Is there photo editing software that will do this....


I want to take 3 separate photographs, each containing, basically, a head shot of a single individual each. Each of those 3 separate photographs are of different individuals. And I want to combine those three photographs into a single photo of a single individual. Basically trying to “create” a photo of what these 3 individuals, if they were combined into 1 person, would look like.

Does something like that exist?




Man, morphing is so 90s.


Maybe run this with each of the three pairs of faces and then combine the two that are most distinct again to get a 3 way (hehe)? Or search for combining face photos and try to find one of these sites that does 3 at once instead of just two.


There are tutorials for Photoshop techniques that a Google search turns up. Might be something to look at, although it looks pretty involved; it’s not like there’s a ‘blend images’ button you can just press (as far as I know, anyway).

#5 ?


If youre limited to merging 2 at a time, you can do some math to get to 1/3 x 3 from repeated pairs. But…that works for math, but probably not for whatever their image algorithm is.

(Uh, I guess just A/B + B/C + A/C)


Thanks for the suggestions. I think one or both of these might be perfect for what I’m looking for.




If you’d dare to share the results, I’m curious as hell.


I have to scan the original photos first, but I will definitely update when I get this done. Probably this weekend sometime.



Ok, so I used both of the sites mentioned above, and I was not really happy with either result. In both cases, it looks as though all they did was overlay the three photos and then played with some transparency settings. I was really looking for more of a photo that looked as though it were just a regular photo of one person.

Both sites cropped the photos (with my help) to show only the heads.

Here are the three photos I uploaded to both sites:

Here are what the sites gave back:



If I had to pick, I’d choose the Moonjee one, but I’m not real thrilled with either.



For hilarity’s sake, I’d say you should feed all the faces into a 3d facial recognition program (e.g. Mass Effect, NBA 2K17), and then just average each of the values that it uses.


back in the dos days when face morphing was new and took a loonng time to process, you had to map some of the meshes manually.


I concur with the Moonjee one, but yeah, I can see how you wouldn’t be happy with either of those results.

If there is better software out there that will morph only two images, could you do something like what @CLWheeljack suggested, or maybe:

Then use that finished image, and then combine that with C.
So, basically (A/B + C)


Well, the problem there is you get 25% A, 25%B, 50% C. Merging in multiple passes normalizes the proportions, but the way these morphers seem to work, with just transparency layers, it may not matter much.

You could probably rotate and scale to get all 3 faces aligned even with those underwhelming tools, to get a slightly better result.


Or you could get them all to donate some sperm and somehow (I’m not a doctor) make kids, and then wait 18 years and take a picture of their kid.