Really good post - you touched on some really good points that I personally fall victim of too. If I don't have the time to flesh out more I definitely agree that it's great habit to practice with keeping in mind the next stages in the pipeline after the concept.
I think this is really dependent on the studio and the phase of production. In the beginning you just want a basic visualization to sell the game to producers. Ortho is great for characters but when it comes to complex enviroments its not really feasible maybe something like a zoomed out shot then a close up and that's all that should be needed. Also consider the speed, the reason concept art is used is because it can be done (at least for now) way faster than 3d. In the end time is money and while its nice to have all the different views and materials down, a good 3d artist should be able to fill in the blanks.
Awesome thread idea. I'm not a particular good concept artist myself, so the most enjoyable work for me typically comes when I have clear, well defined concepts to work from.
I've always liked Martin Shapev aka dfacto's work in this regard and have worked with him on numerous occasions and its always a pleasure. He does a lot of 3d blockouts for his concept work, which is especially crucial for hard surface work (but good in general as well).
Here's an example:
Clean linework for the orthos but also nice rendering to give an idea of the materials, and detailed close up views of all the important angles. He made a 3d blockout so there was essentially zero perspective cheating (another big issue with loose concepts) and all the views are consistent.
So, here is another piece from Martin. This one does a much better job of portraying a specific atmosphere, a mood, and even telling a story. These are all great things for certain purposes, especially overall art direction. However, when it comes to asset creation or generally recreating this scene in the game this sort of concept tends to be a lot less helpful to, for instance, the artists tasked with modeling the robot.
Hmm. I get the point of the thread, but I think it's important to speak with clarity regarding when something fails as a concept - or simply isn't useful for a 3D artist to work from. There is a difference between concept art and illustration. There is also a difference between concept art meant to be modeled from, and concept art meant for any other purpose - as folks like EQ touched on. The first image posted is a fantastic set of character concepts, and communicate a fair amount given their looseness. They could be made better for 3D use with some detailed sketches of specific parts, or photo thumbs clarifying materials - certainly, but they definitely succeed as "concepts".
to me the purpose of concept art is to explore those design decisions in 2d before committing the bandwidth of making it in 3d.
Sure. But its purpose is also to explore design decisions in 2D before committing a significant amount of time making the final concept. I would actually be interested to know, most folks working regularly as concept artists - how much of the work is actually meant to be given to a 3D artist, and how much of it is for other purposes?
That nitpick having been made, I'll say I agree that things like the batgirl concept are great, and I find that sort of thing to be a pretty fantastic "happy medium" (full body character, additional notation and clarifying sketches/notes surrounding) between nice piece of art and practicality for turning them into usable assets.
Sure. But its purpose is also to explore design decisions in 2D before committing a significant amount of time making the final concept. I would actually be interested to know, most folks working regularly as concept artists - how much of the work is actually meant to be given to a 3D artist, and how much of it is for other purposes?
I think some 3d guys can understimate just how much work a turn around is. It is a labour intensive slow process that is very difficult. the concept i just posted is what i would call ideal for a 3d modeller, but that doesn't mean that concept art that polished is necessary for every production because it is so time consuming.
It's all about team work, if you have great modellers that can interperet and extrapolate details then wicked, make your concepts looser and work faster. It all depends on your team dynamic.
And yeah tons of work isn't intended for the modeller, as there is a lot of exploratory sketches.
Line drawings are great, I believe every concept artist should be able to at least do basic line drawings. However, line drawings will never sell a product, you need colors, textures and materials to convey a mood and a setting. If I was to model those drawings ( I am not very good at 3d) I would much rather have value and color than ortho.
Concept art =/= promotional art. Concept art, intended for asset creation, is not art that "sells" a product because that isn't what it's for. Promotional art is. It is absolutely important that people know the difference between the two.
What I see often in new artists - or illustrators that get hired to do concepts are just loose paintings or drawings. This is less useful to a 3d artist than a Concept that Answers Questions rather than creates questions.
I've encountered concept artists who have never drawn orthographic views for example. And I've had leads tell me it was my job to make it look better than the concept. Both of these ideals could be argued for or against. But in my opinion its the responsibility of the pre-production concept art phase to answer as many questions as possible before committing to production.
What I would look for in a concept portfolio is the ability to paint mood pieces and loose design iterations yes. But also the ability to communicate that final design using orthographics and again anticipating all questions and answering them with the art.
@Triksterr - strange as it may seem I have a lot of experience with modelling and I would prefer those lines over a painting. I would only go a little further by placing some values on it to break up the layering and shapes.
The drawings are good but It shows zero shape or form (what is in front of what, facial features high cheek bones shallow or deep eye's?) also I would have no idea what ethnicity the character is supposed to be.
Concept art =/= promotional art. Concept art, intended for asset creation, is not art that "sells" a product because that isn't what it's for. Promotional art is. It is absolutely important that people know the difference between the two.
I did not mean sell as in sell the final product I meant it as in sell an idea to a producer to get a green light on production.
What I would look for in a concept portfolio is the ability to paint mood pieces and loose design iterations yes. But also the ability to communicate that final design using orthographics and again anticipating all questions and answering them with the art.
Sure, but I'm just saying - I think you need to consider a variety of things before you outright declare that something fails "as a concept":
^This. This is simply inaccurate. It's a cool painting, and a nice concept - definitely useful in the design phase, helps establish mood and a variety of other style choices that can be chosen amongst and fleshed out if need be. "Not as useful for 3D as orthos", sure - depending on who the 3D artist is who needs to work with it, but that does not mean that it is a bad concept or that it has failed its purpose. ...I mean, I guess you could say Justin Sweet doesn't produce nice concepts, but I assume that is not your intention (also, notice in that link how other concepts flesh out the designs with more detail). Regarding this, I simply recommend a more specific choice of words.
I also think it's important to keep in mind that most concept artists do not just do concept art, and a significant portion of what is concept art is not meant to become a 1:1 3D, digital asset. I see it a ton (and it bugs the fuck out of me) where I see folks talking about what "concept art" portfolios should have in them, and it's mostly wanting orthos and images that are 50% notes and photographs. Yeah, that's all well and good for the 3D artist who gets a piece of it - I think knowing what 3D artists can benefit from in conceptual work is a great knowledge to have! But concept art is a hard thing to get in to as a singular job. Because of where it falls into the dev cycle there are many places that don't have full time "concept art" positions, and most folks I know who've worked them also sustain themselves off of other work (illustration, for a variety of purposes) - hence, why portfolios might not be full of orthos. Now, that's getting pretty off topic and I apologize for that. I won't dig into that any deeper, as I don't think your intention was to discuss the overall life blood of what concept art is useful in what ways.
Trying to steer myself back into more relevant "Useful for 3D Concept Creation" type discussion, I actually have a question onlooking 3D artists might care to answer:
Regarding character concepts, would you prefer one naturally posed, decently-rendered character with notations for clarification - or an orthographic linework-only image instead?
I can only assume this would depend fairly heavily on the artist and their skillset, but I'm curious to hear some responses and reasons why.
Can i just point out. Teryaki is a 3d guy mainly, two listen is a 2d guy.
As someone that has done both professionally. Teriyaki is correct that there just isn't the detail there on that js concept to be modeled. I have no idea about how that helmet on the left dude looks.
I really think the discussion should be more based on how to find a compromise of time vs efficiency.
Linework is fast, but it is heavily neglected by many because isn't as sexy.
...I really have to say, not every 2D concept is to be immediately modeled. I saw that first image as something fairly loose to start or pull the direction of a discussion. As mentioned previously, something akin to a mood piece, a start to base the shape language from or to tell you where these characters fit into the world. Then, once the actual direction is established, one would go in and make the clean lines/orthos/material breakdowns/etc to model from.
Like, yeah, that's a very loose concept that, if handed to me to model directly from, I'd probably feel was a bit bullshit-- but it's also not likely that it was intended for you to model from. More likely it's something that would come up when no one's really sure of a design direction in the first place. No one wants to spend ages doing nice clean linework when the design language/concept haven't been approved (unless their thing is nice clean linework, in which case they don't want to spend ages doing a painting but I digress). There's no point.
If we're pretending to think from an amateur concept artist perspective-- yeah, no, you can't think that your job will be doing that loose concept/illustration and moving on. You need to be able to make an informative concept, and usually nice orthos. But in the overall scheme of things, getting across the point, where it fits into the world, etc to the art director/designer will probably come easier with a loose painting that needs details filled in but looks pretty than a static ortho.
Going to throw my two cents in here really fast. As someone who likes to do linework concept art I'm going to actually argue for the photobashed quick painted style a bit here. When a concept artist is told to make a design there is almost never a point of time that we have the final idea of what the client wants in their head initially. So we make a ton of thumbnails, which we then select our favorite (and most probably what the client wants) from and take them into presentation. For a character this is usually 3 variations presented nicely on a page which is the standard Wes Burt style presentation.
Now the catch there is that the client will look at those and say "I like this about 1, this about 2 and this about 3 and I hate this about..." and thus begins the process all over. At any point of time that the client hits on the design he likes they will take it away from you and move you to something else. This happens a lot of time without our ability to clean it up or present it in such a way that the 3d can use correctly which is why in house a lot of concept art given to 3d artist is really poorly photobashed, doesn't have perspective or no detail explanations.
Here in lines where good art direction comes in. The better the 3d artist (and the more connected they are with their art director) the less the concept art has to explicitly spell out everything. Apparently at arena net there are seasoned 3D guys who can take what Daniel Dociu draws and directly translate it to 3d.
There in lies the problem though. Most 3D artist would be so lost to try to make a Daniel Dociu piece engine ready. Most 3D guys struggle with even well photobashed paintings. Clean and detailed line art is always the most explicate way to relay information regarding proportion and detail (not mood or lighting). That being said... line art takes about 4 times longer to do (way longer as the scope of the scene goes up.). And if your'e paying the designer by time, which is how most designers operate you're going to be draining your wallet really fast.
Maybe here is something you could add to the list, well it kinda enters in the "detail exploration" section, I don't know : concept artists should think in 3D, and think about how what they are drawing works. I've seen a lot of concepts that just couldn't work in real life, or in 3D. They put fancy shapes here and there because it's cool, they don't care about the articulations for example, or if that part of the mesh will go through another part if the character moves. How many times did you see in a fantasy game those fucking ultra large pauldron just going through the chest or worse, through the head, every time the character moves his arm ?
It applies to perspective and orthos too : sometimes they draw the thing from front and side, but when you try to do it in 3D you quickly see that it doesn't work, so the guy just wasted his time drawing useless orthos.
Last thing, it's about the style. Sometimes the style of the concept doesn't match at all the final style of the game, which can be a pain.
Just wanted to point out that it is very possible to hit both requirements - it just requires time, dedication, and of course leads/art directors who know what they are doing.
The work of Betty Jiang is a perfect example.
And it works for orthos too.
Ironically, in a dynamic team with good synergy there is actually no need to go that tight, as assets can bounce back and forth between 2d and 3d folks to reach a nice, polished result. But it requires the work of a good team, and sadly enough deadlines and arbitrary decisions by people not involved in the art process can put a dent in that.
I'll say that 3/4 drawings are about a million times less tedious and faster to draw than orthos too! That's because it's really hard to get the design looking right when drawing flat on for a character, and we are so in the habit of drawing in perspective.
Props like guns are however faster to draw in profile.
Pior. Those are great examples. But yeah... Not too useful in a production as that concept would take days to paint by itself. I could only see it being justified for main characters, not npc's!
edit:
of course the cyborg monster thing is more complex but point is, for me it's very tiring to visualize complex dimensional forms when just using lines. I imagine this applies to many more experienced dudes as well.
On the other hand pretty rendering doesn't have to be time consuming. And I'm pretty sure that some elements of faces on Pior's examples are photo sourced (or recycled from other concepts that this artist made) which saves even more time.
Oh definitely. That one of herman's is an extreme example. It's an impressively complex design and the man is an absolute beast at linework.
Having watched him draw in person many times, it's quite honestly intimidating.
That being said i don't know many people that could render to the level pf piors example very fast. Honestly the hardest thing about painting concept work is that it's very hard to define a "finished state". Line work very much has a finished state, as you work across the image everything is finished, where as painting is an infinite feedback loop of polishing.
Teriaki : I agree, the renderings you just posted are perfect, especially for a game like Borderlands. Not too much time was wasted on gimmicks and details, which in turn allows the modeler to only focus on what matters the most.
Unfortunately, for the last few years there has been a push for character models that can also hold up during realistic cinematics (AC, Uncharted, The Order) and a lot of production time ends up being spend on movie-style previz not really bringing much to the table in the grand scheme of things. It's a trap really : if the goal is to be 100% realistic, then every edge and every bevel needs to match a real world reference, thus making any art directed "modeling guidelines bible" impossible to put together. But I am getting the feeling that a shift will happen sooner or later, especially now that everyone can see that expensive AAA titles are less profitable than a mobile MOBA.
Also, just for the sake of clarification - without a doubt, I love the craftsmanship shown in BJ's renders and I think that focusing on this style is a very smart move on her end : it is definitely very marketable. But what I like the most about it is not so much the tight rendering, but much more so the fact that the images show a great understanding of surface qualities and materials. Artists with that kind of eye are able to "scale down" after a short adaptation period, embracing almost any other rendering style. And that's the golden ticket !
I think the most important thing in establishing the concept-to-finish process, is being aware of what the function of the concept is in your specific pipeline, and being aware of that on later stages.
Some concepts will be broader, more general, and will require your 3D artist to take design liberties (if he's a capable designer of course).
Some concepts will be very detailed and specific, and will allow no freedom in the creation process (not as fun for a 3D artist who's capable in design).
Both are viable and OK. As a 3D artist who also designs, I prefer the former personally.
The biggest problems, however, arise when your concept fits the former, but the creative lead enforces the latter. This is when the 3D artist will complain that he just doesn't have enough information to create an identical result.
Or the other way around, when the concept is very tight but the creative leads require further exploration in 3D, which is when your concept artist starts to get agitated and defensive.
I think more than anything, it's important to be aware of what the concept needs to convey and create it accordingly, and then be able to treat it relative to what it actually is.
Besides lineart, I think some simple shading to show the forms of surfaces helps a lot, especially on really tough shapes. So lineart concepts, with layers for reference like simple shadows, color/materials, and even shape guides/breakdowns is important for me. I often leave notes and "redlines" on separate layers for modelers so they don't have to try and figure out so much from my concept, I've already done that for them.
I'm going to post this again just becuase I keep seeing the assumption that line art is faster and some how quick. It really isn't. Not only does it take years of training and practice compared to photobashing but even the best line artist generate stuff very slowly compared to what a relatively new digital painter using photos can generate. This becomes exasterbated as you start doing stuff involving complex anatomy or scenes in perspective. If you look at someone like George Hull or Scott Robertson who does gorgeous line work they generate line art very slowly compared to someone like Eduardo Pena or Jonas De Ro who are extremely painterly and rely on photos for quick concepts.
Line art that looks fast and effortless is the result of generally decades of practicing and intense studies. If you look at some of the art from Sergi Brosa or John Park which is incredible line art from another artist perspective you can see the tens of thousands of hours of hard work they put in to make their art look effortless.
While with digital painting techniques I've watched people do brilliant renderings of concepts in as little as 8 hours with only about 6 months of study. Photobashing is incredibly powerful and saturates the industry because of its ability to be made by relatively new artist (who are generally cheaper) and its speed of creation. You can't wait for 2 weeks for a Syd Mead piece all the time.
Line work very much has a finished state, as you work across the image everything is finished, where as painting is an infinite feedback loop of polishing.
I dunno about that.
Back in day I used to do tons of line art designs.
And I would often find myself spending way too much time on cleaning and polishing the lines in order to make the art more precise and pretty. So it was no different than eternally polishing a painting.
If you know when to stop then you know when to stop, regardless of technique.
Greg Westphal: I'm talking about final concepts explicitly made for asset creation. I'm not talking about idea exploration, which is ideal to use quick idea-generating tools and explorative methods. Those two things are mutually exclusive and are different parts of the concept pipeline. We all should know that.
Photo Bashing still takes the same skill and time as good drawing skills, drawing being the foundation of all of it. Few do it well, and it's always very obvious when it's done wrong. For idea exploration, that doesn't matter, so do what works. For promotion and/or final asset creation, it's not always the best practice.
Now I'm repeating myself.
Point is, lineart doesn't have to be clean to be clear and precise. It just needs to have clarity, explain shapes and solve design problems. Different tools for different problems, use them all, and use them right.
teriyaki, you are asking 2d guys what's the best way to do concept art of course it's gonna resolve around technique haha.
Also Greg. Yeah nah buddy, Every video of Scott Robertson is him talking and going slow in a methodical way to demonstrate, and even then when he is sketching he is ridiculously quick.
I wouldn't want to hire an artist who uses photobashing as a crutch, as chances are if that's all they can do, they don't have very good draftsman ship skills.
I dunno about that.
Back in day I used to do tons of line art designs.
And I would often find myself spending way too much time on cleaning and polishing the lines in order to make the art more precise and pretty. So it was no different than eternally polishing a painting.
If you know when to stop then you know when to stop, regardless of technique.
Try drawing in pen for a week. That'll change your mind.
Unless It's a mood painting, don't leave your details a mushy mess for 3d artists to clean up. Unless they want it that way and your name is Ashley Wood.
Also when Ubisoft visited our class and showed us the work they were doing from blackflag very few of the pieces were line art and I after seeing most of the pieces from Last of Us I would probably have to figure it to be the same.
Since Muzz is being a bit fiesty and blowing me off flat out I'll say that most artist, even when they do line art, will use what we call a photo plate for proportion if its anatomy or perspective if its a painting (which Scott Robertson doesn't in his youtube vids which is why those are a bit more slow). I don't know of any artist who doesn't sketch as a way of getting the idea out but what you guys are linking is not sketches but cleaned up line art. If you want a good example of how fast someone sketches and then them translating it to line art I would check out the LevelUp with one of my favorites, Dave Rapoza. That guy is able to hash out a face extremely fast and even when he goes to do his line art its quick, but its still not as fast as if here were to just grab a photo and sample colors out of that really fast to block in form.
Once again its about ideation. I'm not a 3D guy anymore but when I was I loved orthos and 3 quarter views because it took the brainwork out of 3D. As a designer though I realize that my deadlines might be 15 concepted characters a week and If I have to ideate and haul ass through all 15 of those and leave time for revisions I'm going to be using photos even though I would prefer to do line art always.
Dave Rapoza himself went to lineart specifically so he could work faster and work on design...
I think you underestimate just how long those paintings take him.
I'll say that most artist, even when they do line art, will use what we call a photo plate for proportion if its anatomy or perspective if its a painting
Huh... I've very rarely seen that done by good draftsmen. It's just not necessary and it doesn't really provide much of a speed boost when drawing characters.
Using a 3d mockup for enviros on the other hand is super common and should be done.
The other side of this of course that not every production requires realism, and so photobashing falls on it's ass as soon as you need to do something stylized.
Dave Rapoza himself went to lineart specifically so he could work faster and work on design...
I think you underestimate just how long those paintings take him.
I understand how long it takes but Dave Rapoza wasn't doing photo bashing concepts. He was doing finished illustrations as final products, not rapid fire ideation. Line work almost always can produce a clearer design and when I have to produce something that uses a design language that no has ever seen before I think its best to draw it out. But most games and movies play it safe and base their language in real things.
You're playing the "gotcha card" and trying to trip me up on my words. I'm trying to bring my experience as a concept guy as to why a lot of the art out there is using photos and painted just enough to read as a design. If line art and clear rendering was easier and faster people would still produce the beautiful pieces of Ron Cobb and Doug Chiang but that just isn't the case anymore. Why have someone you're potentially paying a six figure salary for their ideas do orthographics or tedious work?
If line art and clear rendering was easier and faster people would still produce the beautiful pieces of Ron Cobb and Doug Chiang but that just isn't the case anymore.
I would say that just as much of it is about production benefits just as much as it is about being flashy and impressive to watch somebody throw together a painting in that way.
I think a lot of my problems with modern day concept art stem from a throw cool shit everywhere with dramatic lighting direction. There has been a bit of a loss of the attempt to make iconic designs that stand out and sell a brand. And there are wonderful examples of it being done recently, but it's not as common. This probably is the source of my aversion to photobashing more so than anything else admittedly.
So after the Ideation is done someone - probably a less senior artist - needs to tighten that up. Passing down an unclear concept can create some hardship for the rest of the team.
This is actually what is done a lot in studios that can afford it and actually the position that i have right now. Most of the senior guys are producing stuff to sell to producers. If it isn't flashy and dramatic it won't get funding to work on so the pieces they want to go forward are generally the ones that make the most impact.
Us junior guys use line art a lot and usually have to express individual pieces instead of a scene. If you get the chance to check out the Dragon Age Inquisition Art Book I would check it out. Most of the line art in there is ortho graphic and very quickly done probably by junior artist to help make sense of a a scene piece done by someone higher up.
It also has some extremely fast photobashed pieces in there in a flat perspective that you can tell was just a quick pass to sell the idea to the art director. A very seasoned 3D guy with a firm grasp on the art directors vision could still build it but it would be incredibly difficult. If you have the book I'm talking about the pieces on page 123 and 185. I didn't work on the project but I assume that those pieces were used internally to greenlight a part of the world to go into a deeper development.
I think you would be suprised just how prolific Herman was on that project. There is a lot of painting in there as well, to help sell specific concepts. While I'm not too big a fan of some of the designs i still maintain it as the single best collection of usable concept art i've ever seen.
Maybe it is a bit ridiculous to expect people to work to that standard because it is just so ridiculously high, but it does show just how much art a man can make if they are efficient at it.
It's just a shame that the 3d guys butchered his concepts .
![[Deleted User]](https://polycount.com/themes/polycount/design/default-avatar.gif){kind=avatar}
Replies
...And happy new year!
I've always liked Martin Shapev aka dfacto's work in this regard and have worked with him on numerous occasions and its always a pleasure. He does a lot of 3d blockouts for his concept work, which is especially crucial for hard surface work (but good in general as well).
Here's an example:
Clean linework for the orthos but also nice rendering to give an idea of the materials, and detailed close up views of all the important angles. He made a 3d blockout so there was essentially zero perspective cheating (another big issue with loose concepts) and all the views are consistent.
So, here is another piece from Martin. This one does a much better job of portraying a specific atmosphere, a mood, and even telling a story. These are all great things for certain purposes, especially overall art direction. However, when it comes to asset creation or generally recreating this scene in the game this sort of concept tends to be a lot less helpful to, for instance, the artists tasked with modeling the robot.
Sure. But its purpose is also to explore design decisions in 2D before committing a significant amount of time making the final concept. I would actually be interested to know, most folks working regularly as concept artists - how much of the work is actually meant to be given to a 3D artist, and how much of it is for other purposes?
That nitpick having been made, I'll say I agree that things like the batgirl concept are great, and I find that sort of thing to be a pretty fantastic "happy medium" (full body character, additional notation and clarifying sketches/notes surrounding) between nice piece of art and practicality for turning them into usable assets.
These are concepts from a guy i used to know called herman ng. Check out the entire thread here.
http://www.conceptart.org/forums/showthread.php/152372-RTS-game-Stormrise-concept-arts-%28update-10-02-2010%29
Lines are fast, clear and precise.
More concept artists should rely on them imo.
I think some 3d guys can understimate just how much work a turn around is. It is a labour intensive slow process that is very difficult. the concept i just posted is what i would call ideal for a 3d modeller, but that doesn't mean that concept art that polished is necessary for every production because it is so time consuming.
It's all about team work, if you have great modellers that can interperet and extrapolate details then wicked, make your concepts looser and work faster. It all depends on your team dynamic.
And yeah tons of work isn't intended for the modeller, as there is a lot of exploratory sketches.
Line drawings are great, I believe every concept artist should be able to at least do basic line drawings. However, line drawings will never sell a product, you need colors, textures and materials to convey a mood and a setting. If I was to model those drawings ( I am not very good at 3d) I would much rather have value and color than ortho.
The drawings are good but It shows zero shape or form (what is in front of what, facial features high cheek bones shallow or deep eye's?) also I would have no idea what ethnicity the character is supposed to be.
I did not mean sell as in sell the final product I meant it as in sell an idea to a producer to get a green light on production.
Sure, but I'm just saying - I think you need to consider a variety of things before you outright declare that something fails "as a concept":
^This. This is simply inaccurate. It's a cool painting, and a nice concept - definitely useful in the design phase, helps establish mood and a variety of other style choices that can be chosen amongst and fleshed out if need be. "Not as useful for 3D as orthos", sure - depending on who the 3D artist is who needs to work with it, but that does not mean that it is a bad concept or that it has failed its purpose. ...I mean, I guess you could say Justin Sweet doesn't produce nice concepts, but I assume that is not your intention (also, notice in that link how other concepts flesh out the designs with more detail). Regarding this, I simply recommend a more specific choice of words.
I also think it's important to keep in mind that most concept artists do not just do concept art, and a significant portion of what is concept art is not meant to become a 1:1 3D, digital asset. I see it a ton (and it bugs the fuck out of me) where I see folks talking about what "concept art" portfolios should have in them, and it's mostly wanting orthos and images that are 50% notes and photographs. Yeah, that's all well and good for the 3D artist who gets a piece of it - I think knowing what 3D artists can benefit from in conceptual work is a great knowledge to have! But concept art is a hard thing to get in to as a singular job. Because of where it falls into the dev cycle there are many places that don't have full time "concept art" positions, and most folks I know who've worked them also sustain themselves off of other work (illustration, for a variety of purposes) - hence, why portfolios might not be full of orthos. Now, that's getting pretty off topic and I apologize for that. I won't dig into that any deeper, as I don't think your intention was to discuss the overall life blood of what concept art is useful in what ways.
Trying to steer myself back into more relevant "Useful for 3D Concept Creation" type discussion, I actually have a question onlooking 3D artists might care to answer:
Regarding character concepts, would you prefer one naturally posed, decently-rendered character with notations for clarification - or an orthographic linework-only image instead?
I can only assume this would depend fairly heavily on the artist and their skillset, but I'm curious to hear some responses and reasons why.
As someone that has done both professionally. Teriyaki is correct that there just isn't the detail there on that js concept to be modeled. I have no idea about how that helmet on the left dude looks.
I really think the discussion should be more based on how to find a compromise of time vs efficiency.
Linework is fast, but it is heavily neglected by many because isn't as sexy.
Like, yeah, that's a very loose concept that, if handed to me to model directly from, I'd probably feel was a bit bullshit-- but it's also not likely that it was intended for you to model from. More likely it's something that would come up when no one's really sure of a design direction in the first place. No one wants to spend ages doing nice clean linework when the design language/concept haven't been approved (unless their thing is nice clean linework, in which case they don't want to spend ages doing a painting but I digress). There's no point.
If we're pretending to think from an amateur concept artist perspective-- yeah, no, you can't think that your job will be doing that loose concept/illustration and moving on. You need to be able to make an informative concept, and usually nice orthos. But in the overall scheme of things, getting across the point, where it fits into the world, etc to the art director/designer will probably come easier with a loose painting that needs details filled in but looks pretty than a static ortho.
http://howtonotsuckatgamedesign.com/2014/02/lets-get-real-concept-art/
It basically says whats on the thread already but adds a bit of things.
In 3d guys opinion, what view is more preferred: front/back or 3/4?
For orthos views, are side views generally a necessity or a luxury?
Now the catch there is that the client will look at those and say "I like this about 1, this about 2 and this about 3 and I hate this about..." and thus begins the process all over. At any point of time that the client hits on the design he likes they will take it away from you and move you to something else. This happens a lot of time without our ability to clean it up or present it in such a way that the 3d can use correctly which is why in house a lot of concept art given to 3d artist is really poorly photobashed, doesn't have perspective or no detail explanations.
Here in lines where good art direction comes in. The better the 3d artist (and the more connected they are with their art director) the less the concept art has to explicitly spell out everything. Apparently at arena net there are seasoned 3D guys who can take what Daniel Dociu draws and directly translate it to 3d.
There in lies the problem though. Most 3D artist would be so lost to try to make a Daniel Dociu piece engine ready. Most 3D guys struggle with even well photobashed paintings. Clean and detailed line art is always the most explicate way to relay information regarding proportion and detail (not mood or lighting). That being said... line art takes about 4 times longer to do (way longer as the scope of the scene goes up.). And if your'e paying the designer by time, which is how most designers operate you're going to be draining your wallet really fast.
Maybe here is something you could add to the list, well it kinda enters in the "detail exploration" section, I don't know : concept artists should think in 3D, and think about how what they are drawing works. I've seen a lot of concepts that just couldn't work in real life, or in 3D. They put fancy shapes here and there because it's cool, they don't care about the articulations for example, or if that part of the mesh will go through another part if the character moves. How many times did you see in a fantasy game those fucking ultra large pauldron just going through the chest or worse, through the head, every time the character moves his arm ?
It applies to perspective and orthos too : sometimes they draw the thing from front and side, but when you try to do it in 3D you quickly see that it doesn't work, so the guy just wasted his time drawing useless orthos.
Last thing, it's about the style. Sometimes the style of the concept doesn't match at all the final style of the game, which can be a pain.
The work of Betty Jiang is a perfect example.
And it works for orthos too.
Ironically, in a dynamic team with good synergy there is actually no need to go that tight, as assets can bounce back and forth between 2d and 3d folks to reach a nice, polished result. But it requires the work of a good team, and sadly enough deadlines and arbitrary decisions by people not involved in the art process can put a dent in that.
Props like guns are however faster to draw in profile.
Pior. Those are great examples. But yeah... Not too useful in a production as that concept would take days to paint by itself. I could only see it being justified for main characters, not npc's!
To me Pior's examples feel 10 times less time consuming to make than this:
http://i.imgur.com/u8qWTjH.jpg
edit:
of course the cyborg monster thing is more complex but point is, for me it's very tiring to visualize complex dimensional forms when just using lines. I imagine this applies to many more experienced dudes as well.
On the other hand pretty rendering doesn't have to be time consuming. And I'm pretty sure that some elements of faces on Pior's examples are photo sourced (or recycled from other concepts that this artist made) which saves even more time.
Having watched him draw in person many times, it's quite honestly intimidating.
That being said i don't know many people that could render to the level pf piors example very fast. Honestly the hardest thing about painting concept work is that it's very hard to define a "finished state". Line work very much has a finished state, as you work across the image everything is finished, where as painting is an infinite feedback loop of polishing.
Unfortunately, for the last few years there has been a push for character models that can also hold up during realistic cinematics (AC, Uncharted, The Order) and a lot of production time ends up being spend on movie-style previz not really bringing much to the table in the grand scheme of things. It's a trap really : if the goal is to be 100% realistic, then every edge and every bevel needs to match a real world reference, thus making any art directed "modeling guidelines bible" impossible to put together. But I am getting the feeling that a shift will happen sooner or later, especially now that everyone can see that expensive AAA titles are less profitable than a mobile MOBA.
Also, just for the sake of clarification - without a doubt, I love the craftsmanship shown in BJ's renders and I think that focusing on this style is a very smart move on her end : it is definitely very marketable. But what I like the most about it is not so much the tight rendering, but much more so the fact that the images show a great understanding of surface qualities and materials. Artists with that kind of eye are able to "scale down" after a short adaptation period, embracing almost any other rendering style. And that's the golden ticket !
Some concepts will be broader, more general, and will require your 3D artist to take design liberties (if he's a capable designer of course).
Some concepts will be very detailed and specific, and will allow no freedom in the creation process (not as fun for a 3D artist who's capable in design).
Both are viable and OK. As a 3D artist who also designs, I prefer the former personally.
The biggest problems, however, arise when your concept fits the former, but the creative lead enforces the latter. This is when the 3D artist will complain that he just doesn't have enough information to create an identical result.
Or the other way around, when the concept is very tight but the creative leads require further exploration in 3D, which is when your concept artist starts to get agitated and defensive.
I think more than anything, it's important to be aware of what the concept needs to convey and create it accordingly, and then be able to treat it relative to what it actually is.
Line art that looks fast and effortless is the result of generally decades of practicing and intense studies. If you look at some of the art from Sergi Brosa or John Park which is incredible line art from another artist perspective you can see the tens of thousands of hours of hard work they put in to make their art look effortless.
While with digital painting techniques I've watched people do brilliant renderings of concepts in as little as 8 hours with only about 6 months of study. Photobashing is incredibly powerful and saturates the industry because of its ability to be made by relatively new artist (who are generally cheaper) and its speed of creation. You can't wait for 2 weeks for a Syd Mead piece all the time.
Well, Aaron Beck has made all kinds of character concepts with lots of photo bashing and paintovers:
I dunno about that.
Back in day I used to do tons of line art designs.
And I would often find myself spending way too much time on cleaning and polishing the lines in order to make the art more precise and pretty. So it was no different than eternally polishing a painting.
If you know when to stop then you know when to stop, regardless of technique.
Definitely. Doug Chiang does amazing job at this using markers
Photo Bashing still takes the same skill and time as good drawing skills, drawing being the foundation of all of it. Few do it well, and it's always very obvious when it's done wrong. For idea exploration, that doesn't matter, so do what works. For promotion and/or final asset creation, it's not always the best practice.
Now I'm repeating myself.
Point is, lineart doesn't have to be clean to be clear and precise. It just needs to have clarity, explain shapes and solve design problems. Different tools for different problems, use them all, and use them right.
Also Greg. Yeah nah buddy, Every video of Scott Robertson is him talking and going slow in a methodical way to demonstrate, and even then when he is sketching he is ridiculously quick.
I wouldn't want to hire an artist who uses photobashing as a crutch, as chances are if that's all they can do, they don't have very good draftsman ship skills.
Try drawing in pen for a week. That'll change your mind.
Unless It's a mood painting, don't leave your details a mushy mess for 3d artists to clean up. Unless they want it that way and your name is Ashley Wood.
Eduardo Pena is a great example of photobash. I watched him do this concept in about 4 hours while talking to our group.
http://www.coloredcg.com/wp-content/uploads/2014/04/winter-hunter-ETBW-concept-art-by-eduardo-pena.jpg
Probably one of my favorite artist is Anthony Jones.
http://img.gawkerassets.com/img/181649rn42syijpg/original.jpg
Also when Ubisoft visited our class and showed us the work they were doing from blackflag very few of the pieces were line art and I after seeing most of the pieces from Last of Us I would probably have to figure it to be the same.
Since Muzz is being a bit fiesty and blowing me off flat out I'll say that most artist, even when they do line art, will use what we call a photo plate for proportion if its anatomy or perspective if its a painting (which Scott Robertson doesn't in his youtube vids which is why those are a bit more slow). I don't know of any artist who doesn't sketch as a way of getting the idea out but what you guys are linking is not sketches but cleaned up line art. If you want a good example of how fast someone sketches and then them translating it to line art I would check out the LevelUp with one of my favorites, Dave Rapoza. That guy is able to hash out a face extremely fast and even when he goes to do his line art its quick, but its still not as fast as if here were to just grab a photo and sample colors out of that really fast to block in form.
Once again its about ideation. I'm not a 3D guy anymore but when I was I loved orthos and 3 quarter views because it took the brainwork out of 3D. As a designer though I realize that my deadlines might be 15 concepted characters a week and If I have to ideate and haul ass through all 15 of those and leave time for revisions I'm going to be using photos even though I would prefer to do line art always.
I think you underestimate just how long those paintings take him.
Huh... I've very rarely seen that done by good draftsmen. It's just not necessary and it doesn't really provide much of a speed boost when drawing characters.
Using a 3d mockup for enviros on the other hand is super common and should be done.
The other side of this of course that not every production requires realism, and so photobashing falls on it's ass as soon as you need to do something stylized.
I understand how long it takes but Dave Rapoza wasn't doing photo bashing concepts. He was doing finished illustrations as final products, not rapid fire ideation. Line work almost always can produce a clearer design and when I have to produce something that uses a design language that no has ever seen before I think its best to draw it out. But most games and movies play it safe and base their language in real things.
You're playing the "gotcha card" and trying to trip me up on my words. I'm trying to bring my experience as a concept guy as to why a lot of the art out there is using photos and painted just enough to read as a design. If line art and clear rendering was easier and faster people would still produce the beautiful pieces of Ron Cobb and Doug Chiang but that just isn't the case anymore. Why have someone you're potentially paying a six figure salary for their ideas do orthographics or tedious work?
I would say that just as much of it is about production benefits just as much as it is about being flashy and impressive to watch somebody throw together a painting in that way.
I think a lot of my problems with modern day concept art stem from a throw cool shit everywhere with dramatic lighting direction. There has been a bit of a loss of the attempt to make iconic designs that stand out and sell a brand. And there are wonderful examples of it being done recently, but it's not as common. This probably is the source of my aversion to photobashing more so than anything else admittedly.
This is actually what is done a lot in studios that can afford it and actually the position that i have right now. Most of the senior guys are producing stuff to sell to producers. If it isn't flashy and dramatic it won't get funding to work on so the pieces they want to go forward are generally the ones that make the most impact.
Us junior guys use line art a lot and usually have to express individual pieces instead of a scene. If you get the chance to check out the Dragon Age Inquisition Art Book I would check it out. Most of the line art in there is ortho graphic and very quickly done probably by junior artist to help make sense of a a scene piece done by someone higher up.
It also has some extremely fast photobashed pieces in there in a flat perspective that you can tell was just a quick pass to sell the idea to the art director. A very seasoned 3D guy with a firm grasp on the art directors vision could still build it but it would be incredibly difficult. If you have the book I'm talking about the pieces on page 123 and 185. I didn't work on the project but I assume that those pieces were used internally to greenlight a part of the world to go into a deeper development.
http://www.conceptart.org/forums/showthread.php/152372-RTS-game-Stormrise-concept-arts-%28update-10-02-2010%29
I think you would be suprised just how prolific Herman was on that project. There is a lot of painting in there as well, to help sell specific concepts. While I'm not too big a fan of some of the designs i still maintain it as the single best collection of usable concept art i've ever seen.
Maybe it is a bit ridiculous to expect people to work to that standard because it is just so ridiculously high, but it does show just how much art a man can make if they are efficient at it.
It's just a shame that the 3d guys butchered his concepts