6
\$\begingroup\$

How to keep the Girls sharp combined with the lighttrace effect?

https://i.sstatic.net/AQtmg.jpg

Improve this question
\$\endgroup\$
7
  • \$\begingroup\$ See When should you use a normal flash vs a second-curtain flash? and What is second-curtain or rear-curtain technique? \$\endgroup\$
    – dpollitt
    Commented May 15, 2013 at 12:32
  • 1
    \$\begingroup\$ @dpollitt - minor point, but in this case, second vs first curtain doesn't matter as we don't know the direction the phone was moved in. There isn't any apparent motion other than the light trail, so it could have been either. \$\endgroup\$
    – AJ Henderson ♦
    Commented May 15, 2013 at 13:21
  • \$\begingroup\$ @heysamhey - do you have the original link to the image on the tillate site, maybe there's a way we can find the author and ask them what they did? \$\endgroup\$
    – James Snell
    Commented May 15, 2013 at 14:02
  • \$\begingroup\$ @JamesSnell I just wrote him a message! \$\endgroup\$
    – heysamhey
    Commented May 15, 2013 at 17:46
  • \$\begingroup\$ @heysamhey - cool, if you get an answer from them then please do share it, it's a cool pic and has given me a couple of ideas! \$\endgroup\$
    – James Snell
    Commented May 15, 2013 at 19:57

4 Answers 4

Reset to default
6
\$\begingroup\$

For this kind of a shot, it doesn't matter if the flash is first or second curtain because there is no direction to the movement. (ie, the phone could have started on the left or the right side and we'd have no idea of the difference). What does matter is that you have a dark environment so that only the light trail is normally exposing. You then take either a long exposure or a bulb exposure (which keeps the shutter open as long as you hold the button down).

The flash exposes the people and since it is dark, they don't continue to expose much, but the light trail from the bright screen exposes quickly and thus you see the light trail but not much else that was exposed during that part of the shot. It also probably took a fair bit of practice (or luck) to get it to come out that well as typically the light trails would be less even.

It's also possible that they faked the effect in Photoshop since that is one heck of a smooth circle. It's certainly possible to get that effect in camera, but they seem to have done a remarkably good job in many aspects, so either they did a lot of time and patience to get it just right or they blended it in artificially (in part or full).

To produce it in post, using an additive blend would properly mix the light sources in the image with the light source of the rainbow such that the sum of the brightest spots would shine through. Any tool of choice could be used to produce the actual arc.

Improve this answer
\$\endgroup\$
7
  • 5
    \$\begingroup\$ +1 on the rainbow being a post production job. It looks a little bit too perfect to me too. \$\endgroup\$
    – James Snell
    Commented May 15, 2013 at 14:05
  • \$\begingroup\$ The "non-post ideas do not account for what appears, to me, to be background intruding over foreground at the top of the arc as shown in my crop. Or, are my eyes/brain deceiving me? \$\endgroup\$
    – Russell McMahon
    Commented May 20, 2013 at 15:56
  • 1
    \$\begingroup\$ @RussellMcMahon - actually quite the opposite. What you are seeing there is additive blending. The greater light intensity shines through as the highest degree of exposure. This is what it would be expected to look like if a point light source was in the path of a dimmer moving light source when done live. On the other hand, doing an additive layer blend would accomplish the same thing in post. \$\endgroup\$
    – AJ Henderson ♦
    Commented May 20, 2013 at 16:08
  • \$\begingroup\$ @JamesSnell I don't think it was faked, it's perfectly possible to hold your arm straight, and a straight arm pivoting around a fixed point (the shoulder joint) will create an almost perfect circle. People are far to quick to cry Photoshop these days, ask yourself: 1) why the phone is in the shot at exactly the end of the arc with what appears to be the right colours on the screen, and 2) who has the time to Photoshop special effects onto event photos (where thousands will be shot in a night) when it can be done easily in camera. \$\endgroup\$
    – Matt Grum
    Commented Jan 10, 2014 at 9:11
  • \$\begingroup\$ oh and 3) why use slow flash sync (evidenced by the blurred highlights in the background) if you're going to Photoshop in the rainbow? Common sense says it was done in camera. \$\endgroup\$
    – Matt Grum
    Commented Jan 10, 2014 at 9:12
3
\$\begingroup\$

You would be using a flash with the camera set for a longer exposure,the flash would freeze the woman's movement and the long exposure will allow for the light trail effect

Improve this answer
\$\endgroup\$
1
\$\begingroup\$

This is how I think It could be done.
1. Set camera to bulb mode. So that the shutter stays open as long as the shutter button is pressed.
2. Set flash to fire either at the first / second curtain.
3. Press the shutter button when the people start to move the light source and release it once the movement is complete

As you realize, I've not specified what ISO or Aperture is needed... this is where you need to experiment.

Improve this answer
\$\endgroup\$
3
  • \$\begingroup\$ It is worth pointing out that second curtain wouldn't matter. We don't know if the phone started at the left or right. I'd actually kind of expect that it started where we see it in the flash. Second curtain only matters when the subject implies movement in a direction (like a car going down the street or a person running.) For something like this, it would look equally natural with a first or second curtain flash. \$\endgroup\$
    – AJ Henderson ♦
    Commented May 15, 2013 at 13:28
  • \$\begingroup\$ Yes, you are correct... \$\endgroup\$
    – Viv
    Commented May 15, 2013 at 13:33
  • \$\begingroup\$ This does not account for what appears, to me, to be background intruding over foreground at the top of the arc as shown in my crop. Or, are my eyes/brain deceiving me? \$\endgroup\$
    – Russell McMahon
    Commented May 20, 2013 at 15:54
1
\$\begingroup\$

This would seem to be a most UNlikely result if done "live"

Bacground over foreground and no apparent mixing even at boundaries.
This very strongly suggests it was done in post processing.

Also, the smooth radius change curve does not feel likely to be "natural". When swept in large arc the arm not only changes length as it moves but also the rate of change of length with position & times varies between locations as the arm geometry changes.

I have some swept arm arc time exposure of people waving "sparklers". I'll post one when/if they show up to allow comparison of arc shapes.

enter image description here

Sparkler arc - full 'circle' in this case.
More circular that I'd recalled but not the perfect arc of the photo above.
Also does not include any of the background hanging over the foreground as theirs does :-).
No post-processing additions in this one :-).
Chinese New Year, Ningbo, China:

enter image description here

Improve this answer
\$\endgroup\$
5
  • 2
    \$\begingroup\$ The background hanging over the foreground is not a sign of post production. It is a sign of the way exposure works. In a long exposure, the brightest and/or most constant light source is going to be the most exposed. For a long exposure, those lights in the background were there the entire time of the exposure. The phone would have only been there a small portion and is probably less intensely bright than the point light. Thus, it is expected that the light would shine through as it does in your crop if it was done in camera. \$\endgroup\$
    – AJ Henderson ♦
    Commented May 20, 2013 at 16:10
  • \$\begingroup\$ I voted down because of that error and because this does not answer the question. The question was not if the image was fake or not, but rather how to accomplish the look. Your answer is a (partially inaccurate) analysis of why it is done in post without any information on how to accomplish the look. \$\endgroup\$
    – AJ Henderson ♦
    Commented May 20, 2013 at 16:12
  • \$\begingroup\$ @AJHenderson - Your stated certainty in the accuracy of your analysis may be warranted, but whether such subsequently proves to be hubris or correctitude is as yet tbd. IF done in post then providing a description of some completely different effect that looks similar is equally "not an answer". I'm "acquainted with" the effects of mixing two sources of differing amplitudes, with one present spatially in a given area only for part of the exposure time. Over many years I've made images with eg ghosts (moving object passively lit), lanterns (moving source), moving strobes and more. ... \$\endgroup\$
    – Russell McMahon
    Commented May 21, 2013 at 14:22
  • \$\begingroup\$ @AJHenderson ... I chose that section as the most evident one but there are many variably uncertain examples throughout the image - some mix, while some over or underlay fully. It's not obvious that this range of variability matches your explanation. FWIW - even IF you are completely wrong your answer has some usefulness - so no down vote :-). \$\endgroup\$
    – Russell McMahon
    Commented May 21, 2013 at 14:22
  • \$\begingroup\$ Btw, I didn't mean to come off as being overly harsh, so I hope it wasn't taken that way. If you add some details about how the effect could be accomplished (either live or in post), I'd be happy to switch my vote. I also do agree that it was almost certainly done in post, I was just pointing out that the blending you pointed to appears to be correct given that the image is clipping (thus blending doesn't appear to occur as it is out of range). That's a minor point though compared to needing to answer how it would be done in order to answer the question. \$\endgroup\$
    – AJ Henderson ♦
    Commented May 21, 2013 at 14:27

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.