Specialize or Generalize?

salimmatta

Hey guys,
I will make this topic short and very straight forward. I have 5 years of work experience as game artist. I'd glad to say that I also have a solid pipeline skills (modelling / sculpting / retopo / texturing / Using UE4 efficiently / game optimization / Lighting and solid understand of VFX in UE4)

Id personally think that the industry is shifting into automation starting from quixel megascan covering almost everything in the environment and recently the metahuman introduction which basically will do the same thing with characters. Things are moving fast, as showing. Apparently game art is heading towards automation to an extend. 

I would like to hear some opinions regarding the matter if as game artists should specialize or generalize taking into consideration how the industry is quickly shifting.

Cheers

oglu

That really depends on the Studio you are working or like to work. In big studios specializing is key cause the pipelines look more and more like vfx. Everything is getting more complex and detailed in some Studios. But there are a lot other they dont follow that trend. Its up to you.

Be warned life may have other plans for you  

fdfxd2

Also depends on where you live, no use for a 3d game weapons artist in a country with no core game industry for example. Probably better to generalise in that case.

poopipe

Robots are not going to take your job.
Robots are going to enable you to keep up with increased demand for fidelity and the ever increasing scope of projects


in answer to the specialise vs generalise question..

If you mean within a specific discipline such as prop modelling then there's always a use for people who can do all the things and always a use for people who are really a good at specific things (eg. hardsurface or organics). 

If you mean in the traditional sense of an art generalist - that's a dead role. Studios used to hire them when technical art wasn't a real job 


Edit: 

Probably worth adding the caveat that I'm speaking from the perspective of someone who works in the UK, primarily at  medium to large AA/AAA  studios

sprunghunt

salimmatta said:

Hey guys,
I will make this topic short and very straight forward. I have 5 years of work experience as game artist. I'd glad to say that I also have a solid pipeline skills (modelling / sculpting / retopo / texturing / Using UE4 efficiently / game optimization / Lighting and solid understand of VFX in UE4)

Id personally think that the industry is shifting into automation starting from quixel megascan covering almost everything in the environment and recently the metahuman introduction which basically will do the same thing with characters. Things are moving fast, as showing. Apparently game art is heading towards automation to an extend. 

I would like to hear some opinions regarding the matter if as game artists should specialize or generalize taking into consideration how the industry is quickly shifting.

Cheers

3d scanning is not really automation. It's not faster to 3d scan a rock than it is for a skilled artist to make one in zbrush. It takes a while to process the images and you have to travel to find a suitable rock in the first place. 

But the amount of content required for realistic games is growing so fast with each generation that even with megascans and metahumans you'd still need a lot of extra art to fill in the gaps.  

But this is a trap beginners often fall into. You're not just trying to make a realistic environment or characters. Realism isn't really the goal. You're trying to make a game that sells well. You need something that stands out from the other games. Not to make something that just looks like you've used metahuman and megascans.

This industry attracts a lot of people who are very technically skilled. However what actually sells games is not just technical skills in art. It's artistic skill. Which is the ability to get people's attention using images. Libraries of pre-made art won't replace those skills. 

gnoop

Imo it's important to keep balance of those too things , too specialize and you could end up  laying off   because  you do it not quick enough and  there is too many people with same skill who agree to do it for lesser money.      The goal really is to do something unique only you understand how .

I personally survive  by focusing on challenging  and typically crappy looking  things in game environments . Often they need just a bit of  an old good artist taste , a subtle touch here and there, a sense of proportion of some kind  that help to hide something or bend  in a way  turning  an engine disadvantages  into like special feature.      Technical guys often  could be  somewhat blind for such  things .

It does require certain  understanding of 3d math  and general principles of how rendering works  so you could suggest an idea  to shader coder  for example.

I'v never worked for big triple A , never  with U4 engine   although.   So maybe no place for me there.

Blaizer

Imho, a specialist has usually more value in big studios, they usually do the job faster and better. A Character TD might be an example, and they are pretty needed.

I've been working as generalist for years, but i needed to specialize in a few disciplines. The more disciplines you cover, the better, but it's good to be specialized in a few of them, like modeling or texturing (-covering all, including mapping-, characters, props, scenery, etc.).

Too many years ago there was only 3 groups, modelers, animators and coders. But right now, we see that animators doesn't rig (at all).

Anyways, being a better artist is what really matters. The more you know, the better.

sharsein

There are right ways to specialize and right ways to generalize. If you specialize just by saying "no" to everything else, you may end up just generalizing in your discipline + procrastination. Specializing is more about paying attention to/developing your eye for detail. Pick an art direction you like for your work, and nail it. Set it aside and/or get critique. Go back to the work. Push it further. Rinse, repeat. Figure out a process that works for you to hit the target style. Repeat that process, figure out how to get faster.
If you generalize by trying a billion 3d programs that do the same thing or dabble in different disciplines "just because", you might get a little further than if you just stared into space wondering what to specialize in, but you won't get much further than that. Generalizing is more about problem solving. Need a character that plays the role of high dps in pvp battle tetris? Learn to design a character the communicates its gameplay role, is expressive, but doesn't take forever to create. Learn to design ui to figure out how a character would appear in a pvp tetris game. Learn to design animations and effects that communicate how its abilities are used. No need to dive into complex software like Ziva if a 8 frame sprite animation works just fine.

Alex_J

You need to know what your long term goals are.

If you specialize I think it's easier to point to specific peak in the distance and say, "that's as far as that route will go." Maybe it is everything you want, maybe not.

If you focus on understanding bigger picture and develop key soft skills you can go a lot higher as long as you willing to do extra work and take more risk.

Go to "know thyself." In a world of devalued humans I see specialization dying out eventually. Things come full circle when a person has to be good at universal problem solving skills again, just like we evolved to be.

gsokol

I've always been told to specialize, but my jobs have always been generalist roles lol.

I would recommend building a "T" shaped skillset.  Meaning, its good to specialize in something and go deep, but still generally keep a broad generalist knowledge.

pior

What about doing the one thing you like, and becoming good at it so that people who don't know how to do that thing or don't want to do that thing themselves, want to hire you to do that thing for them ?

Also, on the following :
" Things are moving fast, as showing"

I may sound like a bit of a contrarian but I would argue that some things are getting incredibly slow, actually. Just because the datasets we can now handle at runtime are growing larger and larger with greater and greater accuracy, doesn't mean that the tools to interact with (let alone create) these datasets get faster too. Even the Metahumans editor running as a streaming service (hence in theory leveraging powerful GPUs) is slow and sluggish ; and creating stylized content is getting slower and slower each year too as the singularity point for stylized art has been passed quite some time ago (in the sense of : stylized art requiring more care, detail and time than an equivalent real-life action figure, and now reaching the level of precision of animated CG feature films ... while also requiring many other extra steps for optimization).

Also, there is a certain level of marketing deception going on with game engines and art creation tools. The trailers may look great, but that doesn't prevent some workflows from being incredibly tedious. Looking at you, UE4 redirectors.

If anything I would say that the most hire-able people are the ones who are very aware of these things, developing workflows that bypass most of the common assumptions about game art ("model everything > retopo everything > bake everything > texture everything").

Meanwhile Nintendo will still laugh their way to the bank by relying on the lightest and fastest art creation workflows possible, and getting fantastic visuals out of them

Ruz

I hate to generalise, but maybe you should specialise

Domslice

I'd some of what has already been said:

- It can be largely dependent on where you are located, and closely related to which studios are in your immediate area. That said, remote work makes the point moot, perhaps especially so during Covid time.

- Becoming a "T" shaped person with a deep knowledge of one particular skill and supporting talents that branch out are true both with work and life.

- A further aside about "the big picture" would be to retain flexibility regarding your expectations as @oglu mentioned above, r.e. life's own plans for you!

Take it for what its worth, but anecdotally I can tell you that most of the people I had gone to school with who wanted nothing more than to model high-fidelity creatures, or badass hero characters, ended up generalists! I think the key would be to become proficient enough with one thing to get in the door, then prove you're versatile enough (your added value) to be kept on for future projects, even if the role isn't precisely what you wanted to stick with!