30 replies to this topic

[glhf]

My GDD is getting near completion so it's time to make it even more detailed and start on creating some GUI art but I need to know what dimensions the 2d images for buttons, background etc should be if it's going to be fullscreen.. since people have different sized screens and resolutions.

And what is best format? PNG or JPG or GIF or BMP?

Edited by glhf, 05 June 2012 - 04:22 AM.

[Madhed]

You will have to find a way to make your ui resolution independent. Usually that means providing fonts with different sizes and a way to scale buttons and other graphical elements without pixelating blurring. There is a method where you cut your sprite into 9 chunks that can be independently scaled but I don't know if it has an actual name. It is widely used however and you should find it easily.

As for formats:
JPG - no, it introduces compression artifacts and has no transparency.
GIF - no, restricted to 256 colors and has only on/off transparency
BMP - no, no transparency and is uncompressed
PNG - ok, has alpha transparency and lossless compression

[kauna]

How about using DDS format (especially if you are working with D3D) ? It is a container format which may contain different surface formats (including alpha channel), compressed and uncompressed. It supports also mipmaps.

Otherwise, a GUI which is more and less resolution independent is the way to go.

Cheers!

[taby]

I prefer TGA format images. Even with just basic RLE "compression" (Apple PackBits-ish), the images can be quite a lot smaller than their uncompressed counterparts, yet still load pretty quickly. It supports an alpha channel (32-bit images). It is an extremely simple format. It just works, and a loader/saver class can be written up in a day or two, tops. Lots of explanation and sample source to go by: http://local.wasp.uw...ataformats/tga/ http://local.wasp.uw...s/tga/tgatest.c

Edited by taby, 05 June 2012 - 11:24 AM.

[glhf]

I've heard that shrinking the dimensions of a picture doesn't reduce quality... only enlarging it.. That should be something to keep in mind maybe?
I will look into that 9 chunks method, thanks..

What is a good starting dimension to go from?
Should be a large one but at the same time not too large then nessassary.
What do you think is good dimension to make art for? Should you make the art so it fits perfectly the biggest resolution available on pc and then let the fullscreen shrink it or enlarge it from there on others computers?
What is the best resolution to make the art for?
What dimension should the screens size be that I fit the game for?

[frob]

Seconding the DDS format.

DDS has been adopted as the de facto standard for about a decade now. It has lossless compression available. Most tools and graphics programs can handle it just fine. Most importantly it is supported in hardware by current cards so no decompression or processing is necessary, and works under Direct3D and under OpenGL.

When it comes to art assets really there should be the psd from artists working in photoshop, and the dds they export for use in game.

[glhf]

What about making the art with vectors?
Can you do that in photoshop? I only know how to do it in flash

with vector you dont lose any resolation no matter what you rezise it too

[szecs]

Vector graphics is great and all, but it's limited. You can only use lines with variable thickness (??? I'm not even sure about this), and solid (or simple gradient) color filling. If that's the style you want, it's fine.

Anyway, Vector graphics is a whole different thing than pixel graphics. You have to handle all the rendering of the lines/curves and filled areas, which is pretty complex. There's nothing like just drawing a polygon and throwing a texture on it.
I may be wrong, maybe there's a library for that, but I'm not sure you understand what vector graphics really is.

I think Photoshop can work with vector graphics and saves it too (but not in a regular image format of course), but there are more feasible applications for that, for example Corell Draw.

Edited by szecs, 07 June 2012 - 09:37 AM.

[Madhed]

Actually, it might be perfectly possible to create vector graphics in a 3d modeling program. just create flat polygonal objects instead of 3d. I'm not sure about the workflow however, could be pretty cumbersome. You would still have edges becoming visible when scaling though.

Edited by Madhed, 07 June 2012 - 09:54 AM.

[Josh Petrie]

My GDD is getting near completion so it's time to make it even more detailed...

I would venture to suggest that your next step should be to actually start making your game, instead of trying to do all this design and "GDD" bullshit up front. You're putting the cart before the horse, and if you spend too much time going too far in-depth with your design and pre-producing your art assets and all that you run the risk of having to throw away a ton of that time when it becomes apparent during iteration that something isn't going to work.

You've already made mistakes in your design and content production, so start iterating on something practical so you can expose those mistakes and correct them sooner rather than later.

Edited by Josh Petrie, 07 June 2012 - 10:15 AM.

[glhf]

My GDD is getting near completion so it's time to make it even more detailed...

I would venture to suggest that your next step should be to actually start making your game, instead of trying to do all this design and "GDD" bullshit up front. You're putting the cart before the horse, and if you spend too much time going too far in-depth with your design and pre-producing your art assets and all that you run the risk of having to throw away a ton of that time when it becomes apparent during iteration that something isn't going to work.

You've already made mistakes in your design and content production, so start iterating on something practical so you can expose those mistakes and correct them sooner rather than later.

You're offtopic.
You also don't seem to know anything about GDD's.
And I haven't made any mistakes.

[Jebbs]

My GDD is getting near completion so it's time to make it even more detailed...

I would venture to suggest that your next step should be to actually start making your game, instead of trying to do all this design and "GDD" bullshit up front. You're putting the cart before the horse, and if you spend too much time going too far in-depth with your design and pre-producing your art assets and all that you run the risk of having to throw away a ton of that time when it becomes apparent during iteration that something isn't going to work.

You've already made mistakes in your design and content production, so start iterating on something practical so you can expose those mistakes and correct them sooner rather than later.

I think that was pretty harsh, and quite frankly unnecessary. A game design doc can be one of the most helpful things when making a game if it's done right, and I don't know why you would accuse someone of making mistakes outright like that.


As for the question,

As mentioned earlier, having a larger imaged scaled down looks a lot better than having a smaller image scaled up. I would probably use that in conjuncture with different sized fonts.

Edited by Jebbles, 07 June 2012 - 12:39 PM.

[Ravyne]

...

You're offtopic.
You also don't seem to know anything about GDD's.
And I haven't made any mistakes.

And you're a bit out-of-line. I assure you that Mr. Petrie is far more familiar with GDDs than you. You don't think you've made any mistakes because you don't know enough to see them.

You're here asking rather simple questions that any amount of cursory research would answer, so either you haven't bothered to do your own homework, or you'd sooner let the opinions of internet strangers, however well-intentioned, set your course. Neither of these inspires confidence that you're willing or able to consider all the angles. The fact that you so readily dismiss an opposing viewpoint reinforces that lack of confidence.

Anyone who's designed anything much more complex than doorstop knows that design has to be iterated upon. As they say in war, even the best-laid plans don't survive contact with the battlefield. All that Josh was saying is that you're front-loading too much design with too little hands-on experience. the more experience you have, the more you'll be able to pre-design to a reasonable facsimile of the end result. This is true of design for both gameplay and technical implementation, and more-so where they touch. If you spend time laying out a poor plan, the best outcome is that you'll throw away the plan, but the worse (and more likely) outcome is that you'll throw away the plan, and also sink a lot more time into correcting the problem, and all the systems that were built around it.

The trick to planning isn't laying out all the details ahead of time, it's knowing what details should be laid out, and which are best to discover along the way. Image formats, while important, are absolute minutia. Choosing a resolution can be important, but a better question would be, "can I, and how do I, support different screen resolutions and geometries (4:3, 16:10, and 16:9 are most common)

[glhf]

...

You're offtopic.
You also don't seem to know anything about GDD's.
And I haven't made any mistakes.

And you're a bit out-of-line. I assure you that Mr. Petrie is far more familiar with GDDs than you. You don't think you've made any mistakes because you don't know enough to see them.

You're here asking rather simple questions that any amount of cursory research would answer, so either you haven't bothered to do your own homework, or you'd sooner let the opinions of internet strangers, however well-intentioned, set your course. Neither of these inspires confidence that you're willing or able to consider all the angles. The fact that you so readily dismiss an opposing viewpoint reinforces that lack of confidence.

Anyone who's designed anything much more complex than doorstop knows that design has to be iterated upon. As they say in war, even the best-laid plans don't survive contact with the battlefield. All that Josh was saying is that you're front-loading too much design with too little hands-on experience. the more experience you have, the more you'll be able to pre-design to a reasonable facsimile of the end result. This is true of design for both gameplay and technical implementation, and more-so where they touch. If you spend time laying out a poor plan, the best outcome is that you'll throw away the plan, but the worse (and more likely) outcome is that you'll throw away the plan, and also sink a lot more time into correcting the problem, and all the systems that were built around it.

The trick to planning isn't laying out all the details ahead of time, it's knowing what details should be laid out, and which are best to discover along the way. Image formats, while important, are absolute minutia. Choosing a resolution can be important, but a better question would be, "can I, and how do I, support different screen resolutions and geometries (4:3, 16:10, and 16:9 are most common)

I don't agree.
The pre-development phase is free since I'm the one doing it.
The more important details I can include in the GDD that I give out to the team later the smoother and faster development will be = cost less money.
So It's just much smarter if I already know what sizes the images should be and even have them ready = saves lots of time because the programmer can just puzzle the pieces together instead of discussing and planning with me = cost more money.

And I really wasn't out of line, The moderator succesfully have derailed this thread now.. Good job on him.
He didn't even leave any comment at all about what the thread actually was for... images.

[glhf]

As for the question,

As mentioned earlier, having a larger imaged scaled down looks a lot better than having a smaller image scaled up. I would probably use that in conjuncture with different sized fonts.

Thanks, But What dimension should I originate from if it's going to be designed for fullscreen and scaling down to fit?
1600 x 1200 ? That's the max of what my computer allows.

[Josh Petrie]

You don't own this thread and have no authority to attempt to dictate it's course. This is a discussion forum. The only people who have that authority are the moderators, and I am the moderator of this forum. Do not attempt to play rules-lawyer with me.

Now then.

I think that was pretty harsh, and quite frankly unnecessary. A game design doc can be one of the most helpful things when making a game if it's done right, and I don't know why you would accuse someone of making mistakes outright like that.

Yes, it was harsh. But I did not say that a game design document is a bad thing, I said that trying to do all the design work up front is bullshit, and quoted the term GDD to express some skepticism at the idea of specific pixel sizes being relevant to the design of the game. It's an implementation detail, and potentially one that belongs more in a technical specification if anything. As for mistakes? Anybody who believes they've never made any is a fool, or as Ravyne suggests, has never developed anything particular complex. All software has bugs in it, all designs have errors or omissions or exploits or loopholes. We make mistakes, probably more often than we made things correctly, and that's good because we learn from them. That's why we iterate.

The pre-development phase is free since I'm the one doing it.

The more important details I can include in the GDD that I give out to the team later the smoother and faster development will be = cost less money.

So It's just much smarter if I already know what sizes the images should be and even have them ready = saves lots of time because the programmer can just puzzle the pieces together instead of discussing and planning with me = cost more money.

Opportunity cost is relevant and needs to be considered. Especially if you going to hand off your design to a third party to actually get it implemented, you cannot expect to actually get anything produced without any kind of back and forth dialog and iteration. It is just impractical, it doesn't work that way. You will invariably have forgotten something that your developer will need to ask you about. You're only human.

Worrying about specific image resolutions isn't something you need in the GDD. What you want to call out is, as Ravyne suggested, that you want scalable resolution support. Your developer will be able to translate that into the technical details he needs to produce the implementation you desire.

[Ravyne]

Spoken with child-like naivety.

Yes, requirements are important, but the one's you're concentrating on are, relatively speaking, of the smallest order. Many other things will have a far-greater impact on how your software will be structured and constructed. For example, target system requirements. Your artists almost certainly shouldn't be working in native resolution on any modern platform, so even the final resolution of the game has relatively little impact on work-flows.

What really costs money later is not discussing things as they come up, but blindly executing a poor plan, and then having to backtrack and fix it later. If you're talking in terms of having a paid staff, just think how much money it would cost to have most of the staff sitting on their hands with no work to do while a few key people figure out how to tear apart the code and re-build it in a way that only now makes sense.

In short, your assumption that more planning means less costs relies upon the plan being flawless in every non-trivial regard. Even NASA occasionally and spectacularly gets things wrong, and they have a thousand of the brightest and most-experienced minds in the world focusing on one big task.

In the real world, being committed to the wrong plan always costs more in the long run.

Having the humility to admit what you don't know saves a lot of pain in the end.

[frob]

Thanks, But What dimension should I originate from if it's going to be designed for fullscreen and scaling down to fit?
1600 x 1200 ? That's the max of what my computer allows.

It depends.

If you are talking about a splash-screen full screen background, then you will want to include at least a 4:3 (most common), 16:9 (second most common) 16:10 (third most common) ratio images. Exactly how large those images should be ought to be a decision for your art people. This detail generally IS NOT part of a design document. It is part of the art direction, setting the image size to whatever the art director or lead artist decides it needs to be.

If you are talking about textures for individual UI objects, you will want to include at least those three most common image ratios. Again, this detail generally IS NOT part of a design document. It is generally decided by the person doing art direction based on the visual style being created.

Finally, if you are talking about textures for modeled objects, those are power-of-two textures. You can specify an approximate size to pixel ratio, but again, this detail is generally not part of the design document. It is generally decided by art direction and technical artists based on the actual performance of the engine.

[glhf]

Thanks, But What dimension should I originate from if it's going to be designed for fullscreen and scaling down to fit?
1600 x 1200 ? That's the max of what my computer allows.

It depends.

If you are talking about a splash-screen full screen background, then you will want to include at least a 4:3 (most common), 16:9 (second most common) 16:10 (third most common) ratio images. Exactly how large those images should be ought to be a decision for your art people. This detail generally IS NOT part of a design document. It is part of the art direction, setting the image size to whatever the art director or lead artist decides it needs to be.

If you are talking about textures for individual UI objects, you will want to include at least those three most common image ratios. Again, this detail generally IS NOT part of a design document. It is generally decided by the person doing art direction based on the visual style being created.

Finally, if you are talking about textures for modeled objects, those are power-of-two textures. You can specify an approximate size to pixel ratio, but again, this detail is generally not part of the design document. It is generally decided by art direction and technical artists based on the actual performance of the engine.

Thanks and I know it's not technically part of the design document but that's indy for you.. We need to take the roles of more than one job and I will simply try and do as many jobs I can because everything I do myself in pre-development doesn't cost me anything. It's true about opportunity cost but I'm not really in the position to have to worry about that because I don't have other projects going on or waiting for me.

I think I am good on what I need to do for art now.. time to get back to work =)

[yckx]

I concur that you're dealing with this issue before its time. And if you are giving the design document to a team, these are really issues that the UI artist should already be familiar with and understand.

But if you want to have some idea of the resolutions you should support, do a search for "browser display statistics". This should give you a decent idea of what resolutions people are using. You can use this data to determine for yourself the resolutions you'd like to target.

Also, I've often seen games provide UI elements in 2 or 3 different sizes and choose which to use based on screen resolution. Even if you slice up buttons and whatnot into corner/edge/middle sections, the same graphic may not look good at both your minimum and maximum supported display size. Having multiple sizes of the UI elements can help to mitigate this issue.