Head gen modeling apps like FaceGen and CrazyTalk??

melviso

So I noticed there are softwares like:
Face Gen:
https://www.youtube.com/watch?v=ZVGn7ljTHmY
Crazy Talk
https://www.youtube.com/watch?v=CPMkAqoLCAM

Are these software used in the cg industry? If no, why is this so? Is this because it wont be seen as original work. In a production environment, this would definitely speed up time with Zbrush or Blender used to add more details. I would like to think Studios have something like this as an inhouse tool. Also, is the topology of the computed mesh animation friendly?

poopipe

Facegen has/had a middleware license.
If Fallout3 didn't use it I'll be extremely surprised, there were a few other games out at around the same time that looked like they did too.

I built a character customisation system on it many years ago and I quite like it,  it's very clever and a lot more customisable than you'd expect given that most implementations never quite manage to get rid of the 'facegen'  look.

You can build your own meshes for facegen so whatever you need for your animation system can be done - it even handles  custom blendshapes

Disclaimer... 
It's been over a decade since I did this - I'm not sure how well it'd all hold up these days

melviso

What I am wondering is if this is something that is embraced in a production environment especially with time as a huge factor. These software are basically tools using some sort of calculation to fit a pre- modelled head as close as possible to a certain character's features sort of like ZWrap but instead of to a sculpt, 2d references are used. This might not work in all cases especially for none human characters.
If I remember, there was this artist that was very good at using one of these game character creators to make celebrities, e.t.c. Can't remember or I might be mistaken. But if someone is good at using these kind of character creators to replicate multiple characters' likeness. Is that something that the industry would frown at or will be seen as a skill in a production environment.
In future we might have AI doing this so I wonder.

Alex_J

Not an industry pro, but this is why I'm focusing efforts not on realistic characters anymore. By the time I git gud at it, large parts of it will be scan based, if not already. That's just my hobbyist-in-his-basement speculation, of course.

Not too many people are talking/maybe can't talk about how work goes in the AAA industry, but from what I know about business in general, if the necessity of skill can be reduced or eliminated, and if a job can be done cheaper via improved technology, it will be. I did read somewhere that COD or something liek that has built a new engine that is designed to work heavily from scanned data. You could say this is all taking the art out of the work, reducing skill, etc., and probably thats true. But you know, if I didn't have PBR tools I probably wouldn't be making realistic metal textures either. Not so soon anyway.

Good reason to focus on those soft skills that make you a force multiplier versus a specialist that will be out of work once technology renders you not quite so special. And also good reason to think of yourself as a leader, no matter your position, so that you can keep afloat in a job market that increasingly favors algorithms over human labor. Don't just be a pair of hands that do what the boss tells them to do. Be a brain and a personality that catalyzes the team into accomplishing things nobody imagined.

melviso

@Alex Javor
Exactly, u have raised some very important points bc I have always wondered why some artists frown at softwares like Poser, Daz3d. So what about scanned 3d characters then. Although, I think its important to always state that u used these softwares as they are procedural in nature. Take substance designer for example, u use these procedural tools to create unique looking textures. Now this is seen as an important skill to have in the cg industry. So think of an artist posting a photorealistic cg artwork of Angelina Jolie, generated using say a character generator with animation friendly topology with some sculpting in ZBrush to add details. Isn't that saving time like using scanned humans.

I remember one concept artist pointing out the need to deliver quality and at a good speed. Whatever gets you there doesn't matter because the client doesn't care what technique u used, how u got it done. What they care about is the result and time delivery.
Imagine if u are good at using a particular game character creator to create spot on likeness of celebrities like that artist I mentioned earlier,  ppl could offer to pay u to create characters in the game using their photo or a celebrity they liked. Game studios can easily make characters especially npcs easily with that person's skills. Directors and concept artist can play around with certain parameters until they are satisfied with the look they want for a certain character.

I remember one very good professional artist mentioning how AI is going to take his job in the coming years and I mentioned how AI would always need some level of guidance and there being a new job arena where some ppl would learn how to use these AI and would be in high demand due to their skills. So its wise to start learning how to be good using the new tech. Just like how substance designer and painter now have ppl who are experts in creating photorealistic textures using its procedural nature. 

I even see an app for creating non human characters coming out in the coming years. Time is money. Anyone now who can create photorealistc unique looking charcter from a generator app is going to be sort after because of the time it will save the studio but I am not sure why that isn't a thing. Scanning on the other hand is already replacing some areas of character 3d art.

Like in archviz, certain scenes require a good number of 3d human models. There are already great websites where scanned human models are sold for use in archviz scenes but there are times u might need custom looking characters. I think these softwares should not be overlooked but developed further. We would always need 3d character artist as well and I have the outmost respect for their craft.

Where the future leads can be scary especially with how fast technology is nowadays and keeping up with it surely will help keeping us artist afloat.
@Alex Javor
Wait..a new game engine that is designed to work heavily from scanned data? So this engine handles retopology automatically? Maybe they are using Zwrap?

Alex_J

https://www.pcgamesn.com/call-of-duty-modern-warfare/engine

I didn't dig deep into the details, just a cursory read. Not sure if this was the same article or not... I saw it linked on the Unity forums.

But yeah, my opinion is that the art in game art is going to become more specialized stewards of complex technology, which isn't something I'm super interested. You have a vision but the computer does the work. Maybe that sounds good to people who hate doing the UV's and figuring out new workflows and all that, but for me, the work is the magic. The work is what you do day in and day out. Results are tiny little blips that will be quickly forgotten. So I want to make sure the work I am doing is right for me.

melviso

Thanks for the link :-)

Yep, nothing beats the satisfaction when u put in the work and the results are on point. However, time has become a factor as the requirements for both games and films continue to climb higher. So learning how to be efficient is key and practice above everything else yields results.

I am currently learning how to sculpt photorealistic characters when I can manage time. I still feel it is important especially when u look at characters with distinct features like wings, scalely skin, scars, zombies, outworldly appearance e.t.c. Celebrities that may be too busy or occupied to come down for a scan, here a 3d character artist is needed. 

Alex_J

Yeah I dont the skill is not going to lose necessity very soon. There's been discussion about this before and I think most the  in-the-know artist weren't expecting realistic character artist to dissappear in the next 5 years. But I am thinking about what I'm gonna be doing in 30 years, you know. A ten year plan isn't good enough.

A lot of artist are just gonna love the struggle of developing that skill regardless of how marketable it is. For myself, I've got stronger interest in other skills so that coupled with the fact that technology is slowly reducing the need for such skilled artist makes me think my efforts will best be put elsewhere.