"Thumbnailing"
I've been looking into ways of getting past some difficulty getting started on some game art for learning/portfolio, as I get stuck on the fear of failure to have a "good" idea. I was curious about how many game artists utilize this process and how it fits into a typical pipeline when generating ideas, if it fits in. If not, try to explain how you use it for portfolio work. If you don't thumbnail, please explain your alternative process.
Some questions I have:
1. How much time do you put into thumbnails and do you use any photo source (cut and paste shapes) to generate ideas for those thumbnails?
2. Do you make modular thumbnails? Does this save time?
3. Do you often remix your thumbnails into new ideas? In what ways?
4. Do you save your thumbnails to use for future projects? Could you share some examples?
5. How much detail do you put into a thumbnail (e.g. silhouette, course detail, fine detail, etc)?
6. Is it faster/easier to do it on traditional media vs. digital media or vice versa?
7. Do you ever build on another artist's related or similar work as a platform for new thumbnails (i.e. copy, transform, combine)?
Some questions I have:
1. How much time do you put into thumbnails and do you use any photo source (cut and paste shapes) to generate ideas for those thumbnails?
2. Do you make modular thumbnails? Does this save time?
3. Do you often remix your thumbnails into new ideas? In what ways?
4. Do you save your thumbnails to use for future projects? Could you share some examples?
5. How much detail do you put into a thumbnail (e.g. silhouette, course detail, fine detail, etc)?
6. Is it faster/easier to do it on traditional media vs. digital media or vice versa?
7. Do you ever build on another artist's related or similar work as a platform for new thumbnails (i.e. copy, transform, combine)?
Replies
The thing with thumbnails is that you kinda have to do them fast or you get bogged down working on one. Turn and burn, keep moving and then once you've made enough you can pick and choose what parts you like and what you dont like. You have to fail over and over to get something good.
1. each thumbnail/silhouette shouldn't take very long, a couple minutes tops. I don't personally use it but I've seen some people use Alchemy which gives them some funky cool shapes sometimes.
2. Not sure what a 'modular thumbnail' is? Modular pieces are things that are duplicated over and over so I don't see how that could help.
3. Sometimes I take parts from one thing and combine it with a different one. Mixing and matching really helps get different ideas.
4. I don't necessarily save thumbnails but generally I have certain body types or overall shapes that I go back to.
5. This depends on what stage you're at. Myself and most others start at the silhouette stage. Basically just black filled, no details just the outside silhouette is important. Then when I get something I like, I fill in large details on a few. Then I select 2 or three I really like and work on smaller details
6. It really just depends on what you're comfortable using there's no real trick or advantage using either.
7. If there's a shape or something I like I might reuse it somewhere, or a body proportion I like too. But I don't rip and combine different pieces of other peoples work
http://www.ctrlpaint.com/imaginary-landscape-thumbnails/
2. No, if this is actually saving you time then you're not making thumbnails, you're making tight sketches.
3. Dunno what this means
4. Yes. You burn through a lot of thumbnails in a sitting, you end up with lots of pretty decent ideas that you don't use simply because they just aren't right for this project. Doesn't mean they're bad.
5. Again if you have detail in your thumbnail, it's not a thumbnail. This is why a lot of peoples thumbnails are basically just silhouettes. MAJOR ELEMENTS ONLY.
6. Purely in terms of output potential, digital is always faster for basically everything. In terms of the speeds individual artists work that will vary based on their knowledge/experience of each medium.
7. If you're talking about frankensteining other peoples art into cut and pasted abominations... I don't do this and would not recommend it.
All of your questions seem to revolve around saving time, the whole POINT of thumbnailing is to save time later. It's the planning stage. The more problems you can solve in the planning stage the less you spend later. As the old business addage goes: 1 hour of planning is worth 4 hours of doing.
Focus on IDEAS not technical merit, if the perspective is off, or something isn't well rendered, or a shape is a little wonky, it doesn't matter, make sure the ideas are sound, that as many obvious problems are solved before you move on.
KEEP IT LOOSE MAN.
I had a wohle thing typed up but I think its better to just link you to Paul Richards blog, he's been posting up his Darksiders 2 work for the last couple weeks and most of it is thumbnails for environments.
Most of an environment is single objects. When it comes to big picture things like composition and scene layout I think it's best to paint over a blockmesh map when the gameplay pace/flow has been established and you know the level plays well.
You want to be trying to make a level that plays well also look pretty, not trying to make a pretty level play well. That's just my opinion though.
3. What I'm asking is, do you break your old thumbnails into pieces that you use to create new thumbnails?
7. I don't mean literally, but use the coarse shapes (cut them out, fill with black) of other works and fill in your own fine details after. For instance, you find a really nice robot concept that has an element or two that you think would look really good with one of your ideas. You break it down to its simplest form (i.e. coarse shapes) and drop it in like a Lego set. What's wrong with that? Doesn't everyone build their art from others in some way? How would you make a fire hydrant without prior art?
I'm trying to figure out how to generate ideas quickly so that I don't have to feel anxious and lost on what I want to create.
I found a video that explains what I'm thinking of:
http://vimeo.com/45077843#at=0
7. Yeah, yeah, all art is derivative, nothing is original, etc., etc., etc. Deep thoughts. I assumed this was a practical question not a philosophical one, hence my answer.
I'm talking about creating a library of thumbnail parts similar to a greeble/nuts & bolts library, a collection of random pieces from many sources. A veritable box 'o parts, if you will. Some ideas are derived from other artists' ideas. Is this not done?
I'm not advocating this:
Anyway, thanks for the insight.
But having say... a library of arms that you just slap into your thumbnail every time you need an arm is a surefire way to get repetitive designs. Which is why modular type designs are usually reserved for less important characters in games, like zombies where you need a shit load of them, they all need to look fairly similar, but not like they're just cloned.
Basically a good shape is a fairly abstract, non descrip one.
also: alchemy = awesome
Thanks
hehe! how quaint!
Nice to see those again - it's been a while!
Some questions I have:
1. How much time do you put into thumbnails and do you use any photo source (cut and paste shapes) to generate ideas for those thumbnails?
It depends on what i'm working on. But I try and do my initial thumbnails within 3-5mins each to try and get as many ideas down in a short time. Then I go back and work on about 3 of the better ones a bit. Then I chose my favourite and work it up. I never use photos but imo its completely acceptable to use any and all media you like to get things going.
2. Do you make modular thumbnails? Does this save time?
I don't but its a good technique and if it gets you going. Do it.
3. Do you often remix your thumbnails into new ideas? In what ways?
Yes. When thumbnailing I might take 2 or more of the thumbs and overlay them to create new shapes.
4. Do you save your thumbnails to use for future projects? Could you share some examples?
I usually start with fresh thumbs as I like doing them and I think it helps me improve or try new techniques.
5. How much detail do you put into a thumbnail (e.g. silhouette, course detail, fine detail, etc)?
I start very blocky and refine from there. How much detail?... How long is a piece of string...Until i'm done with the piece.
6. Is it faster/easier to do it on traditional media vs. digital media or vice versa?
Well that depends on what tools, media you prefer. I do digital only because its faster/easier for me.
7. Do you ever build on another artist's related or similar work as a platform for new thumbnails (i.e. copy, transform, combine)?
No. But like question 1. Use whatever you can to get started. If you use other peoples work make sure none of their work is recognizable or is removed from the final piece. Otherwise get permission and credit the work.
Here's a piece I did the other day which shows my process from thumbnail to concept. This is my usual workflow.
1. 1024x1024 canvas
2. zoom out so the canvas is very small (about an 1.5X1.5 inches)
3. Use a hard edge square brush with size and opacity set to pen pressure
4. make shapes
4/5. duplicate and mirror (in the example below, I don't usually do this)
5. refine
6. refine
7. er...refine.
8. etc
and the finished piece is here: http://1.bp.blogspot.com/-7d5SsEaXgcg/UGIPWamKMKI/AAAAAAAABKU/AMFfkZlbDhk/s800/sentinel_01.jpg
and here's more thumbs I did not long ago using the same workflow. http://3.bp.blogspot.com/-MiD6Wm-fk_E/T_n-PIIAdLI/AAAAAAAABCU/WXQRzoZEaMw/s1600/Mechs.jpg
here's examples of awesome artists that do awesome thumbs.
Lemonhead: http://www.polycount.com/forum/showthread.php?t=61393&highlight=lemonhead
Wizo: http://www.polycount.com/forum/showthread.php?t=54182&highlight=wizo
Tacit Math: http://www.orderindebris.com/
dii: http://www.polycount.com/forum/showthread.php?t=87094
anyhoo. Hope that helps somehow.
In short. Thumbnailing is simple.
Start with huge shapes.
keep it lose and fast.
then refine, refine, refine.
In game design yes, there seems to be alot of blocking out of large shapes, and emphasized large forms as you say.
But the form of thumbnailing I use, is more focused on getting many designs or concepts out, just small of size and in less detail as opposed to a full concept piece - but not concentrating on huge shapes.
So I often thumbnail about 10 designs / concepts, which is pretty much me just quickly sketching out various ideas, on a project. And then picking 2-5 of them and taking them further, which sometimes involved taking ideas from other thumbnails and combining details or elements.
http://forums.gausswerks.com/viewtopic.php?f=3&t=202
[ame="
[ame="
[ame="
Also check out Creating Unique Environments with Scott Robertson. If you can. He uses a great technique of starting with very simple marker sketches then mixing them up to create new shapes. This technique can be used for anything not just environments. I use the same technique a lot but in digital form.
http://www.thegnomonworkshop.com/store/product/331/Creating-Unique-Environments#.UGVjvK5VUe4
@Ninjas, I like your point about making use of existing reference to inform your design rather than trying to pull it out of your ass. I agree with the concept of plausible design. I see so many things out there that just don't make sense visually and seem to be there just to make it "look cool". As far as I see it, if you're going to put something on a weapon, a piece of a costume, or a prop of some kind, that it should look like it has a real purpose for being there and not just to be greeble.
@Stinger88
FZD really shows how experimental you can be, but not focus on one shape and present a dozen subsets. However, I think you can still do those subsets for yourself and select the one you want to present.
I really like the responses I'm seeing here. The whole point of this thread was to see how many methods people use to generate ideas so that I could incorporate that into my own method to perhaps make it easier for me to generate ideas, not just for characters and creatures, but also for props, weapons, and static objects.
Simply put, it's not ideal or feasible, I think if you're designing an aircraft, taking into account size, compartment, etc, it will send you on a tech-chase for about a week or two before you even comeback to start working on it, especially if you can explore inside and out of the object in question.
Trust me, you don't want to go halfway either, since you will get a mushy middle of things, which just looks wrong on soo many levels, it's not even funny. It's not "OK" to take the images of google of Sci-Fi looking ship that was inspired by some movie, and then getting alot of the design wrong in terms of inner workings, loss of volume, etc.
For example, Jetpacks have always been a bad idea, yet that's one example of PD that people like and use alot because it 'possible', problem is Jetpacks are NOT possible, they'll burn your legs off, that's a preconceived case and even 'illusion' of "Plausible Design", they're a historic and culture thing that stuck with us for ages, just because something 'could' go in that direction doesn't mean it should or will.
Here's an example of 'jetpacks on feet' that actually does make sense, problem is limiting yourself to Google wouldn't work in those examples since when was the last time a mass media-culture knowledge of HPB existed?
I find it unfair to shoot thumbnailing out the window when all that thread is comparing it to is using Google to get simple designs, and most of the follow up designs are about how to get a scary monster with extra limbs in a DS or IPhone game, it's just tacky, not to mention very, VERY, dull. Everyone might as well create the same designs for everything, I mean it's not like anything else would make sense, so much for Art Bibles in a studio.
Also, lastly, Fun-Fact: Star Wars, with their 'floaty Droids' from the old Trilogy actually inspired the US army in creating the drones we see today, as well much of the smaller ship and other such designs today from the movies. Sometimes fiction is what drives the designs we see today.
I don't thumbnail much either, ESPECIALLY on the mechanical designs I've done, but I don't see it as being mutually exclusive to function whatsoever. The knowledge it takes to draw functional thumbnails (restricting yourself to certain shapes in certain areas, at its core) is the same knowledge it takes to draw functional designs in the first place.
Ace, I kinda agree with ninajs here. I dont see your post as being entirely bad, but what experience are you drawing from to make these judgements? Throwing out function just to get a good design is a bit like throwing the baby out with the bathwater. Good design is expected, functional or not. I guarantee I could come up with 25 interesting jetpack/flying suit designs that WON'T immolate the pilot if somebody hired me to!
In a broader sense I think the art in games is a vehicle for ideas, so it should either be communicating an idea or staying out of the way. In that sense, the actual art is not super important beyond just not having obvious flaws.
Sometimes strong silhouettes are useful in fast-paced gameplay (in helping things to be identifiable faster), but in other situations, I keep wondering why I should have a preference for 'good' forms over 'bad' forms when they are both plausible -- it is less plausible to have everything with just nice forms. I find things that are engineered without a single thought put in to their aesthetics still have a sort of beauty.
At the end of the day you still have to sell a game, probably to teenagers, so I don't know the final absolute answer on any of this. I mostly just wanted to post to say I don't personally use a lot of thumbnails, my art is less appealing for it, and I am not sure that's a bad thing at all so long as it's in service to the game as a whole.
Ninjas, I enjoy the robots you used in Reciever quite a bit. Aren't they an example of the same kind of thought as thumbnailing? Kitbashing functional shapes and inspirations together to create a variety of believable, iconic designs? They certainly aren't real-world functional either, but they sell to the viewer as if they are and suit the art style... Isn't that the exact same kind of design you're criticizing in here?
edit: not trying to badger you here, just curious about your viewpoint.
I did do some simple concepting work on the robots. In the game they are meant to be the product of extra-dimensional critters who had been spying on the thoughts of human engineers and designers -- a difficult idea to communicate. I mostly mashed up 70's and 80's consumer electronics with modern combat drones via image search. In the case of the tazer drone, I duplicated the fan and reversed it mostly because I wanted a plausible design, but didn't have time to model a tail fan, so the most unique aspect of it was mostly due to laziness
I think it's hard for me to not sound like I am talking out of my ass here. You can look at plenty of examples of more appealing, better executed art than mine done with thumbnailing (in this very thread!), and that my explanation of "well, my art supposed to look bad" is the noobiest kind of cop-out. I am still struggling with these ideas I have mentioned, but I think you can probably get where I'm coming from.
Thumbnailing is just a starting point. Throw down lines or shapes and work from there. People with a good knowledge of car design or whatever would be able to turn those thumbnails into something functional.
I think these can be considered thumbnails.
http://www.bmw.com/com/en/insights/bmw_design/process.html
edit: about other methods, on my first game project via Polycount I worked with a guy who had been an industrial designer at Black & Decker, and the first step of their design process was to come up with a mood board. Apparently developing the final mood board with the other guys on the team, through discussion, was a big part of the initial process.
http://www.mostinspired.com/blog/2009/09/16/mood-board-101-branding-and-image-development/
Designing for entertainment and designing for real life are two pretty different things. Like it or not games are a commercial project not fine art and therefore the most important thing to take into account when designing for them is to make something that looks cool and is somewhat believable as this is what appeals to the majority of the audience out there. Sure there are always exceptions but they are usually in the minority.
woah those are some pretty damn complex thumbs if they are thumbs, I mean look at the amazing line quality and perspective, maybe it has to be that complex to get an idea down though? I mean I cant imagine a car concept being much more simple than that
I was thinking about doing a video series about automotive design a while back and here are some things I found that I liked:
[ame="
[ame="
(this video just came out, but I like it too)
http://nbry.wordpress.com/2012/08/14/when-drawing-is-thinking-by-patrick-le-quement-renault-design-creator/
Really, thumbnailing is just a way to force the artist to think of the entire subject without getting bogged down in the details. They can be silhouettes or line drawings, or even those BMW sketches - speed is relative to the artist.
Snader thats a great link, the top page of little car thumbnails is exactly the kind of thing Im trying to learn to do fast. It takes me like 20 minutes to do a simple concept whereas that guy on ninjas link does a beautiful concept in 5 minutes! There must be a technique...or its just do it all day every day and practice will make perfect