Documents... and a chimp

With the run up to our group project we have been looking at communication through documents. I never realised that there are so many different document steps in the game production process.

• The pitch
• Idea generation & refinement
• Visual guide (reference swatches, mood boards etc)
• Technical specifications & goals
• Plan of actions/break down, deadlines

Briefs/documentation really comes into play when working in groups and handing off work to others, when working alone we know our visual and goals, but with others this has to be conveyed as clearly as possible.
One comparative I can make is musicians – When playing music in a group you all have to understand each other, but not everyone understands what a bar is for example. The guitarist thinks a riff is a bar, instead of a pre-chorus it is a called a "little but before the next bit". I know fundamental musical terms, but the point I'm getting at is that not everyone uses the best language, so in a way briefs should be treated as if you were explaining to a chimp (or a punk guitarist), leave all uncertainties out, throw in a photo or a diagram to explain if the point warrants it.

Having a strong brief is like having strong outlines already there and we just fill out the space, if the brief is crap then the goal and process will be crap (if the brief is followed). I've had problems recently following my own briefs for personal projects, not (totally) because I'm lazy or stupid, but because of the reason, the purpose and the expected outcomes and timeframe. For example in my last post I started looking at making a character, I was stumped coming up with a design to fit my initial idea & brief. What I did was rethink the brief, instead of concentrating on my character must have "x" object and look like "y", I went back and filtered what I actually want to achieve

1. High poly character
2. Female form
3. Learn Strip Modelling, zbrush, and retopology/baking to medium poly
4. Render and look nice/export

(WIP)  So I immediately cut out specifics which weren't integral to these points. Within half an hour of this thinking I drew up an alternative idea/character which I felt was much stronger & appropriate. I converted the idea of having a massive weapon belt to accommodate many swords to the character having six arms. In a way this is a totally different project, but the fundamentals are all there and shining brighter. The reason I don't stipulate an exact result is so that I can explore as I'm going – I'm still learning, and there is a lot to learn, but would be neglecting by repeating the exact processes I already know.

Above I thought for a minute about adding chains and worked out a decent way to do it in 3DS. From this I learned that you can apply (and stack) modifiers to linked/grouped objects – This is amazingly useful to know and would not have known without the leeway to experiment.  I think that's what personal projects are all about - exploring & developing.

Paris, Hitler & Jazz

Last week we watched Win Wyman's Paris, Texas. We were shown it to help us understand and think about composition and visually communicating emotion and story-telling.
The cinematography and directing told the story for the first forty-minutes. I rented the film from the library to see its conclusion, but unfortunately I was visually distracted doodling so I don't think I reaped the Benefits of the visual awesomeness. The story was told backwards somewhat and led towards the reason why Travis was lost to the world and behaving odd. I got a little emotional at times, but the last scene made me ill, most of us have had tough breakups at points and it seemed a little "emo". I basically got the impression the writer never got over his girlfriend and cried into his script, so the end was a bit of a cringer, but uplifting too. The main actor Harry Dean Stanton, is the guy out of a film I used to like in my youth – Repo Man, it's similar in a way as it supports a dystopian ideology, and everyone seems to be cool and has strength in their own ways… It also has an alien in the car boot which burns up anyone who sees it. Harry's character is on a similar journey as in Paris Texas, but instead of returning to his family and love he follows a blind faith & an attunement with this alien. Is worth a watch. The feel to it also reminded me of Happiness and Little Miss sunshine – These are all kinda "mainstream" alternative cinema and use the same kind of subversive direction, palette and pacing... Seems common in foreign films

 
Colour seems a bit of a tough concept for some of us students. I have seen quite a lot of digital fails when attempting to colour, along with a lot of colour picking from reference. I actually did a little exercise to see how to "rip off" pallets from pictures, here I did a couple of ten minute thumbnails. I just took what I thought were the main colours (keeping a small palette, no more than 6 colours) and blocked out the image like a thumbnail.  The picture in the middle is a doodle with an extreme palette - looks a lot like eighties graphical art  .I have also been reading Colour & Light, which was surprisingly awesome, I actually prefer the artists studies over his Dinotopia stuff. I would recommend it to every art student; it's very thorough and informative. I've never really spent hours on end thinking of colour, light & shadow. I actually thought the thumbs looked pretty decent considering was only ten mins, it just goes to show that strong composition and colours make all the difference. It's a lot quicker and more natural using a limited palette, I favour mixing on the image and you can colour pick locally while using the flow and opacity, then undoing the mix. Anyways...

 
I've been doing a lot more 3D than 2D recently, there is simply more to learn and not a lot of time to learn it. UDK is a bit of a treat, it seems really flexible, yet a little ugly and takes time to navigate and learn. Zbrush on the other hand makes me laugh how easy and intuitive it seems to use its functions – seems like cheating after using 3DS Max. Anyways I was looking into UDK and watching the tutorials, I actually thought I was awesome how you can import a library of assets (along with the other libraries there) and make a scene, much like Lego. It got me thinking about how to make my own efficient sets. The trick is to make a lot of modular bits (walls, pillars, vents, trims, furniture, decorations and the like), this should be made to work together. For example instead of making a whole building in 3DS max we can import different floors and elements and then stick them together creatively in UDK.

Early in the term I was doing small personal projects regularly, I have to structure this better… To learn and apply 3D it can take days of slogging to do the whole process, along with a lot of trial and error. So I have set a couple of longer projects.

Character:
I have been drawing up ideas for a character, I want to do a female as I need the practice. I want a lady with a lot of different swords.
I want it to look good, so will need to bake with Zbrush, I have also been looking around and see you can bake down further in 3DS to make a really low poly version too – I want to learn this also.  I have to get a balance as I want to create an unrealistic character, but make her believable, starting anime and going realistic.

[ABOVE]  I have decided to try and pin down the general details - I know I want feminine features, but have to refine it a lot more.  The main feature is to be her weapons and her holster, something to contrast against her female bits, thinking about weapon belt/holster shapes here

Environment:
I've started a simple tower in a scene, and am using it as a learning tool. How to modulate and share map space, making efficient maps for this and making asset sets for UDK.
Also an opportunity to fill gaps in my 3DS learning, and to try out new scripts to improve workflow.
So far I have actually been going back to the drawing board a lot and trying to visualise and rationalise the assets. This is also great for practicing concept art – I will try and get it to a standard where others can read it too! :D

I have realised that to learn everything properly to do them well it takes a while and re-iterations, and while I can throw out some crap every day or two I wouldn't be learning or progressing myself as much. I have set deadlines for Christmas, and realise the outcomes won't be perfect, but is good practice and helps me not to "faff" between uni projects. From these projects I have also been building up my vision and reference library and adding more and more links and books to my collection.

Drawing n Stuff

Being relatively new to using digital tablets and more at home with traditional medial I have noticed lots of my skills are lost or don’t translate so well digitally.  One of my strengths is having quick expressive lines, which let’s be honest is way too loose to create the quality images needed.

I quite liked this 5 minute sketch I did at the pose workshop arranged by Shay (sketch starring Kit).  The loose lines aren’t appropriate for concept art.  Think I’m possibly influenced by Quentin Blake; I need to think how I can apply that in a concept manner.  I drew this in class the other day which got me thinking

When I draw naturally I scribble in and imply detail, it’s quite easy to blag bits and with appropriate reference could be more authentic.  I also watched a DVD tutorial I got with Imagine FX magazine (can't locate it now to say name or link to online vid if there is one), it was interesting to see his line of thought and work process.  Some interesting points were that to look cool sometimes you have to push the reality a bit, even if a part of a vehicle for example isn't needed it doesn’t matter soo much.  I have wings on my thing (reminds me of the Thundertank… Maybe it helps create lift for faster ground speed?  It sure as hell can’t fly… And what are the paws about?  

See, it doesn’t really matter until it has to be implemented in the universe (which I make up for this vehicle project).  He also stated that he would draw semi-detailed images just to get a feel for the subject matter and try out ideas early on in his design process (day 1).   Also he mentioned perspective, he makes a few guide lines after drawing the initial shapes – I found this a lot more useful, also perspective comes second to coolness (getting pleasant balance can be hard).  On my pic the placement and perspective is odd, but it displays what I wanted to show ok and has given a little organic twist to the design in its asymmetry.

Anyhow I’ve been doing a little more “penning” digitally.  I have already scanned traditional media marks in photoshop and made some of my own “traditional brushes”.  I have since making them learned quite a bit to help make them more loyal to their traditional counterparts.  Much mark making in Photoshop is done through clever textured brushes, along with using the lasso to make a sharp selection edge, filtering and blending back (among many other ways).  I’m currently looking into making manual marks quicker.  This somehow all combines as you learn, and making brushes for your vision becomes more fluid and appropriate.

Above is an example of bad brush use, here I spent about 1 minute drawing a simple leaf and scattered it randomly.   In my eyes it looks naff, but luckily lead towards the death of the diabolical image.
e is an example of bad brush use, here I spent about 1 minute making a simple leaf and scattered it randomly.   In my eyes it looks naff, but luckily lead towards the death of the diabolical image.
Here is some better use of custom brushes.  Spending a little time experimenting and analysing other brushes is possibly essential to getting decent at digi-painting.  Here I just used random foliage photos I took as a textured background.  Using silhouettes from nature also seems to work well.

Above are a couple of examples – The square brush emulates chalk in a way, using the negative of branches against the sky, using big brushes and photos allows for more virtual detail and can hide a possible seam if the texture repeats.  You can also zoom in the texture loads to make it less noisy and soften the texture somewhat.  Here I’ve used a close-up of a tombstone and some ivy.

One trick good digital artists pull off is hiding the photo elements and making you believe it is all one… Is it cheating?  Don’t think there are any rules.

Composition

I’m re-writing this one as my last draft is a little old now.  Lots of the discussion we had about composition was meant as a follow on from Mitch’s class.  Unfortunately I missed that class (think I was ill on the Tuesday it was communicated).

There are many accepted precepts to composition, these have developed over all of existence and I like to think applies to all living things.  There is a primal appeal visually to anything we see – for example symmetrical faces are typically beautiful, we like forms and values we can easily interpret subconsciously.  If there is a large imposing object we may feel threatened, if there are curvy aerodynamic lines it puts you at ease (and the eye enjoys following them), if there is a lot of “noise” or contrasting detail it will get our attention (for better or worse).

Composition is a quite a personal thing, and is a matter of taste.  Below is one of the drawings I had feedback on concerning composition (1st year).  Meg suggested that I should consider putting detail on the left bank. In my eyes I wanted that space sparse so the viewer is drawn towards the detail of the trees on the right (using the rule of thirds here too).  I like the idea that the eye knows where to look, but is also invited to follow the stream, balancing the image.  For me it’s a tranquil composition – very light and soft in places, but also with a contrast and a little interest.  Here also I used the foreground as a frame, omitting detail and contrast and using hard lines to suggest detail and form.  It was by no means a masterpiece, but I find it interesting knowing what others think.  Feedback is nice, but I think people are scared of hurting other’s feelings , also lots just stick to “Use the rule of thirds” “Have the foreground become gradually darker “ – I much prefer personal opinion and questions.

The trick to a successful composition is using these precepts to not only create an accurate image (perspective, light and shadow values etc), but to also invoke an emotion and create an attachment to the subject visually.  I personally think the best way of exploring this is through examining nature, for example a dark image would remind people of night time or shadow, in the past before CCTV and home insurance this would be a time of danger.  In the animal kingdom colours are utterly important, plants naturally tint their fruits to attract animals to aid procreation, animals use colours to attract a mate, warn off predators, or to conceal their presence.

That’s all good, but how does it apply to game art?  Game art is slightly different, I did a graphics course and the presentation is very similar, especially for 3D work.  Not only do we want pleasing compositions but we also need to present a certain level of detail in the work also and make sure the viewer can see the workings.

Above here is my trash project.  As you can see arrangement wise it is a little drab, but I would say it is a successful presentation.  I used the variety of primitives to create contrasting shapes among the items here.  You can see everything I modelled, along with the wire-frame and three maps all in one shot.  I actually spent quite a bit of time arranging the objects so they can be all be seen at once, I previously had the crap scattered around the bin as I was experimenting with a turntable, which was quite different.  When It was rotating on the turntable I in reality had many different compositions (from the rotating view) and had to arrange the items so the eye had time to move from one object to the next and always have something interesting to see as the scene rotated.  But alas we are posting photos and not turntable videos, so I went back to a single “glamour” shot.

Here I went back after receiving some feedback, I toned down the shine and rounded off the pumpkin & bottle... I think the composition isn't as good, but it should help to improve the assets and grade.

I have seen quite a few “bad” 3D compositions posted on Facebook, where not only is it unpleasing to look at.  This may be via bad bordering, clashing backgrounds, and generally erratic compositions, also where it’s actually a challenge to see exactly what’s there.  (note: above box is hidden by magazine = bad)

Wizadora & Wizzbit.. Some old school references, along with Razzle

Planning n Concepting

Week 2...



My last Blog touched on this subject, as well as going on about doodling, so I wont be repeating myself too much.

Basically the theme today is "working smarter, not harder".  That's is to say that I could spend 18 hours a day  doing pictures of people in armour, but without any thought or process the work will be unfocused and simply lame.

So for part of my learning I'm taking a step back and thinking about the whole work process.  For starters when I feel myself going on a random tangent I will stop.  Last night I had my first go at Zbrush - Instead of just opening the program and clicking on everything (trial and error) I read quite a few tutorials and looked online.  This hour or two spend slowly acquiring information is more productive than a hole day of clicking and hoping.  Also I noticed after a point (I was watching a one hour tutorial) I was not absorbing further information, which indicated that I should put into practice what little I learned before that was lost too.

I was quit happy with the results for a first go, and it was actually a "doodle" as I used no reference.  By reducing the workload/goals it freed up time to experiment with the fundamental tools and navigation.  I spent a couple of hours "playing around", but sometimes there is no substitute for experience.

This was one of my own mini-projects - operation learn a bit of Zbrush.  Goal was to make a detailed head and get the program going.  The next step is to create & import meshes from 3DS Max - full body ones, and to learn to bake.  I could see myself wasting loads of time using zbrush as it seems super intuative (compared to Autodesk) and easy really damn easy to get results I thought above me.  I must resist however, and keep goals in mind.

The sculpting process in my eyes is exactly how it should be - a lot like moulding clay.  To make the head you get a sphere, pull and push it.  Here I started by making the bone frame then added on bits with a brush to make the muscles and features - Just like with traditional media, but much more flexible and accessible.

I read the article "Concept Art process" and attended a talk by a third year about the work pipeline within a functioning company.  Even though the article is concerning the composing of an image for a magazine and Mike's talk was about Lockwood - a studio that makes assets for the virtual PlayStation universe they were about the same thing - work flow.

An important aspect of any workflow is planning and procedure.  While the concept artist plans his work for a tutorial, ensuring that the content and composition are ideal the Lockwood have a meeting to plan what assets are needed.

The actual physical work is not the whole process.  Here's a breakdown of Lockwood's work pipe-line

5%    Reference
10%  Concpet & Block out (thumbnails for 2D comparative)
10%  Whitebox with place holders (refining favoured thumbnails)
40%  Model and texture (Paint and details in 2D)
35%  Refinement (going off feedback or analyzing personally, adjusting levels etc in 2D)

I think that's a very reasonable balance of time.  As you can see, not even half of the time is spent on the main body of work, yet it is most efficient!  Why?  Well first off they aren't learning their trade, so it is a little different to an educational course, but the first quarter of the time is spent planning and acquiring the right tools for the job.  The hardest part sometimes is not only formulating the best solution/desired outcome but also visualizing it before the work is even started.  A bad idea will produce a bad outcome and ensuring you are working accurately towards the vision is essential.

I find I tend to use this kind of structure - Finishing work early leaves you time to fix any errors you that may have passed you by before.  I tend to have a break and spend a while observing my work, normally making tweaks changing any parts I am unhappy with and gradually polishing off the work.  This touches a point in the previous post about finishing work - Make sure you have a goal and set some standards otherwise you could get lost here.

As part of my learning I am trying to get into the concept art thing, I think I'm a fine artist at heart.  So I have set myself a continuous task - To do a daily concept piece... Well not every day as it does take up time.  My main goal is to emulate the concepty style and to get work out a nice process.

I made myself a random image content generator type thing to keep me trying different genres and assets.
I roll a (virtual) dice...

Dice	Genre	Focus
1/2	Urban	Character
3/4	Sci-Fi	Environment
5/6	Fantasy	Vehicle

This will give me something different to do each time and removes part of thinking.  I'm not sure where to focus yet - I like doing a bit of everything.

I want the process to be quick so for characters I used posemaniacs and painted over the figure.  I know that most professional digital artists cheat in every way possible, it's just a matter of not making it obvious or hopefully apparent at all.  Anyways, I missed a vital part here - I'm basically just drawing the costume and I made it up on the spot.  It's a bit crap to be honest... I thought "sci-fi, she could have an astronaut's helm! Nothing really fits - The mini-skirt looks feminine but would she really have those two items together?  It's a bit "comic book", I made her eye glow white just because.  I does somehow work in a retro (Buck Rogers) way.

I could post rationalize all day and say she's this and that, but this should all be told in the image.  So next time I need more reference and actually decide these things - the picture would be less fail as it actually achieves some visual goals.  This would in turn help to draw the viewer in and add credibility and character to an actual character.  Also on analysis I think the picture is "unfinished" - I would like to see more geometric shapes and hard lines to fit with the sci-fi theme, along with a bit of detail to explain and interconnect the clothing better.  Needs to be added to the process earlier on

See it's a long process to draw a decent picture, and as I need to do more prep work I need to allocate more time to get good at that - Micro managing each piece of work.  Any how, the next one should be interesting.  I will spend a bit more time, up to an hour at the planning/reference process.  That will be a four hour project if sticking to the "Lockwood formula"

Catchup

Easy now, I started a new blog and emailed the address, but will add here instead... week 1

I feel that I learned a lot in the first year, not just 3DS Max, and theory but more importantly I learned how to learn better... If that makes sense?  What I mean is that I was in a comfortable (yet not sociable) routine, putting a lot of hours into university work while exploring & learning at the same time.

I can spend a lot of time working but I want to try and manage that time better.  Setting regular short term goals which I can say are "finished" and move onto the next thing.  I have a bad habit of obsessing over the current thing and getting slightly lost.  I'm going to take a step back and rethink my routine.

first thing - have a purpose to each piece I do.  What I have often been doing is "doodling".  I'd start drawing something and not really know where I'm going.  I'm not saying I don't learn anything at all or enjoy it, but it is just bad practice and the results are powerless as I have no real goal, or reason why it should work.

Here's a doodle I did.  Even though I like it myself I could understand that not many others would.  I should be trying to make a quality picture.  Many of my doodles start with what I consider strong lines and tones... Otherwise I would just start all over as it is a quick process.  What I need to do is use that skill better and actually finishing images.

This picture was a "fail" as I know it's a doodle, By default a doodle has no goal, its just a random brain fart, doodling is what you scribble out when on the phone.  On the plus side it is a very quick process and helps develop judgement, technique, and calmness (is great for a break).

But year two is a step up.  One of my main goals is to set myself more personal projects - Including managing my time better and working on my portfolio.  That's the reason I played around with blogger a bit - A little practice.  I thought a serious portfolio shouldn't have weekly rants about random crap (these blogs I do).

Another mini-project thing I have been doing involves http://www.posemaniacs.com/ it's a pretty popular site for us Game Art students and is a pretty simple site.  I'm spending about twenty minute some mornings drawing quick 30 second human figure in Photoshop.  It's great for developing judgement (where and how to draw the lines), speed (every second counts), the eye (many poses and perspectives and orientations), and is a pretty nice way to wake up a little.

There is also a second bonus as it gets me acquainted with using references - Which I often don't use, and my work suffers... It used to feel like copying.  But the Mona Lisa had reference, and that's a pretty good painting.

As for 3D work I am going to try and learn a lot of different things.  Me poppy bought me Marmoset for me birthday, and it's pretty damn easy to use. I plan to go back over Mr Pickton's tutorials as there was a lot I didn't impliment/learn and will be very useful.

High poly modelling, Zbrush, more 3DS Max, and UDK are all pretty vital to learn... Gonna leave Maya until it's vital too.  So some personal 3D projects would be a good idea.