Dark Frame for astronomy

Started by skypix, August 21, 2017, 09:30:27 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

skypix

Hi all !
Please help,  if this has been discuss before I can't find thread.
i have a question , can magic lantern take dark frame for astrophotography?
can we add menu in time lapse and for some frame set photography without open shutter ! in dark frame the camera shutter closed and the sensor does not receive light.
Thank you for your help

a1ex

Take a look at changesets 51da5cdb8dde and 99be9612028c, and this discussion. They are not included in the current codebase, as I wasn't happy with the increased code complexity; to get started, you can just compile ML from the second changeset.

I suggest moving them to a new module (dark.mo? other name?), and maybe have it average the dark frames as well, see EekoAddRawPath. From the silent picture module, you will need the "basic" full-res capture functionality, but outside LiveView.

One issue to solve: the captured area from a ML DNG doesn't match the CR2 exactly (for some reason, Canon code crops a few pixels from the raw buffer before saving it). This will have to be adjusted for every single camera model. Besides, as this one is not a widely used feature, you can't rely on bug reports, so you'll also have to come up with some testing procedure (ideally automated).

As I'm not really into astrophotography, are you (or anyone else) willing to step up and polish the above patches?

calypsob

Quote from: a1ex on August 22, 2017, 01:28:54 AM
Take a look at changesets 51da5cdb8dde and 99be9612028c, and this discussion. They are not included in the current codebase, as I wasn't happy with the increased code complexity; to get started, you can just compile ML from the second changeset.

I suggest moving them to a new module (dark.mo? other name?), and maybe have it average the dark frames as well, see EekoAddRawPath. From the silent picture module, you will need the "basic" full-res capture functionality, but outside LiveView.

One issue to solve: the captured area from a ML DNG doesn't match the CR2 exactly (for some reason, Canon code crops a few pixels from the raw buffer before saving it). This will have to be adjusted for every single camera model. Besides, as this one is not a widely used feature, you can't rely on bug reports, so you'll also have to come up with some testing procedure (ideally automated).

As I'm not really into astrophotography, are you (or anyone else) willing to step up and polish the above patches?

The cropping that you mention was actually addressed in the program pixinsight. I think canon does this because the pixels at the edge are used for calibration of the columns.  It seems that the best option here would be to allow the user to enter the sensor size and pixel size independently so  you do not need to write code for every camera model.   That being said however I am a bit off put by the OP desired results.  Is he wanting to do dark frames by having the camera close its shutter similar to LENR?
Full spectrum T2i
T3i

a1ex

I assume he wants to take pictures with the mirror down (dark frame without having to put the lens cap).

With the revisions linked above, ML does exactly that (but the resulting image is slightly larger than Canon's CR2, so you'll have to match the active areas).

Whether the resulting dark frame is any good, or is affected by light leaks, I have no idea.