I got some requests, so I’ll try to explain how I put someone’s face onto someone else’s body without my final result looking super weird. Some examples of this are here: 1, 2, 3. My techniques might be useful to:
- put your fancast into that dress/suit of armor/scene
- make your OTP that never interacts be in the same scene together
- add that Lannister lion pendant/Tyrell rose necklace/Targ dragon pin to someone’s outfit
- basically merge 2, 3, 4+ images as seamlessly as possible
I use PS5. I don’t use a tablet or anything special, but what you’ll need is time and patience. Sometimes body swapping takes me as long as 16 hours. Sometimes it only takes me 2 hours. Every body swap is different, so this tutorial is more about tips and techniques than a step by step process. This got super long omg. If you could please like/reblog if this helped you, I would appreciate it. All of the manips in this tutorial are mine and PLEASE do not repost them or claim them as your own. They can all be found in my edits tag if you really want them on your blog.
1) SELECTING YOUR IMAGES
My favorite screencap gallery is grande-caps. This post lists some others, but usually when I’m looking for screencaps I just google “MOVIE/SHOW NAME screencaps.” You could also use promotional images from shows, or go to celebrity image galleries. Unless you are already very very familiar with all the images, you want to take your time looking through galleries before you choose.
Higher quality images are always better. Work with the highest quality (highest resolution, biggest) images you can find.
You cannot insert any image into any other image. You must select images where the perspective/camera angle is nearly the same. If the camera is near the ground looking up at the person from below in one image, and the camera is looking straight ahead in another image, I don’t know how to combine these images without freaky results.
It’s definitely not essential, but your photoshopping process is going to be easier when you select images where the person in the Inserted Image (ie young cat stark) is in a position similar to the person in the Background Image (ie older cat stark, up at the very top). In this case, michelle fairley is talking to someone in both images, and it’s a headshot in 3/4 view. Is your inserted image in profile, 3/4 view, dead on or ??? Whatever it is, it’s easier when the position of your background image matches the position of your inserted image.
It’s also not essential, but it’s easier when you have the shadows falling in the same place. Otherwise you might have to make your own shadows like I did here and that is hard. (More on that later.) Bonus points if you have images with similar coloring, lighting, brightness, etc.
BASICALLY THE SIMILAR YOUR 2+ IMAGES ARE, THE EASIER AND FASTER THE PHOTOSHOP PROCESS WILL BE.
Finally, when you’re selecting the images, THINK ABOUT THE HANDS. Hands reveal age. For example, Natalie Dormer is 31 years old. You cannot take a 16 year old’s face and put it on Dormer’s body while Dormer’s hands are in view. If you do, it will look weird!!! You can still put your fancast in Margaery’s dress, but you might want your final step of face swapping to be cropping the older person’s hands out of the picture.
2) HEAD SIZE
Usually at this point people just make a guess about head size and they cut out the head and paste it on, and they usually end up with a head that looks too small for the body. And that looks ugly. I use a different method that involves a lot less guesswork. It makes the heads on my manips come out the right size every time.
The Inserted Image:
The background image:
Copy the inserted image and paste it onto the background image.
Reduce the opacity of the inserted image, until you can see the eyes and nose and mouth of both images at once. I usually use around 50% opacity for this step. It’ll look like this ugly thing, except 1280px wide:
Move the inserted image over the background image. Line something up in both images, like have the chins line up, or one side of the face line up.
The person in your inserted image should have a bigger head than the person in your background image, because you’re going to shrink the inserted image to align with the background image. You want to shrink instead of enlarge whenever possible. Shrinking preserves quality, while enlarging reduces quality.
If necessary, reduce the size of the background image, while holding down the shift key to keep everything in proportion.
Reduce the size of the inserted image, while holding down the shift key to keep everything in proportion. You want to try to make the heads (about) the same size. Try to make the lips match up between the two pictures. Try to make the eyes match up between the two pictures. It doesnt have to be exact, but usually you want the eyes and nose and mouth to be in similar places. NOW THEIR HEADS ARE ABOUT THE SAME SIZE AND YOU DON’T HAVE A WEIRD TINY HEAD.
You want to try to cover up the person in the background image
3) LAYER MASKS - WHAT TO GET RID OF AND WHAT TO KEEP
Bring the inserted image back to 100% opacity.
Now we need to get rid of all the extra stuff around the inserted image.
DO NOT USE THE ERASER TOOL HERE. DON’T DO IT.
Instead you want to add a layer mask on the inserted image.
Select the layer of the inserted image. Hit the layer mask button at the bottom of the layers palette (). A white layer mask will appear next to the layer.
If you don’t know about layer masks, the basic gist of it is, you paint on the layer mask. The white parts of the layer mask show stuff on your layer. The black parts of the layer mask hide stuff on your layer. If you paint part of the layer mask grey, it shows stuff at reduced opacity, depending on how grey you paint it.
So how is a layer mask different from the eraser? Layer masks let you get part of your picture back later. Oh, you accidentally hid too much hair on the inserted image? You can just paint that part of the layer mask white again and you get it back!!
So use the layer mask to get rid of the parts of the image you don’t want. I start with a soft-edged brush and erase the edges of the inserted image. As I get closer to the face, I reduce my brush size. By the time, I’m really close to the face, I’m usually working with a brush size of around 3px. Sometimes I switch to a hard edged brush close to the face. The more precise you are and the more you take your time, the better it will look.
Something that you want to be thinking about here is, where can I put my seams? Blending the hair together is better than just using the inserted person’s face. Try to keep as much of the inserted person as possible.
For example, in this Sansa/Willas pic, I inserted Sansa’s whole upper body. (REMEMBER WHAT I SAID ABOUT HANDS!!! DON’T MIX OLDER HANDS WITH YOUNG PEOPLE FACES)
My seams are around the skirt of her dress.
I made a new layer and set the layer to overlay or soft light or color (you just have to experiment). I selected a purplish color and painted over the bottom of Helen MIrren’s dress in the background image to make the colors of the skirt match the bodice of Sansa’s dress.
There’s other ways that you can come up with creative seams that aren’t so obvious.
When I made this Lyanna Stark picture
I used Sansa from 1x01 as the background
but i kept the neckline of the dress from the inserted image
So what I’m trying to say here is, YOU DON’T NEED TO JUST INSERT THE OVAL THAT IS THE FACE INTO THE BACKGROUND IMAGE. Keep your fancast’s hair, keep your fancast’s neck, keep your fancast’s chest, keep your fancast’s whole upper body. If you put your seams in creative places where people aren’t expecting them, your image will look better.
Something else to keep in mind here is, avoid hard lines. If you have a perfect crisp line, it’s gonna look weird. There usually aren’t perfect, crisp straight lines in real screencaps. Things blur together a little. Let them blur a teeny tiny bit. Let it be soft and natural.
If you’re just inserting small things into an image, like a necklace or a belt or a pin or some embroidery, you can try using lighten or darken and those might do the work for you; it also works well if you reduce the opacity of the necklace or whatever to 85-95% so that it’s not standing out so much.
4) DUPLICATING STUFF
So, say you have that Lyanna Stark/Sansa image up there. If you look at the differences between the two images, Lyanna’s dress has an extra one of those little fabric ribbons hanging down on the right because I needed to cover up that piece of Sansa’s hair.
I also duplicated Winterfell’s chandeliers to get rid of some of Sansa’s head on the right.
You just wanna keep duplicating things and flipping them and rotating them until you cover stuff you don’t want to see.
If you look in the top left corner of the two Cat Stark images up at the top, you can see that Young Cat Stark on the bottom has a little bit more of the wall behind her visible. I duplicated that wall and moved it down to hide a piece of Older Cat Stark.
If you need to cover up stuff in your background image, look for stuff you can duplicate and move over.
Sometimes you’re gonna want to cover up something bigger:
I made a new layer and used the clone stamp tool to keep cloning the wall bricks to cover the Kingsguard knight.
5) SELECTIVE COLOR, BRIGHTNESS, SATURATION, VIBRANCE
If you look at these two images again
Young Cat is REALLY RED AND ORANGE and Older Cat is NOT RED AND ORANGE.
When your inserted image doesn’t have the same colors as your background image, you need to add new adjustment layers adjusting the colors. Clip the new adjustment layer to the Inserted Image Layer by clicking the little round symbol to the left of the eye on the bottom of the adjustment panel. Clipping it means it will only affect the Inserted Image Layer.
I usually add (and clip) adjustment layers reducing the vibrance, reducing the brightness, and / or reducing the saturation, but every image is different. In this case, I also add a selective color layer to reduce the amount of reds and yellows in young cat’s face. It doesn’t have to be perfect; you’re never going to get all the colors to match exactly, but the more you experiment and play with it, the better it will look.
REDUCE THE BRIGHTNESS OF YOUR INSERTED FACE. REDUCE THE SATURATION OF YOUR INSERTED FACE. SERIOUSLY.
So now I have this image and it’s nice but there isn’t a lot of shadow on the left side of Cat’s face
So I make a new layer and paint on it using the color #9f8e8c
you dont have to use that exact color, it’s just a color I sometimes use to make shadows. A good idea is to select a color from the face and use that color to make a shadow.
Set the painted layer to multiply and reduce the opacity to make the shadow. I reduced the opacity to 45%, but, again, it’s specific to the image every time.
You might need to add more shadows depending on your light source.
Sometimes it works well to make a new layer, paint a big black splotch on the layer, and then use the filter Gaussian Blur to make it look like a shadow.
For example, if you’re facing this nightmare of a screencap:
and you wanna put this image of Hugh Dancy in it:
the lighting on Hugh Dancy is all wrong, and it needs lots and lots of layers of shadows, mostly browns, set to multiply at varying opacities.
omg dont ask me to relive that, it was a nightmare, the lighting on both of those caps is a nightmare, here, this is the final thing, it’s looks a little weird because there was such a difference in lighting omg
In addition to the shadows above, there’s kind of a glowy effect around the people. I got this effect around Hugh Dancy by sampling the background color and then, on a new layer, painting lightly around him with a soft brush.
Sometimes you just want to paint an area black to make it darker and hide stuff. SHADOWS ARE YOUR FRIEND. THEY CONCEAL A LOT.
So, PAINT STUFF. PAINT. SERIOUSLY. IM NOT AN ARTIST AND I CAN’T DRAW, BUT I PAINTED OVER THIS STUFF TO GET SHADOWS AND GLOW AND YOU CAN TOO.
6) MOVING STUFF
Sometimes you really REALLY want to use a certain background image but you can’t make it work despite all of the above. That happened to me here:
I really REALLY wanted “Joanna’s” head on Sansa’s body, and I tried and tried but it wasn’t working. So eventually I gave up and got rid of Sansa. I moved the pink woman with her hands on the table back and merged her hair with the hair of the woman behind Sansa. I also copy/pasted some of the pillars. I had to use the blur tool to drag the pink woman’s cloak down. This was HARD and TIME CONSUMING but I REALLY WANTED IT.
So you have to get creative. If one thing isn’t working to merge your images, try something else.
um, I think that’s everything I do? I apply some coloring/textures at this stage and that’s about it.
soooo idk if this was helpful, it you thought it was helpful could you please like or reblog this post? Or come to my ask and tell me what you thought?
THANK YOU AND IM SORRY IF I EXPLAINED THINGS REALLY BADLY. If you want me to clarify something, or you have a question, just come ask me.