| Author | Thread |
|
|
07/19/2005 01:34:59 AM · #1 |
And yet there still aren't enough hours in the day.
Yep, this is a kiwiness ripoff, but I couldn't let him have all the fun. Plus, the overlapping mirror images were a pain to work with. Still, good times. |
|
|
|
07/19/2005 01:54:47 AM · #2 |
| Good Job! I like the way they are all doing something different. |
|
|
|
07/19/2005 02:13:43 AM · #3 |
| interesting how did you do it? |
|
|
|
07/19/2005 02:28:33 AM · #4 |
Nice job, Mark. Did you use the same procedure that kiwiness described on this image? I'd be interested in having someone write up a more comprehensive tutorial of how to do this. Kiwiness had mentioned doing so at one point, but I don't know if anything ever came of it.
Edit: fixed broken link
Message edited by author 2005-07-19 02:33:01. |
|
|
|
07/19/2005 05:04:42 AM · #5 |
Warning, long post to follow...
Originally posted by metoecus:
Nice job, Mark. Did you use the same procedure that kiwiness described on this image? |
Essentially, yes. I don't know what a vector mask is (I use GIMP for photo-editing), but skill with layer masks is essential. In fact, it is the only tool I used to achieve this effect. At no time will I actually be editting any pixel data--only transparency.
You have two choices when applying layer masks:
(1) Isolate the clone by placing the new layer on top and making everything except the clone transparent. I think this is the easiest method both conceptually and in practice since all you're doing is placing cutout characters on top of a background scene.
(2) Isolate the clone by placing the new layer beneath the current layer and making the portion of the background that lies directly above the clone transparent. This is not as nice conceptually, but sometimes necessary and the result is equivalent to (1).
The first thing to do is to take pictures of yourself (or your subject) in all of the desired positions. Use a tripod to make sure the perspective is exactly the same in every picture. Here are my raw materials:
The Cameraman
The Reader
The Sleeper
The Singer
The Writer
The Gamer
(Making stupid faces really seems to help the process...)
You'll want to use your camera's LCD to preview each picture in succession to make sure that you don't crowd yourself or end up sharing the same space.
Next, open up Photoshop or equivalent and open a new document with the same dimensions as your pictures. Drag each photo from your windows folder onto this document to load each one as a new layer. What will make this easier is to have the picture with the character furthest from the camera on the bottom layer and each succeding layer containing a character that is closer to the camera. You'll end up with a layer list like this:
Turn off the visibility (click on the eye) of all but the bottom two pictures--in my case, Cameraman and Reader (I like having a redundant Background layer so I can move the bottom picture in the future. The Background layer cannot be moved once it is created, nor can anything move below it.) You can set the layer opacity to less than 100% to make sure that the pictures line up right and there aren't two people taking up the same space or blocking each other or interacting in an anatomically incorrect way.
For this layer (and this layer only) I'm going to use method (2). By the time I took the second picture (Reader) the sun had moved such that all the room looked to be a different color from the rest of the pictures. I decided to make transparent the portion of Reader that covered me with the camera. To do this, go to the Layers diaogue box, right-click on the layer above the subject you want to reveal, and select "Add Layer Mask...." In the box that appears, select "White (full opacity)" to initialize the layer mask. You can also choose Layers -> Masks -> Apply Layer Mask... from the Image Window.
Now, since there is no one near the Cameraman, I don't have to be too careful about this layer mask. Select a brush tool with a large radius and paint over the character you wish to reveal with black. When working with layer masks, painting a section with black makes it transparent (alpha value = 0), white makes it opaque (alpha value = 1) and gray makes it semitransparent.
What we have so far:
+
----------------------------------------------------
The great thing about this layer is that the background is the same for all the pictures (blown out white) so if I make more than the portion of the picture covering me transparent, the borders will be invisible since they are the same to begin with. The tricky part will be when two images have characters that overlap.
Now, to use method (1) to insert the sleeper. Make the next layer (Sleeper) visible. Use your favorite method to select just the newest character and any relevant shadows, reflections, and other evidence of a changed environment (in this case, a deformed pillow and mattress).
With the selection made and feathered (5-10 pixels works well for me, experiment on your own), add a layer mask to the current visible layer (Sleeper). Make sure the layer mask is selected, return to the image, invert the selection (so everything but the new character is selected), activate the paint bucket tool, and paint the selection black. This will reveal all your previous work with a new person.
+
----------------------------------------------------
One problem, the sleeper's arm passes in front of the book when it looks like it should pass behind. No problem. Activate a brush tool with soft edges, select black, and paint the layer mask black where it should pass behind the book.
Ergo, (concordantly, vis-a-vis)
+
----------------------------------------------------
Rule: If an object belongs in front of another object, paint it's layer mask white. If it belongs behind, paint it black.
Fine tune the layer masks with small, soft brushes (for finely detailed borders) and gradients (for large areas like the wall behind the sleeper's and reader's heads) to make the borders of the various pieces of the picture invisible.
Repeat for all the other layers. If you're smart (unlike me) and the lighting conditions were the same for all the initial pictures, then you can flatten the image and post-process like it was an original photograph to prepare it for display. Otherwise, adjust each layer as needed, flatten, save, print, and astonish the world.
If you have further questions or if any of the above was unclear, feel free to ask.
Edit: You can never proof-read enough times...
Message edited by author 2005-07-19 05:08:49. |
|
|
|
07/19/2005 05:26:55 AM · #6 |
| Wow, nice write-up/tutorial!! Cheers, I may have to try this out sometime!! |
|
|
|
07/19/2005 05:30:12 AM · #7 |
These are fun to do - but I rarely do them! Everyone needs an image with a few of themself in it.. confuse your friends and family ;)
|
|
|
|
07/19/2005 06:57:49 AM · #8 |
|
|
|
07/19/2005 07:18:05 AM · #9 |
HA... I did one of those a while back. Not as nice as yours though. Maybe I should go and try another one.
 |
|
|
|
07/19/2005 07:22:54 AM · #10 |
Cool, and great tutorial..
One thing I would add is to use the same exposure for all shots, manual metered, or use P/Auto to get the settings then put those settings into manual mode.. It's much easier to make it all look like one photo if the exposures are the same!
I did these two in the gimp as well...
For a local camera club 'self portrait'.
And just messing about:
Message edited by author 2005-07-19 07:25:32.
|
|
|
|
07/19/2005 08:02:33 AM · #11 |
An awesome few I found a few weeks ago here at DPC from a 14 YO
|
|
Home -
Challenges -
Community -
League -
Photos -
Cameras -
Lenses -
Learn -
Help -
Terms of Use -
Privacy -
Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 09/12/2025 03:59:00 AM EDT.
