4 ms·

Does anyone know of a similar guide for digital sensors, since they can do these images with very different exposure times?

lm28469 6y ago
Digital has other issues like hot pixels and more noise on longer exposures, they usually do a couple (dozens, hundreds, ..) of shorter exposures and stack them in post prod. It's basically the same process otherwise.
joshvm 6y ago
There is a huge volume of literature for astrophotography - I think most random hits you'd get on Google will be OK (the vast majority will be digital oriented). For example: http://www.astropix.com/html/i_astrop/toc_ap.html http://www.astropix.com/html/i_astrop/toc_ap.html I can probably give a better recommendation if you have a particular goal in mind. There are plenty of guides on how to take photos of the Milky Way with an SLR, but astrophotography is a broad (and quickly very expensive) field to get into. If you're interested in a more technical angle, you can look at the Handbook of CCD Astronomy (Howell, 2006) which also discusses noise issues in detail. Nowadays we're moving to CMOS though. However when you go for astrophysics books they tend to focus more on measuring intensities accurately (photometry) rather than getting pretty pictures. Although one nitpick with the article - tracking mounts do not follow the rotation of the Earth, they rotate in the opposite direction.
- Stratoscope 6y ago
  > Although one nitpick with the article - tracking mounts do not follow the rotation of the Earth, they rotate in the opposite direction. Maybe I missed something, but I don't see where the article contradicts what you wrote above. The one thing I did find seems correct (as it doesn't specify the direction): "With film and these very long exposures, the Earth’s rotation is a big problem — you’ll end up with star trails (which is a fun type of photo to take). The solution to this problem is a star tracker, a mount that precisely rotates your camera at the same rate as the Earth (known as sidereal rate)."
  - joshvm 6y ago
    This is true, I interpreted that to mean with the Earth, rather than against. But as you say, the direction isn't actually specified.
miahi 6y ago
Most of the guides for digital will concentrate on post-processing, as the actual shot is quite simple, and you can tweak everything as you have instant feedback. If you can see the milky way with the naked eye, you need: tripod + wide lens (24mm on full frame is ok) with wider aperture, manual focus (using the display zoom function for best results) as autofocus on stars doesn't always work, manual exposure for something like ISO 1600-3200, around 20 seconds exposure on f/2.8, 10 second timer exposure, press the shutter and leave the camera to stabilize on the tripod. The key is to find the places where you can see the milky way and to plan ahead - a moonless and cloudless night is important. There are websites[1] and tools[2] or that. Some results based on the above, with minimal post processing (as I'm not good at it): https://miahi.ro/photos/perseide-2017/ https://miahi.ro/photos/perseide-2017/. The first series is at ~1500m altitude and 70km distance of any large city. [1] http://darksitefinder.com http://darksitefinder.com [2] https://stellarium.org/ https://stellarium.org/
- glup 6y ago
  Adding to that: at least historically, full frame DSLRs took better low-light pictures; smaller sensors had more noise because of the density. Not sure if this is still true.
  - snazz 6y ago
    It's definitely still true, but the difference is getting smaller as sensor tech improves. My DSLR is a Nikon D80 from 2006. While it's not full-frame, its sensor is much bigger than the one on my iPhone 8, yet my phone works far better in low-light conditions. The highest non-expanded sensitivity on the D80 is ISO 1600, which is already very grainy. Bump that up to the highest possible ISO 3200 and you're in for a lot of touch-up work.
gerikson 6y ago
The site I’ve found to have good info about this (including lenses) is lonelyspeck.com

lm28469 6y ago

Digital has other issues like hot pixels and more noise on longer exposures, they usually do a couple (dozens, hundreds, ..) of shorter exposures and stack them in post prod. It's basically the same process otherwise.

joshvm 6y ago

There is a huge volume of literature for astrophotography - I think most random hits you'd get on Google will be OK (the vast majority will be digital oriented). For example: http://www.astropix.com/html/i_astrop/toc_ap.html http://www.astropix.com/html/i_astrop/toc_ap.html I can probably give a better recommendation if you have a particular goal in mind. There are plenty of guides on how to take photos of the Milky Way with an SLR, but astrophotography is a broad (and quickly very expensive) field to get into. If you're interested in a more technical angle, you can look at the Handbook of CCD Astronomy (Howell, 2006) which also discusses noise issues in detail. Nowadays we're moving to CMOS though. However when you go for astrophysics books they tend to focus more on measuring intensities accurately (photometry) rather than getting pretty pictures. Although one nitpick with the article - tracking mounts do not follow the rotation of the Earth, they rotate in the opposite direction.

Stratoscope 6y ago
> Although one nitpick with the article - tracking mounts do not follow the rotation of the Earth, they rotate in the opposite direction. Maybe I missed something, but I don't see where the article contradicts what you wrote above. The one thing I did find seems correct (as it doesn't specify the direction): "With film and these very long exposures, the Earth’s rotation is a big problem — you’ll end up with star trails (which is a fun type of photo to take). The solution to this problem is a star tracker, a mount that precisely rotates your camera at the same rate as the Earth (known as sidereal rate)."
- joshvm 6y ago
  This is true, I interpreted that to mean with the Earth, rather than against. But as you say, the direction isn't actually specified.

miahi 6y ago

Most of the guides for digital will concentrate on post-processing, as the actual shot is quite simple, and you can tweak everything as you have instant feedback. If you can see the milky way with the naked eye, you need: tripod + wide lens (24mm on full frame is ok) with wider aperture, manual focus (using the display zoom function for best results) as autofocus on stars doesn't always work, manual exposure for something like ISO 1600-3200, around 20 seconds exposure on f/2.8, 10 second timer exposure, press the shutter and leave the camera to stabilize on the tripod. The key is to find the places where you can see the milky way and to plan ahead - a moonless and cloudless night is important. There are websites[1] and tools[2] or that. Some results based on the above, with minimal post processing (as I'm not good at it): https://miahi.ro/photos/perseide-2017/ https://miahi.ro/photos/perseide-2017/. The first series is at ~1500m altitude and 70km distance of any large city. [1] http://darksitefinder.com http://darksitefinder.com [2] https://stellarium.org/ https://stellarium.org/

glup 6y ago
Adding to that: at least historically, full frame DSLRs took better low-light pictures; smaller sensors had more noise because of the density. Not sure if this is still true.
- snazz 6y ago
  It's definitely still true, but the difference is getting smaller as sensor tech improves. My DSLR is a Nikon D80 from 2006. While it's not full-frame, its sensor is much bigger than the one on my iPhone 8, yet my phone works far better in low-light conditions. The highest non-expanded sensitivity on the D80 is ISO 1600, which is already very grainy. Bump that up to the highest possible ISO 3200 and you're in for a lot of touch-up work.

gerikson 6y ago

The site I’ve found to have good info about this (including lenses) is lonelyspeck.com