Computer Graphic Design References

[Mithril Studios]

After my logo fiasco, I started looking for some beginner's references on doing web graphics / computer graphics and I was inundated with information - to say the least.

I would really like to learn the theory or fundamentals behind creating good cg graphics and websites. Usability and technical issues I'm pretty comfortable with - I'm ok with figuring out how to use Photoshop/Paintshop/Fireworks on my own - but I'm looking for something that will give me the reasoning behind why this logo is very good and these aren't.

If that sounds wacky, then my apologies! I'm just beginning to realize how clueless I really am when it comes to this stuff - I simply don't know what I don't know.

Would anyone be able to recommend some good books or websites that would help me out?

At some point I intend to outsource the artwork, but for now it's a personal goal to become rudimentarily familiar with the fundamentals involved.

Thanks!

Anthony

[Anthony Flack]

Interesting approach, but if you can't tell already, I'm not sure if there's some kind of algorithmic solution to let you assess their relative merits scientifically.

It brings to mind a quote that I think is from the Kurt Vonnegut book "Bluebeard". Apologies if it isn't, and apologies for paraphrasing, but...

The artist is asked how you can learn how to spot a great painting, and he replies that the best way is just to look at thousands and thousands of paintings.

[oNyx]

I cannot provide any links, but I can tell you (roughly) what a good logo is all about.

Not many colors and the shape is most important (get some contrast with your color selection). The shorter it takes to recognize a logo the better. It's also good if you can create some associative link to your product/company/whatsoever.

If I read "mithril" I think about... uhm... a military underground organisation. Most likely that isn't what you had in mind. After asking google... that fictional Tolkien material?

Oh and you should try to create your logo in a vector based format such as SVG. That allows lossless scaling (small for your page, bigger as splash, gigantic for tshirts).

Try these for doing SVG stuff (both are free):
http://www.inkscape.org/
http://www.sodipodi.com/

[botman]

Re: what oNyx said, mithril makes me think of your stereotypical fantasy stuff.

Anyways, my suggestion would be to do some study of typography (the study of text of course). My artistic talent is minimal like most programmers, but I know enough about fonts to be able to do some nice things with them. As I work as a web developer, it's pretty handy. You can make things (including logos) look good without being good at art.

[Nutter]

A good start might be colour theory (complimentary colours, etc) and typography theory (spacing, layout, styles, etc) books. I can't really recommend any unfortunately, but I'm sure doing a search on Amazon and reading the reviews will turn up a few gems (and the "you might also be interested in..." links will give you even more related books).

[mahlzeit]

The Non-Designer's Design Book by Robin Williams (not the actor) will teach you about typography and composition. It's a compact book (you can work through it in a couple of hours) and it contains a wealth of information.

My suggestion if you want to learn how to do artwork is to take a (non-computer) drawing class. You won't be drawing with pixels, but at least you'll learn the basics of colors and composition. Once you know that, your drawing skills are easily transferred to the computer. Using a real life brush isn't really that different from using a computer brush.

Other than that, just draw, draw, draw until you drop.

[Rainer Deyke]

I'm not an artist, but I can still tell what's wrong with your logos.

1. Small, difficult to read text. Bending it in a circle doesn't improve readability.

2. Lack of shapes that stand out. You've got a square and a circle, combined in the most obvious way.

3. Lack of contrast. Light blue and white are both bright colors. Put some black outline between them. As a general rule, your logo should look good in black and white.

4. Gratuitous special effects. The pattern on the background looks too complex, the "button" effect on the background look amateurish and out of place, and the bevel effect on the text is hard to make out.

[super_e]

Hi!

I'm currently doing all artwork stuff for eSoft Interactive and I have a few advices for you if you want to create and work on your own logo.

1. Focus on the ff.:
- impression you want to reflect to your customers once they see your logo the first time
- must be simple, catchy (even for black and white logos), and straight to the point
- main concept (for example, what do you exactly mean by "Mithril"?)
- secondary concept (for example you're making games, your logo show somehow show that if possible)

2. For company names like yours ("Mithril"), its hard to draw a simple image that will get your company recognized on a first impression. I'm not saying that you should change your company name but you could do some workarounds like emphasize the letter "M" or something like that.

3. When you decide on the colors, think of at least 2 or 3 colors that you would like to use for your company (website, letterheads, logo, themes, etc.). More colors would be great but you should anticipate future costs like if you wish to have a professional calling card printed, it would cost more for color separations. A combination of brighter and darker colors would be nice if you wish to cater to "black and white" documents (for example, your logo should still stand out on photocopied documents, etc.).

I know there are myriad ways and techniques in designing "killer" logos but that's about what I could think of for now. I hope this helps. Feel free to drop me a message if you need help.

Erick
http://www.esoftinteractive.com/

[maksum]

but I'm looking for something that will give me the reasoning behind why this logo is very good and these aren't.

If that sounds wacky, then my apologies! I'm just beginning to realize how clueless I really am when it comes to this stuff - I simply don't know what I don't know.

Anthony

Anthony, your post is flattering I'm thrilled you would think my logo is "very good" I'm actually still refining it, and when I post my "real" site I'll take off that glare and cast shadow. I just threw it on for fun, but ultimately, I feel it will take away more than add-to when it comes to presenting a professional image.

I agree with what others have said. This is what I have done...

Become proficient with graphics programs, such as Photoshop, Illustrator, Free Hand, etc.. If you're sure that you want to be somewhat serious with logos, then think Vector (ie Illustrator.)

Next, buy some books or find some great online resources of LOTS and LOTS of LOGOS that you like. If you go online, then print them off so you can just thumb through them and look at them. There are some great collections of logos in books out there. Search for "David E. Carter" on Amazon and/or Half.com for example. I LOVE books like those and love looking through them. As you look through them, you will pick out common practices, trends, patterns, etc..

Next combine the two. Try and re-create those that you like in Illustrator (or whatever). If you're able to recreate pretty much anything that you see, then it's just a matter of formulating ideas in your head and working them out.

Finally, peel your eyes away from the computer screen for a couple hours (at least) and just sketch ideas. Brainstorm, listing every element you can think of that has to do with your name or what you want to portray. As an example, when I was asked to come up with a logo for "Go Academic" I wrote down every thing I could think of that had to do with "go" (arrow, green light, etc.) and "Academic" (glasses, books, pencils, apple, chalk board, teacher, etc. etc.). I then did LOTS of sketching and finally came up with a green arrow that forms the letter "G" and an apple that forms the letter "O". You can check it out at GoAcademic.com. It's really not that well refined (low-budget and a bit rushed), but I think the idea is good. Something I NEVER would have come up with without sketching and brainstorming.

Anyway, those are just a few of my thoughts. I'm excited to see what you can come up with. PM me if you have any other questions about the BBG logo... again, thanks for the compliment

[Reactor]

Have a good read through a number of these articles...

http://www.webreference.com/dlab/

[SuperBug]

I'll take my shot at "explaining" why one logo looks good and the others don't.

The key is simplicity. The first logo looks good because it needs very little information to encode it - it's drawn by following very simple rules. You can almost immediately comprehend the underlying (hidden) structure. Circles, circles, circles. Almost everything is a circle or is derived from it. Even legs and hairs are shaped like circles were used as stencils. You can find a circle in every part of the picture and the coloring is very simple (a few shades of the same color used in a way that doesn't add too much detail and follows the "circular" rule, which harmonizes well with the rest of the picture). The brain appreciates that regularity and you percieve it as "good looking". It can "understand" the picture (it's "syntax") very easily and needs minimal effort to encode that pattern (a circle here, a circle there, a cut of a circle here...).

The "not good looking" logos are, on the other hand, designed by following the "rule of exception". You have a square, inside you have a circle, inside that you have a bunch of small letters (excessive detail), the texture doesn't match anything (not squares, nor circles) and it's also full of (different) detail, the square and the circle obviously don't match (and don't repeat anywhere, except for the circular text), the 3D look is inconsistent, shapes are quite unrelated in their proportions ... lots and lots of complexity. You would need relatively lot of information to encode all that "exceptional" detail. It's like music with a bunch of strange notes (unrelated shapes/patterns) and without a good rhythm (unrelated proportions), making it really hard for the listener to comprehend the underlying rules. Basically, the more complex the rules are, the more the emerging patterns appear like they are random or "noisy". The brain finds that too "frustrating" to understand (can't generalize much) and rejects it.

I found this very interesting:
http://www.idsia.ch/~juergen/locoart/node12.html

[Nauris]

I can heartily advise "Digital Color and Type" by Rob Carter. Its a great book that not only explains some simple laws, but also gives thousands and thousands of various color and type combinations. I often use just the example sections for quick inspirations. It would help not only with logo stuff but with anything else involving color and type, like interface design, website stuff etc.
Also, I like his approach that first he talks about the "iron laws" of type and then, the final chapters are devoted to how to break them

Not to sound harsh, but main problems with your logo are:

*zero "first glance" readability (if your logo was shown only for couple of seconds, it would be impossible to read what it is about.

*zero recognizability: there`s nothing that would make people remember your logo.

*It`s busy. Generally, texture in logo is a bad idea. Logo is the abstraction of idea, like cartoon is abstraction of human face. When its simplyfied to the point that you cant take anything away, yet, the idea is still communicated clearly, its a good logo. This could lead to the next problem:

*There`s no idea underneath. Its just textured block with type over it. you should think first, why did you choose Mythril for a name, and then think of how effectively visualize that reason in an abstract manner.

* and finally: avoid cheesy filters obvious bevel and pattern overlay just make image busier, but dont make it look pro.

I`m sorry if i sound grumpy-harsh, i`m actually not like that

[GameStudioD]

When I think of the word Mithril, I think about console RPGs. Specifically Shining Force 2 for Sega Genesis, where you collected pieces of Mithril to build better weapons. I assume your company is making RPG or D&D type games.

I threw this together in Illustrator just now, maybe this is something to start with (see attached file).

[paulm]

The design itself has been analysed fairly well here, so I don't know if I'm adding too much to the conversation.

However, the starting point really should be your brainstorming about who you are. Making your own logos and branding is the hardest thing to do (trust me, I know). So ask yourself, what kind of games does Mithril Studios make?

The Blue Bug logo does look great, as it has character and simplicity, but part of that is because you can also see the logo within the context of a greater design. Think of not only the logo itself, but where it's going to be used. If it's going to be used for a website, research websites that have the aesthetic you're trying to achieve.

After that, you need to get some books and learn the fundamentals of design. No matter how much wizardry you pull in Photoshop, design fundamentals are what really count and will pull you through no matter the technology. This includes such concepts as harmony, contrast, unity, composition, and colour.

The logo needs more contrast in terms of type, colours and composition. You need to bring out the type a lot more, through size or colour. The colours need to be simpler, as the textures are too muddled and the colours in them don't capture they eye.

The composition needs more contrast, as it is all too symmetrical and doesn't lead the eye in any particular direction. In the end there is too much harmony, too much symmetry that there is nothing pulling the eye and leading it in a direction.

The logo needs more unity in terms of your typeface, as that doesn't fit at all with the theme you're going for (as far as can be determined). The bigger question is, what IS your theme or idea?

What this all adds up to is a visually confusing message. Nauris has already pointed out why, and I've shown it in terms of more abstract concepts. Unfortunately, I can't recommend any good books on design fundamentals, but do a bit of research on the Internet, and you'll find something.

Cheers,
Paul.

[Mithril Studios]

Thanks everyone for all the insightful suggestions, comments, and ideas! You've given me a ton of new leads to work, and I really appreciate them!

Anthony