Fhotoscanned face into ready to use facial rig

onera

Hey folks,

I'm working over a project of mine in UE4.

Also, I have facial motion capture solution and photoscan skills
However, I have no knowledge at all in 3D modeling/texturing/rigging/uv

What will be the best approach to do it? What is the easiest way? Can the texture be directly applied on let's say MakeHuman head that has rig?

Or a reto to be done and then be rigged afterwords?

Please let me know the options I have

[SKETCHFAB]782f96aeff724336bb1f9f0194bc94ed[/SKETCHFAB]

[SKETCHFAB]866ba46b293b485f9076a8d88a6b51de[/SKETCHFAB]

leslievdb

with your limited knownledge its going to be harder but have a look at http://www.russian3dscanner.com/ wrapx
if you have an already rigged face you could use this software to blend it to your scan

onera

Thanks a lot @Ravenslayer

It still looks advanced to me in the matter, I will try to find a 3D guy to help me who is eager in photoscans and MoCap.

It's just too much work for single person

jfitch

I'd do it just for fun if I didn't have a million other things to do. You might try posting in unpaid work (unless you can pay?) and see if anyone is willing to lend their services. http://www.polycount.com/forum/forumdisplay.php?f=44

Jonathan85

1) BTW quite a nice scan, did you do it with only 1 camera?

2) Did someone other has taken the pictures while the person was sitting?

3) How many photos were used in the scan?

onera

@jfitch

If you want, we can try it. There's no time limit maybe a month or two.

@Jonathan85

1) Thanks, Yeah it's with a single camera.
2) Nope, only me! And had a nice actor who was willing to stay still
3) I've used around 35-45 as long as I remember

You guys think in the unpaid category I'll have luck to find someone?

jfitch

You'll have more luck if the animated head will be featured somewhere, but idk. Some people just really love doing that stuff and having a nice scan to work from would be enjoyable.

onera

Hmm, then I'll probably give it a try!

It just doesn't feel right to make someone work over it, it's more like cooperation.

I can photoscan and Motion Capture, and someone else to do the rest. As you told it, it should be fun.

I can't afford investing in that current hobby because there won't be any return.

You think in the unpaid section there's people having the skills to do it? Also if I have to pay someone, how much green wizards it will cost?

p.s
Sorry for the interrogation... But hard to figure it out by myself :P

Jonathan85

onera: Interesting, did the subject of scanning had a hard time standing/sitting still?

How long did it take you to shoot those aproximatelly 40 photos?

im asking all this because i aswell am planning to headscan but im not sure if the people will manage to sit still until i take all their photos...

onera

@Jonathan85

It took me around 40 minutes. The biggest problem is that the actor should watch in the lightning so it's not really cool for his/her eyes and can be an issue.

It's really important they don't move and watch at the same point, because the results can get really weird.

Biggest issue is to have lightning from even distance/strength around the subject, because you can notice only with one lightning source there's shadows.

Jonathan85

Wow... the actor was able to sit still for around 40 minutes :-)...? Is he really an actor or model or something that he is able to sit still for such a long period? I would have great problem mainting the same pose and facial expresion, and so does a lot of people i believe... is there some trick how you achiaved this?

onera

A little psychology... I guess? :P

I've challenged him with saying... "Not a lot of people are able to do it... so it will be remarkable if you do it."

And he gave his best, lol! That's my only trick,

Jonathan85

Lol, thats tricky and clever, i like it :-)!

Neverthless im quite worried my actors wont be able to do it, do you think its possible to do it more faster? I have hear in tutorialk via JefreyIanWilson that he was able to do it i 15 mniutes (or even 5-10 i dont remeber), you have used stativ i gues?
Neverthless really good result!

onera

Yeah, I use tripod also I do manual focus that's why it's taking more time.

In few days, I'll make a new photoscan with better lightning!

I'll report you with the progress. Makes me very happy that you like the results! Sadly there's no one to help me turn it into ready asset. =(

Jonathan85

Yes please you can report :-), you could also try to scan some other person and see how still can he/she sit

Also if you will be using a good lighting i believe you could try to shoot from hand and not from stativ/tripod, it will allow you much quicker shutter speeds (from 100 up), which might help if the other actor wont be able to sit still so long .

Also please what camera did you use? Nikon d3200 cannon something....?

try to look elsewhere on net, there has to be someone who has time and is willing to help.

onera

You wont believe me... But I've used 1/15 shutter, aperture on 4 and ISO on 100. 35mm 1.8 lens. I focus between the nose and the eyes as depth. And using Nikon d5100

I don't have enough money at the moment for better lightning, but that is not concerning me.

Next model will be another person and I'll shoot light outside with a single lightning source from in front of his face to kill shadows.

Will try to get a white bed sheet to reflect as much as white light as possible.

Jonathan85

wow interesting, you should probably btw use F8, it will get you greater depth of field, but since your original scan looks ok you probably even do not have to, but usually its adviced to shoot F8 (or even a little bit higher if you got good lense).

shutter 1/15 sounds great, you probably have a firm hand.

btw, your facial motion capture includes what? Faceshift?

onera

Yeah F8, sounds great, but... it's with my equipment I need to aim for the sweet spot. The closer the shot, the higher F is better, but my camera has 16mpx and best approach for me was f3.2 and fairly nice distance. It sacrifices from density of the texture, but it pays off through the workflow.

About the Facial MoCap, I plan to use Adobe After effect with marking points

Also I don't do the shots on hand, I use a tripod. In hand will never managed to do a sharp photo at 1/15 even with a still target. If the photo is not sharp enough, I retake it from the same position.

Jonathan85

Hmmm interesting (facila mocap via after effects with markers on face), did you do some research, how does it stand in comparison with faceshift? I believed that faceshift is the best way to go with a limited budget...?

onera

Yeah, Faceshift is great, but it's for stationary recording.

I want to do the facial and body motion capture together.

Jonathan85

hm.. what would prevent you making your own rig with the faceshift camera on some helmet? that would have to be a custom made solution but totally doable, or am i wrong and im missing something?

onera

It won't be wireless, I am not sure if you can power-up and record the faceshift camera with some DIY.

And a GoPro camera is completely wireless shooting at 60fps, pretty cute. Also it's light and small

Jonathan85

Interesting... what about some LONG LONG usb cable for the faceshift camera.

By the way you are a second person who told me in recent days that he will go for a go pro camera solution (and using maya i think for capturing/processing the facila mocap data) instead of using faceshift....

I will be also playing with these solutions in few months but i would prefer to use faceshift, if you already got some results and you dont mind, you could post them...? Because i think the faceshift solution has to be a lot superior.... (?)

Aabel

GoPro cameras can also sync together making the capture of 3d tracking data possible.

I did some facial mocap with ps3 eye cameras using ipisoft to sync them, results were ok, the resolution wasn't high enough to get good data though.

onera

I don't have any results with the Facial MoCap yet, because I don't have any good rigged model available.

Yeah, Faceshift is better, because it records in depth and has IR camera so you don't have to worry about errors from the lightning around.

However, I don't consider it "Lot superior" just better. But definitely not good enough to sacrifice the mobility of wireless Body + Face MoCap.

Because the project I'm working could take the work mostly outdoor than indoor.

@Aabbel

I'll be relaying on IMU sensors for the body. So I can focus mainly on the face.

onera

A new photoscan added

[SKETCHFAB]866ba46b293b485f9076a8d88a6b51de[/SKETCHFAB]