0
\$\begingroup\$

I am from sri lanka. I want to photograph milky way.

Canon 750D EFS 18-135 lens

I know that i should increase aperture & increase iso.

What should be the shutter speed for this camera to capture milky way without startrails ?

\$\endgroup\$
2

4 Answers 4

4
\$\begingroup\$

To avoid trails you have to know how long a star stays in the same pixel. A quick way to figure it out:

  • Take a picture of the moon with the your lens.
  • Measure the diameter of the moon (in pixels) in the resulting picture.
  • The moon is about half a degree, the sky rotates by 15°/hour (30 moons/hour), in other words one moon per 2mn (or 120 seconds).
  • So the sky moves by one pixel every 120s/N, N being the size of the moon in your shots

Depending on the sharpness of your lens, star trails can show as soon as you get over one pixel, or with a softer lens you can extend this to 2/3 pixels.

\$\endgroup\$
2
\$\begingroup\$

The very best way to do astrophotography without star-trails is to use a motorized tracking head for the tripod (such as a Sky Watcher "Star Adventurer" head or an iOptron "Sky Guider Pro" head.) Preferably this would be on a sturdy tripod to avoid vibrations.

You'll find you are able to use longer lenses and take narrower field images (although to astrophotographers these are all "wide field" images).

For a non-tracking mount... you have to deal with the fact that the Earth is rotating (about 15 arc-seconds of angular rotation per second). There's a complicated formula ... and an easy (less precise but good enough) formula. Let's do the easy one.

For your APS-C size sensor, you can divide 325 by the focal length of the lens. This provides the maximum number of seconds you should expose before your stars begin to become elongated due to the rotation of the Earth.

Doing this with your 18-135 at 135mm... you can see 325 ÷ 135 = 2.4 seconds ... not enough. But at 18mm it's about 18 seconds (much better).

But another challenge is how much light you can collect in that amount of time. At 18mm it's an f/3.5 lens.

Rokinon makes a 16mm f/2 lens that in a Canon EF-S mount version. It's a completely manual lens (no auto-focus, no auto-aperture, etc.) but that's ok because in astrophotography you wont use those features.

This would get you a 20 second exposure with no star elongation. At f/2 it will collect considerably more light than the 18mm at f/3.5 (which means you can use a lower ISO to hopefully get less "noise").

Focusing is a challenge. Stars are not very bright. But everything in the night sky is focused at the same distance. So if you can focus on anything in the night sky, then everything in the night sky would use that same focus. This means you can swing the camera around to find a nice bright star to make it easier. The Canon also uses "exposure simulation" when using live-view. So turn on "live view" mode. Manually rotate the focus on the lens to near the "infinity" point (that wont be accurate but at least you'll be close). Set the exposure time to 30 seconds. Set the ISO to max. This will brighten up your display. Gently adjust focus until you see some stars. Center them (actually it's better if you slightly de-center them ... maybe 1/3 away from the center of frame). Use the live-view zoom feature to take it to the 10x zoom. Now carefully adjust focus to make the star as small as possible.

Swing the camera back around to the area of sky you want to image (being especially carefully NOT to touch the focus at this point).

You can return the shutter speed to 20 seconds and return the ISO to something more sane (maybe ISO 800 ... test it).

Use a remote shutter release. If you don't have one, set the self-timer so that the vibration created from touching the shutter button will have time to settle down before the camera takes the photo.

Take a test exposure to confirm your framing. Also... VERY important... inspect the test exposure carefully (ideally unload the image from the camera and inspect it on a computer ... but otherwise hit the review button on the camera and zoom in to max so you can inspect the pixels). You are checking to see if you really nailed the focus. This is a step I am guilty of brushing over ... checking the image in the field (without careful inspection) and thinking "that looks pretty good to me". When I got home and inspected the image ... they were a little soft and mushy (if ONLY I had taken more time). Your patience up front will be rewarded.

I don't normally do Milky Way photos... I'm using shooting deep-sky objects with a telescopes (that gets a lot more complicated).

\$\endgroup\$
1
\$\begingroup\$

To capture images of the Milky Way without star trails, you must use a wide-angle lens along with a shutter speed of no longer than 30 seconds, with an aperture of f/2.8 and an ISO of 3,200. A shutter speed longer than 30 seconds will result in the stars being recorded as smudges or trail. The longer the focal length, the faster your shutter speeds will need to be in order to capture stars as pinpoints.

Since the largest aperture for your lens is f/3.5, which is 1/2 stop smaller than f/2.8. Therefore your exposure should be 30 seconds at f/3.5 with an ISO of 4,000. I'm not sure what the exact number is, but your if your camera is set up to advance ISO settings by 1/2 stop (as opposed to 1/3 stops) then choose the number between 3,200 and 6,400.

One very important thing to do is shoot RAW. DO NOT SHOOT JPEGs. You will need the extra data in RAW files to properly adjust the brightness, contrast and color of the image. When you shoot in JPEG, you are literally deleting 94% of the data that is in a RAW file. So shoot RAW.

\$\endgroup\$
1
  • 2
    \$\begingroup\$ I am interested in your statement about losing 94% of the data when shooting in Jpeg. Can you elaborate on that? If I post it as a question, can you answer it with detail? many thanks \$\endgroup\$ Commented Mar 4, 2019 at 23:58
0
\$\begingroup\$

Blur trail length depends on your lens zoom mm used, and also your sensor size and shutter duration. My web site offers a calculator for this specific blur trail question, at https://www.scantips.com/lights/stars.html

However, this blur trail length is independent of proper exposure, which are not the same factors. We always need longer exposure for the milky way stars, but which is a serious problem on a fixed mount that does not track the star motion. Other than a rotating tracking mount, using a short wide focal length is the best tool to minimize the blur.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.