Request: Script for Solar Eclipse

aparis1983

Request: Script for Solar Eclipse

« on: 10 / May / 2017, 09:40:07 »

Advertisements

Back story leading up to request:

In preparation for this year's solar eclipse on August 21st, I made some modifications to my Meade ETX-90 telescope and bought a Thousand Oaks solar filter.

In order to create the best composite image of a total solar eclipse and capture all of the different eclipse features (corona, chromosphere, baily's beads, prominences, etc) one needs to take as many shots as possible while cycling through different shutter speeds.

So, say I chose an ISO of 200, with f-stop at 11, I would ideally need to cycle through the following shutter speeds/exposures over and over during the 3 minutes that the total eclipse lasts: 1 second, 1/2, 1/4, 1/15, 1/60, 1/250, 1/500, 1/1000, 1/2000, 1/4000. Over and over again during 3 minutes. Without touching the camera so as to not ruin the exposures and of course just enjoy the eclipse without fiddling with the camera.

If there was 1 second in between shots, every cycle would take 10.3 seconds, which means I could cycle through these exposure about 17 times during the 3 minutes, taking a total of 156 pictures. And then use Registax and Photoshop to try and create an amazing composite image.

Request:

Now you see where I'm going with this. Is there an existing script or can a script be designed to set an ISO, an f-stop and then cycle through the different shutter speeds over and over and over again??

I have CHDK on both a Canon Powershot SD880IS and Canon EOS M3. Ideally, I will be running this on the M3.

« Last Edit: 10 / May / 2017, 09:41:41 by aparis1983 »

Logged

waterwingz

12083

Re: Request: Script for Solar Eclipse

« Reply #1 on: 11 / May / 2017, 02:15:36 »

I don't recall seeing anything that does exactly what you are looking for. I can take at converting something else is a couple of days.

Logged

Ported : A1200 SD940 G10 Powershot N G16

aparis1983

Re: Request: Script for Solar Eclipse

« Reply #2 on: 12 / May / 2017, 08:15:13 »

Hey waterwingz.
Needless to say, it would be great if this could be written into a script. Most eclipse photographers spend the entirety of the eclipse changing the settings on the camera and taking pictures. They end up missing the main event because they were never able to look up. So, a "hands-free" script would be invaluable.
Take a look at the attachment. It's a table for selecting shutter speeds during a total solar eclipse. Select an ISO, choose an f-stop on the right and then go down to see the different shutter speeds needed to capture the different solar features.

IMG_0977.GIF (44.93 kB, 759x949 - viewed 146 times.)

« Last Edit: 12 / May / 2017, 08:17:51 by aparis1983 »

Logged

lapser

1093

Re: Request: Script for Solar Eclipse

« Reply #3 on: 12 / May / 2017, 16:20:31 »

Thanks for posting that solar eclipse exposure table. It's very helpful, and much easier to understand than the ones I've seen.

I've been planning to jump back into script writing specifically to write something for the solar eclipse. I just got CHDK running on my EOS-M3, so I'm ready to go.

I have a script and CHDK build from a few years back that I use for time lapses of sunsets and lunar eclipses, but it doesn't run on the M3. It reads the raw buffer between shots and sets the exposure based on the average brightness, and the histogram. I see Reyalp has added this capability to CHDK now.

After hundreds of time lapses, I've discovered that I don't need a full histogram to set exposure. I just need to know the 2nd brightest pixel. I'm hoping Reyalp will consider returning that value along with the average brightness? If so, maybe we can discuss it in a different topic.

Regarding the proposed eclipse script, it could run in continuous (burst) mode and adjust both shutter speed and ISO. Saving 1 raw file per second is in the ball park. It would basically do continuous exposure bracketing with around 10 different exposures.

You can check out all the time lapses I've done here:
https://www.youtube.com/user/DrLapser/videos

This was a partially clouded out partial solar eclipse I did with the SX50. The script adjusted the exposure (using the 2nd brightest pixel) so the Sun remained visible through varying cloud densities..
https://youtu.be/Z7HZMdE64lA

This lunar eclipse with the SX50 used the 2nd brightest pixel (ETTR), which showed details on the lit part until totality.
https://youtu.be/tIjfgxDB9Pc

In this lunar eclipse (SX50), I set a limit of 2.5 ev below the average meter reading to show more of the eclipsed Moon at the end of totality.
https://youtu.be/dOdG1-hzo50

Finally, this is the same eclipse with the EOS-M through a telescope, adjusting the exposure manually.
https://youtu.be/SAvtH7wOIDI

Logged

EOS-M3_120f / SX50_100b / SX260_101a / G1X_100g / D20_100b
https://www.youtube.com/user/DrLapser/videos

c_joerg

1100

Re: Request: Script for Solar Eclipse

« Reply #4 on: 13 / May / 2017, 04:48:15 »

Quote from: lapser on 12 / May / 2017, 16:20:31

You can check out all the time lapses I've done here:

I think I've seen them all and I'm pretty envious at times…
Good job…

Quote from: lapser on 12 / May / 2017, 16:20:31

Thanks for posting that solar eclipse exposure table. It's very helpful, and much easier to understand than the ones I've seen.

Yes, of course. But my understanding was, he wants to use this tables in a script. Do you think it’s useful?

Quote from: lapser on 12 / May / 2017, 16:20:31

I just need to know the 2nd brightest pixel.

I have not really understood what that means. Do you have some more information about this?
What is different with what rawopint does?
rawopint:

Code: [Select]

The metering is done using an adjustable sized rectangle in the middle of the screen, and the script normally tries to keep this at the "neutral" value. Additionally, a histogram of the entire sensor area is used to limit over and under exposed pixels in the entire scene to user specified ranges. The relative influence of these limits may be adjusted, and the target exposure may be scaled with absolute scene brightness.

Quote from: lapser on 12 / May / 2017, 16:20:31

If so, maybe we can discuss it in a different topic.

Good idea…

Another ‚offtopic‘ Question. When you use the SkyTracker, do you have a synchronization between the SkyTracker and the camera?

I plan to build a slider and I want to send a signal from the script on the end of the shoot for moving…

« Last Edit: 13 / May / 2017, 04:50:49 by c_joerg »

Logged

M100 100a, M3 101a, G9x II (1.00c), 2*G1x (101a,100e), S110 (103a), SX50 (100c), SX230 (101a), S45,
Flickr https://www.flickr.com/photos/136329431@N06/albums
YouTube https://www.youtube.com/channel/UCrTH0tHy9OYTVDzWIvXEMlw/videos?shelf_id=0&view=0&sort=dd

lapser

1093

Re: Request: Script for Solar Eclipse

« Reply #5 on: 13 / May / 2017, 20:41:56 »

Quote from: c_joerg on 13 / May / 2017, 04:48:15

Yes, of course. But my understanding was, he wants to use this tables in a script. Do you think it’s useful?

Yes, for a basic solar eclipse script, you could look up the starting exposure in the table and enter it as a script parameter. Then you'd shoot a series of more and more overexposed shots until you get the dimmest part of the eclipse reasonably exposed. Then, repeat for 3 or 4 minutes until the total eclipse is over.

To process the photos, you'd need to use an HDR program (not Registax) to combine the photos into a single image.

Quote

Quote from: lapser on 12 / May / 2017, 16:20:31
I just need to know the 2nd brightest pixel.
I have not really understood what that means. Do you have some more information about this?
What is different with what rawopint does?

My script sets the exposure so the value of the 2nd brightest pixel is 1/2 of the white level (adjusted for black level bias). My reference point is 1/2 of white level, or 1 ev (f-stop) below white level.

My script also computes an average pixel value and calculates the exposure change needed to bring it to 1/8 of white level, or 3 ev below white level. I've found that this duplicates the camera meter very closely (12.5% gray).

Quote

Another ‚offtopic‘ Question. When you use the SkyTracker, do you have a synchronization between the SkyTracker and the camera?

I plan to build a slider and I want to send a signal from the script on the end of the shoot for moving…

I don't have any synchronization. The SkyTracker actually makes stars much sharper, especially to the south. Close foreground objects may be blurred in front of sharp stars. The SkyTracker has a 1/2 speed switch to split the difference, but I haven't used it.

I suspect that you can get away without synchronization in the daytime using a rail, at least for wide angle shots. At night, it might be helpful, especially with a close foreground object. You'll have to experiment!

Logged

EOS-M3_120f / SX50_100b / SX260_101a / G1X_100g / D20_100b
https://www.youtube.com/user/DrLapser/videos

c_joerg

1100

Re: Request: Script for Solar Eclipse

« Reply #6 on: 15 / May / 2017, 07:58:39 »

Thanks for explanation…

Quote from: lapser on 13 / May / 2017, 20:41:56

The SkyTracker actually makes stars much sharper, especially to the south.

Yes of course, that’s the reason to use a The SkyTracker…

Quote from: lapser on 13 / May / 2017, 20:41:56

I suspect that you can get away without synchronization in the daytime using a rail, at least for wide angle shots.

I plan to use the slider on the ocean. I want to have longer exposer times (with a ND3.0) at daytime, to smooth the waves. In this case I have probably to synchronize.

Logged

waterwingz

12083

Re: Request: Script for Solar Eclipse

« Reply #7 on: 25 / May / 2017, 21:23:06 »

Quote from: waterwingz on 11 / May / 2017, 02:15:36

I don't recall seeing anything that does exactly what you are looking for. I can take at converting something else is a couple of days.

New thread to discuss this (with working code) started here.

Logged

Ported : A1200 SD940 G10 Powershot N G16

topshot

26
S100

Re: Request: Script for Solar Eclipse

« Reply #8 on: 09 / August / 2017, 12:46:09 »

Quote from: lapser on 13 / May / 2017, 20:41:56

My script sets the exposure so the value of the 2nd brightest pixel is 1/2 of the white level (adjusted for black level bias). My reference point is 1/2 of white level, or 1 ev (f-stop) below white level.

Would you mind sharing how that is done? Since I only intend to do landscape shots, the sun will appear as a small orange dot moving across the black frame during the partial phases, which I suspect will throw off any type of exposure calculation. However, if exposure could be set to 2nd brightest pixel found in the frame, I'd think that would solve the issue, right?

Logged

lapser

1093

Re: Request: Script for Solar Eclipse

« Reply #9 on: 09 / August / 2017, 14:25:06 »

Quote from: topshot on 09 / August / 2017, 12:46:09

Would you mind sharing how that is done? Since I only intend to do landscape shots, the sun will appear as a small orange dot moving across the black frame during the partial phases, which I suspect will throw off any type of exposure calculation. However, if exposure could be set to 2nd brightest pixel found in the frame, I'd think that would solve the issue, right?

I should have a script ready for the eclipse in a few days, which I'll post here. Measuring the 2nd brightest pixel required some changes to reyalp's raw meter code, but I just got that part working. Stay tuned.

I'm thinking that I'll first write a script that does extreme exposure bracketing for the total eclipse part. It will start at a specific time (hhmmss) or with a keypress (or possibly when the solar filter is removed). You'll set the start time, minimum and maximum shutter time and ISO, interval etc. as script parameters.

I'm planning to use the 2nd brightest pixel to set the exposure of the first bracketed shot, and then increase the exposure in steps to around 8 to 10 ev above the first shot.

Then if I have time, I'll try to add exposure adjusting script code for the partial phase. That requires smoothing if you want to show it as a time lapse.

I'm also thinking that if there are still blown pixels at the minimum shutter time, i.e. 1/2000 secs or so, that means the filter is off and the diamond ring or the beads are visible. I won't start bracketing until there are less than 2 or 3 blown pixels. I'll stop bracketing when there are blown pixels again at the end of the total phase.

If anyone has any suggestions, let me know (soon).

Logged

EOS-M3_120f / SX50_100b / SX260_101a / G1X_100g / D20_100b
https://www.youtube.com/user/DrLapser/videos