Take a look at the following image by Franky Chou:

金龍二亭 by FrankyChou, licensed CC-BY-SA

I'm trying to understand how it was even possible. I mean, it's clear how you do a stars' trail shot, but with a long exposure, I feel like the bright area should have basically "light-polluted" the entire image.

Ok, you can stack 2 images with different exposure in post-processing to achieve the effect of a well illuminated area with a beautiful starry sky, but let's say the photographer didn't have possibility to turn off the light there... How is it possible that the landscape appears on the left too? It should have been totally eliminated by the light anyway.

The photo is a composite for SURE, but I wanted to know how come the left of the landscape seems not to have been destroyed by the light

  • 2
    \$\begingroup\$ What makes you believe that shot isn't a composite? \$\endgroup\$
    – Philip Kendall
    Jul 1, 2015 at 7:29
  • \$\begingroup\$ Check out this related question, about a photo of a brightly-lit tree with a starry background — although in that case, a tracking mount was used so there are no star trails. photo.stackexchange.com/questions/30206/… \$\endgroup\$
    – mattdm
    Jul 1, 2015 at 14:49
  • \$\begingroup\$ I'm going to go the other way on this: what makes you sure this is a composite? \$\endgroup\$
    – mattdm
    Jul 1, 2015 at 14:53
  • \$\begingroup\$ How would you expose (I'm going to wildly guess here) 1m30sec to obtain star trails, without polluting the left part of the image? The gazebo there seems to me a public place, which leads to think that the illumination would consist at least of a light bulb (not so dim as stated in the answers below) \$\endgroup\$ Jul 2, 2015 at 6:02
  • \$\begingroup\$ Turn the light bulb on for only part of the exposure. When that's the most likely explanation, why rule it out? \$\endgroup\$
    – mattdm
    Jul 5, 2015 at 14:17

3 Answers 3


but let's say the photographer didn't have possibility to turn off the light there...

Why not? You are making it artificially complicated by not allowing it without knowing if it was indeed not possible. Maybe he just did it this way.

Turning a light off can be as simple as covering it with something. Take a look at how the shadows of the framework fall: the main light (which is outside the image) is positioned very low. I guess a flash with a gel on a low stand.

What you see in the ceiling is either a reflection from the main light source (most likely) or a very dim light. It isn't very strong casting almost no shadows at all.

And of course, we are in Photoshop land. Who says that the foreground and background are photographed at this loaction together? I'm not a PS wizard, but given a star trail background I think it is also plausible to add this sky to the picture. It's a bit of masking work on the tree, but the leaves are fairly distinct from the background. It wouldn't take long.


One possibility is the combination of a long exposure time + flash; you do the long exposure for the sky and, at the end of the exposure, you use a flash/strobe to expose the house/model (in this case).

Of course you have to take into account the placement of the flash, so it wouldn't shine directly at your lens, but I believe this is possible.

  • \$\begingroup\$ I don't think so. In this case, the left part of the image would have been overcome by the light \$\endgroup\$ Jul 6, 2015 at 5:43
  • \$\begingroup\$ Not if you leave the lights off until the end - remember, we are talking about an exposure of many seconds - to a few minutes. \$\endgroup\$ Jul 7, 2015 at 8:34

They don't need flash. In bulb mode, a long exposure would bring plenty of light. The model would have to stay firm at least 3 hours for that much movement of the stars. I think is possible only with a low light long exposure and photoshop balance

  • \$\begingroup\$ Absolutely not. With a minimum of wind you would have a picture with blurry tree and so on. With that stars' movement I don't believe that the exposure was more than 5 minutes for the sky \$\endgroup\$ Jul 6, 2015 at 5:44
  • \$\begingroup\$ in the picture you can see blur in the tree, and 5 min? for that movement? i don't think so "Typical exposure times for a star trail range from 15 minutes to several hours". link \$\endgroup\$ Jul 6, 2015 at 14:30
  • \$\begingroup\$ A 3h duration star trail would have each star tracing out 1/8th of a circle (45 degrees). The arc of each trail is clearly much less than 45 degrees. \$\endgroup\$
    – user13451
    Jul 6, 2015 at 15:07
  • \$\begingroup\$ I totally agree. Hours if you want FULL arches... And blur in the tree? Where? \$\endgroup\$ Jul 9, 2015 at 6:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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