作者：Maria Sifnioti

今天我要谈论的是资产的处理，这里的资产管理主要涉及质量、预先计划及平衡性。

资产是除源代码/脚本文件外的游戏组成元素。资产源自各种不同媒介，由不同的人员负责。例如：

音乐：音乐人

音效：声音工程师

插图：美工

用户界面：UI设计师和美工

动画：美工

3D物体：美工

视频：制作人和美工

对话/文本：文案和本土化团队

这里涉及众多不同职业。所有这些人员在创建游戏资产时都需就下述惯例达成共识：

1. 资产质量

2. 资产

3. 资产备份和存储

负责通知相关成员这些惯例的，或是首席美术设计师/音效师/音乐人/文案，或是小团队中的制作人/助理制片人。

沟通和良好的团队合作在此非常重要。非专业人士需要全面把握专业知识，方能向专业人士传达项目需求。

下面我将以《Maya’s Dress Up》为例剖析上述各惯例。我的例子主要瞄准图像资产，但也会涉及如何将其应用至其他富媒体。

质量&成本权衡

《Maya’s Dress Up》是我们最新的iOS游戏，游戏融入众多图像资产。下面就来看其中的典型资产——项链。在制作人和美工就着色达成一致看法后，游戏就需要确定相应的质量标准。图像可以通过DPI（每英寸点数）进行呈现。

* 资金成本：DPI越多，图像越大，投入的成本就越多（游戏邦注：因为美工需要在插画中添加额外细节内容）。

* 时间成本：DPI越多，图像越大，美工所要投入的时间就越多。

下面是Maya的一款项链，我们对于所有插画都坚持300 DPI标准。

左侧呈现的是插图300 dpi的细节处理程度——。而右侧则是72 dpi画面的详细程度，此时我们无法看到项链的链环以及蝴蝶的玻璃光泽。这无疑能让美工省下许多时间。假设我们有500个道具。

* 72 dpi道具平均耗时1小时（总计500小时）

* 300 dpi道具平均耗时2-3小时（总计1000-1500小时）

乍看之下，仿佛如果我们选择低质量标准，就可以省下许多时间。下面是Maya项链在第一代iPod（游戏邦注：分辨率480×320）的呈现效果：

你也许会觉得“相当小”。想想如果你的所有资产都采用低质量标准，然后苹果推出分辨率1024×768的iPad，接着没多久，他们又推出支持Retina技术的新iPhone。这令设备屏幕像数翻一番，要求图像达到960×640分辨率。

因此若你想要转投这些新设备，就得提高自己的标准。

提高标准

我听到你说，“美工负责制作适合新设备的图像。”美工显然会表示，她需要投入多一倍的时间，重新制作这些内容，将其提升至更高标准。

总时间=原先时间+新投入的时间=1500-2000小时

下面来看看另一张图。

你可以明显看到分辨率高低所存在的差异。图像2源自72 DPI的源文件，而图像3的是300 DPI。图2的质量很糟糕，画面很模糊。

将所有资产由图2标准提升至图3意味着你需要重新制作所有内容。这需要优秀美工投入上百小时的时间。这会给你的钱袋带来沉重负担，同时也会挫伤美工的积极性，因为她需要针对相同内容进行返工，而非制作不同内容。

而由图3转变至图2只需调整相应尺寸。下述图反映的是若干我们能够植入300 DPI源文件的平台图像效果（游戏邦注：作者在这里假设苹果将推出支持Retina技术的iPad）。

300dpi-results(from gamasutra)

降低标准

当美工都瞄准高品质分辨率时，所呈现的细节内容就会非常清晰、轮廓分明。选择降低标准会牺牲众多细节内容。其中技巧在于牺牲非关键细节内容（游戏邦注：那些无法定义图像本身的细节）。

需牢记的一点是，即便美工基于高标准制作内容，他也需要查看最终成果在其他分辨率载体中的表现情况。美工需在基于预期比例制作游戏画面的过程中预先检查内容。

即便是处理矢量图像（其中提高或降低图像标准不像试映图那般费劲），记住线性图像也会相应缩放。若你的图像原本轮廓分明，降低图像标准会让细线消失。若此线性图像旨在“定义”你的插图，那么你就需要付出额外劳动，加粗这些线条。

细节必不可少

植入众多细节内容的前提是你打算进行长期投资。多数时候，制作极高品质的内容毫无意义。因为这可能拖垮你的软件运行速度，导致文件受损，硬盘驱动器的空间很快就趋于饱和。

不妨想想资产在最初媒介的实际尺寸。思考通过投入更多时间换得更多细节内容是否值得。

运用其他资产

记住若你想要将游戏资产运用至其他媒介，例如海报、收集册子、DVD印刷封面、T恤或杂志广告，你就需要采用高分辨率的内容，否则它们看起来会很糟糕。没有什么情况比优美画面基于低分辨率载体更糟糕。

辨别组成游戏的关键资产，例如商标、音乐、视频和角色，然后瞄准高品质标准。

总结

首个关于资产的惯例可以归结成：

降低标准比提高标准更简单。着眼于未来的资产值得投入成本。

试着分析自己的项目，寻找质量、数量和资产尺寸间的平衡点。寻找游戏成本和先进技术的黄金比例。不要将自己局限于当前所瞄准的设备。在将项目移植至各种平台的过程中，高品质的资产能够帮你省下许多资金和时间，你因此能够将这些资源运用至优化游戏内容等方面。（本文为游戏邦/gamerboom.com编译，拒绝任何不保留版权的转载，如需转载请联系：游戏邦）

Managing Assets – Part 1: Setting Asset Quality

by Maria Sifnioti

I’m Maria, and I’m the Associate Producer for Total Eclipse, a game development studio based in Thessaloniki, Greece.  We’re a small studio, with a core team of five. We have been publishing some articles on production, game design, and more on our blog. We will reproduce them here at Gamasutra. We hope you find them interesting and useful!

[This article was originally published in the Total Eclipse blog]

Today I’m going to talk about the way we work with assets here at Total Eclipse. This first post on asset management will discuss quality, forward planning, and balance.

Assets are everything that makes up your game, apart from the source code/script files (managing source code is a different topic and will not be discussed now).  Assets can come from a variety of media and different people may be responsible for creating them. For example:

Music: Musicians

Sound effects: Sound Engineers

Illustrations: Artists

User interface: UI designer, Artists

Animations: Artists

3D objects: Artists

Movies: Producers, Artists

Dialogues/Text: Writers & Localisation teams

A lot of different professions, there. All these people have to be aware of, and in agreement with, the following project conventions whenever they create one asset:

1. Asset quality

2. Asset naming

3. Asset backup and storage

The person responsible for informing and enforcing the creators about conventions is either the lead artist/sound/music/writer or, in smaller teams like ours, the producer/associate producer.

Communication and good teamwork are very crucial here. A non specialist (i.e the Producer) has to have a good level of knowledge of the specialist’s (i.e. artist) field of work in order to explain what the project requires.

I’m going to take a look at the conventions by using Maya’s Dress Up as an example. My examples will focus on image assets but what I’m going to say can equally apply to other rich media.

The quality vs. cost trade-off

Maya’s Dress Up, our most recent game for iOS devices, is extremely asset-heavy with thousands of images. Let’s have a look at one of the game’s typical assets, a necklace. After the producer and artist agree on what will be painted, the quality must be established. In images this can be expressed in terms of DPI (dots-per-inch).

* Cost in money: The more DPIs and the larger the image is, the more you will pay for the art, as the artist needs to add extra details in the illustration.

* Cost in time: The more DPIs, and the larger the image is, the longer it will take for the artist to deliver.

The following image shows one of the necklaces done for Maya. Here, in Total Eclipse, we generally stick to 3oo DPI per illustration.

On the left side you can see the level of detail that it was illustrated – at 300 dpi. The right shows the level of detail that would be needed if one would make it at 72 dpi. The chain links are no longer visible nor is the glass glow on the butterfly. Surely that should save a lot of the artist’s time – or not? Let’s suppose we have 500 items.

* Avg time needed for a garment at 72 dpi -> 1 hour (500 total)

* Avg time needed for a garment at 300 dpi -> 2-3 hours (1000-1500 total)

Looks like a ton of hours would be saved if the lower quality is chosen. This is how the necklace looks on Maya on a venerable 1st-gen iPod Touch with a 480×320 resolution:

“Pretty small” you might think. Suppose for a minute that all of the assets were done in low quality. Then Apple goes out and throws an iPad at you with a 1024×768 resolution. And after a bit, it comes up with a new iPhone sporting a Retina display, of all sorts. This doubles the device’s pixel density of each image and effectively requiring an 960×640 resolution.

If you wanted to move to those new devices, you would need to scale upwards.

Scaling Upwards

“Artist”, I hear you say. “Make me images for this new display”. The artist will rightly tell you that she’d have to work double the time and re-do all of the art to a higher quality.

Total time = Old time+new time = 1500-2000 hours

Have a look at the next image. Click on it to see it full-size.

You can clearly see the difference between scaling up and scaling down in resolution. Image #2 is derived from a low, 72-DPI source file while #3 from a 300 DPI one. The quality level in #2 is bad, the image very blurry and non-crisp.

To go from #2 to #3 for all your asset library roughly means doing all your assets again. It would take a good artist hundreds of hours of work.  This would put a burden to your pocket and probably to the artist’s morale as she’d have to do the same things all over again instead on working at something different.

To go from #3 to #2 is a matter of resizing to the dimensions you need.  The following image shows some platforms that we could export to from one 300 DPI source file. Don’t worry, this is not a sneaky announcement for iPads with Retina display (although, you never know).

Scaling Downwards

When the artists draw at the high quality resolution, details seem crystal clear and fine line-art shows. When you scale it down you must sacrifice a ton of detail. The trick is to sacrifice non-crucial detail -  detail that does not define the image itself.

One very important thing to remember is that even though the artist might draw at a high resolution, she must often check how the end result looks like in other resolutions as well. Previewing the work while it’s being created, on the intended medium scale, should be an integral part of the artist’s task.

Even when dealing with vector graphics, where scaling up or down is much less painful than raster,  keep in mind that line art will scale accordingly. If you have fine line-art, scaling an image down will make thin lines disappear. If this line-art actually ‘defines’ your illustration, then there’s extra work to be done, thickening those lines.

Detail Overkill

Put only as much detail as you’re prepared to pay long-term. Most of the times, there is no point dealing with assets with a ludicrously high quality. The software that you use might get slow and unwieldy, documents may crash, hard drives fill up faster than you’d like.

Think about what the actual size of the asset will be on your primary medium. Consider if the slower completion times are worth the steep increase in detail.

Using assets elsewhere

Keep in mind that if you want to use assets from your game to other mediums, like posters, collector’s booklets, printed DVD covers, t-shirts or a magazine advert, you will need high-resolution assets, otherwise it will look bad. And there’s nothing worse than beautiful artwork printed in bad resolution.  Or wait, there is: Bad artwork printed in bad resolution. (Here’s the antidote to that link).

Identify the key assets that make up your game such as the logo, music, video, and characters and invest towards higher quality.

Conclusion

The first convention about assets boils down to this:

You more easily scale down, than scale up. Future-proof assets justify the cost.

Study your project carefully and strike a balance between quality, quantity and scalability of assets. Find your game’s golden ratio between cost and future-proofing. Do not limit yourself to the current device you are building on.  When porting a project to different platforms, high-quality assets will save you both time and money which you can then spend to properly polish the game for each platform.（Source：gamasutra）