the light from the Sun, not the dark sky surrounding. Else use Curves, which give a slightly better result but requires a bit more knowledge;
 - Lower brightness and increase contrast, moderately; aim for a centre brightness of about 220 out of 255 or as it appeals to you;
 - Colour: remove any residual colour by converting to B&W. If done perfectly in Levels/ Curves any leftover colour would have been removed, but this is near impossible to judge;
 - Sharpening: I find that if working on a jpeg image any sharpening gives the sun a fine splodgy appearance. One could use a brush for a little local sharpening on the sunspots only; best not to overdo this effect because it'll show.

Shadows and Highlights, unlike for the Moon, are not appropriate for the Sun. Likewise, Levels is appropriate for the Sun but not the Moon. Clearly they are 'opposites' in this respect which reflects the Sun having a fraction of the tonal range of the sensor (known as 'flat', low contrast, like grey cloud) whilst for the Moon there is overlap on either side - too contrasty.
So, what would be the case for a solar eclipse? I think it won't be a challenge to fit in the bright prominences with the outer edge of the corona, maybe stretching the histogram in Levels (sliding the two 'points' towards either edge of the curve, depending on how you've exposed it).
For a lunar eclipse, the tonal range resembles the Sun.
Each eclipse is a different sense of 'opposites' (geometric - I'm sure this occurred... :).

If the Sun is white where would any residual colour come from? The intervening sky, which is predominantly but not completely blue, is still there and accounts for most of the difference. The filter is much of the rest.

I took 70 imagers, discarded all but 9, shortlisted further to only 3; turbulence was terrible. This evening looking at the crescent moon, about the same altitude, much calmer.

Thank you for looking, William

=CanR7_021874_M-sb.jpg (311K)=
CanR7_021874_M-sb.jpg
Reply all
Reply to author
Forward
0 new messages