Open

by damned truths

submit your photo


Hall of Fame
View past winners from this year

Please participate in Meta
and help us grow.

Take the 2-minute tour ×
Photography Stack Exchange is a question and answer site for professional, enthusiast and amateur photographers. It's 100% free, no registration required.

I just saw this image on the deviantart. Does anybody know how I could shot a photo like this? From what the artist says in the comments, it seems that it has been created using only one single shot (not so many frames). My guess is that this has been created using Adobe After Effects. Does anybody have any experience creating anything like this? Could you direct me to any step-by-step tutorial?

enter image description here

Update: the original image found here:

enter image description here

share|improve this question
3  
It's a gif file, most likely it's a burst of photos taken after each other and then put together in post processing showing the movement(changes). –  fluf Mar 8 '12 at 12:49
    
Couldn't the photo just be panned and zoomed? That's what it looks like to me. –  Matt Giesbrecht Mar 8 '12 at 13:48
1  
@Matt A pan and zoom of a static photo wouldn't give the perspective shift you see here - the store windows behind the model in the first frame are fully visible in the last frame. I think fluf's suggestion is most likely. –  j-g-faustus Mar 8 '12 at 14:59
1  
The artist does say that it's only one shot in the forum thread. Given that, I assume the store window and flickering light to the right is drawn / cloned in. –  mattdm Mar 8 '12 at 15:37
3  
@fluf — I think it's done using the sophisticated 2D-to-3D tools in either Photoshop CS Extended or After Effects, or both. See youtube.com/watch?v=nDrr5kMl96k for example. –  mattdm Mar 8 '12 at 17:32

2 Answers 2

The file is a 1.5 MB GIF file using LZW compression and contains 84 images.

To say that it is made from a "single shot" is misleading at best.
You can clearly see 3D relative motion of the model against the background.
In a "single shot" all the information that is visible in only some of the images would not be available at all. So the creator has either used multiple photos, which violates the "single shot" claim, or they have used a separate background and moved the model relative to the background in each frame. In that case the picture of the model may arguably be a "single shot", but the background is also a "shot" so the claim is really not true in any sensible meaning of the term.

Even the claim for the model to be a "single shot" seems false as parts of the model appear to move relative to each other. In the two frames below the models arm is more covered in the left hand (later) frame than in the right hand one. This covering of the arm and hand occurs as the point of view moves from left to right.

enter image description here

The quality is so low that it is hard to see exactly what is happening but if you watch the GIF sequence repeatedly you will see that the costume and models arm have a 3D relationship that changes. Again, it would be possible for the "artist" to have added the costume separately and to have moved it between frames, but this again brings into question what a "shot" is. If the costume and the backgrounds are not considered to be 'shots" and have been added separately , then the claim MAY be considered to still possibly be true, but it is getting rather hard to believe.

It is possible that this was done by a program which endeavors to create a 3D effect by identifying various parts of the image that can be moved relative to each other. If so, this would explain the claim, but the amount of processing required would be large, and correctly determining what should move relative to something else would be "a challenge".


Re a comment on "suggesting foul play"

I did not mean to be be 'harsh" about what has been done or to suggest that it was improper in some way. I wouldn't have used as harsh a term as "foul play". More I'm saying that the very act of combining images, even if they have been derived from an original single one, and of adding material which must necessarily have been missing if a single image was used as source, somewhat stretches the meaning of "single shot". It's entirely possible that the creator added material that was hidden when the shot was made. Watching it replay repeatedly while looking at a selected area show many non-idealities which show how it may have been done. eg a piece of the dress by the right shoulder springs out substanially in the early part of the seuence. Watch the floor outside and behind the left leg and you see a sudden springing up of a shadow effect which does not occur gradually as would be the case with a real 3D pan. Neither of these effects would occur if this was a real series of shots and neither would probably happen if this sequence was completely automatically machine generated. This does not preclude use of manually directed automated tools. This all points to but does not guarantee that the source is a single frame which has had extensive infilling of originally hidden material and adding of effects along the way. An interesting exercise.

share|improve this answer
    
I think the last paragraph is the actual case. The amount of computing power on our desktops today is astounding, particularly with GPGPU-enabled processing. –  mattdm Mar 8 '12 at 17:34
4  
Everyone is so quick to call foul play these days, I see nothing there that couldn't have been done by hand from a single image given a little time, patience and expertise. I have see no real reason to doubt the creator's claims. The apparent movement of the model is just an artifact of the .gif compression and lens flare effects. –  Matt Grum Mar 8 '12 at 18:02
    
@Russell I have already given a link to the original image here with higher quality, if you want to examine it. It's 4.3 MB. –  Meysam Mar 8 '12 at 20:52
    
See addition to my answer re "foul play" - not what I meant to suggest. –  Russell McMahon Mar 9 '12 at 0:15
1  
@mattdm seems like this is the original image: http://psychiatrique.deviantart.com/art/Nastasia-Noir-133818038 –  Meysam Mar 9 '12 at 8:29

This is not that hard to do by hand but it is labour intensive. The process is roughly as follows:

  • Take the original still photograph and extract the subject (the model) to a new document
  • Fill in gap in the where the model used to be, by hand or using content aware fill.
  • Warp or texture map the now empty background image onto the inside of a cuboid
  • Add the model as flat image to the 3D representation of the room at the right depth.
  • Create a path for the camera in 3D.
  • Add lens flare and focus effects.
  • (Optional) render to low resolution .gif to hide the worst of your handiwork!
share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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