游戏邦在:
杂志专栏:
gamerboom.com订阅到鲜果订阅到抓虾google reader订阅到有道订阅到QQ邮箱订阅到帮看

开发者阐述创造基于加速计手机游戏的过程

发布时间:2013-02-25 14:43:36 Tags:,,,,

作者:Gyorgy Dragon

引言

你是否听过鞋商前往非洲的老故事?没有的话我现在讲给你听:“两个推销员前往非洲想在此打开全新的鞋子市场。在到达非洲两天后,其中一名推销员打电话给总部说道‘我将尽快回到美国,在这里根本不可能卖出鞋子!因为这里几乎所有人都是赤脚走路。’而在同一天,另一名推销员给工厂发了一封电子邮件说道:‘在非洲卖鞋具有无限大的前景。因为现在这里还没有多少人在穿鞋。”

在手机设备领域分析这一故事便意味着,使用从未见过的技术可能意味着巨大的成功,也有可能等同于惨痛的失败——而不管是成功与失败都是基于我们的看待方式。

手机开发世界拥有大量开发商:有些公司尝试着去创造“终极”应用,“能够带来百万收益的”游戏,也有些公司只希望能够维持每日生计。想出新理念不仅非常困难,而且充满危险,因为几乎所有理念都被前人探索过了,甚至有些理念是因为没有意义而被遗留下来。

我们还是一家新公司,并且是由在娱乐产业中打拼多年的游戏产业资深人士所创建的公司。我们决定通过整合新技术到手机用户的设备上而带给他们全新的体验。我们的首件作品便是基于创新方式而使用加速计,呈献给用户全新的发现体验——例如现今很少用到的某一手机功能。这款游戏便是《Fun Steps》。

第一步便是创造乐趣

这一应用包含了一个动态检测器,以不同音效(不管是流浪汉拖拉声还是机器人抨击声)去替换人们正常走路的声音。我们的目标是让人们的行走能够带些幽默感或惊悚感(游戏邦注:就像电影《警察学校》便利用了Michael Winslow的超强口技)。最终结果也比预期好多了,我将在之后进行详细解释。作为《Fun Steps》项目的首席程序员,我将帮助你理解基于动态检测器和相关技术原则创造游戏的难度与益处。也许某一理念非常复杂,但是隐藏于该理念背后的技术却是值得我们去开发的。

Fun Steps(from gamasutra)

Fun Steps(from gamasutra)

快速和精确的脚步检测——首个挑战

我们可以从任何商店中下载各种脚步计算应用,所以有人便会认为检测脚步是一种很常见的功能。但现实是,这一特殊任务让我们费劲了心血,因为实时动态检测器比计算脚步复杂多了。

在脚步计算程序的例子中检测实际脚步非常简单:不需要获得脚步落地的具体时间。而在我们自己的例子中,应用最重要的部分在于我们会在碰触的最近距离触发声音,即当用户的脚碰到地板时。为了实现这一想法,也因为预料中的处理延迟,我们必须避开复杂的的算法。所以我们决定使用简单的方法去检测脚步。并检测到amplitudo的改变与绝对阀值息息相关。虽然这种检测方法在一开始会引起种种问题,但是在进一步完善后它便能带来许多好结果,不过前提是我们必须做出合理的界面选择。不同用户的脚步力度不同,,所以应用能够检测到用户的高矮胖瘦以及他们所穿的鞋子的材质。

Android上一些待解决的问题——第二个挑战

由于技术原因,音频延迟是Android上最大的问题,而这一问题也变得越来越严重。不同制造商使用不同质量的加速计也带来了各种麻烦。我们发现一些具有非常棒的功能设置的设备能够不断探测到其它设备的移动——即使是在非移动状态下!然后我们创造一种测试应用去降低加速计的噪音。最终检测结果非常极端,即我们在不同手机设备间经历了200%至300%的不同。

我们决定搁下敏感度去获得噪杂设备的更全面的结果。同时还出现了另外一个问题:有时候测量值所呈现的是一些难以相信的事实:有些制造商的设备上的扬声器会对加速计带来负面影响。这并不是一种机制反馈,而是某种程度上的电磁畸变。我们所测试的一些设备只能够检测到2G的加速(因为其扬声器的声音)。除此之外也有一些知名的品牌产品也会出现这种情况。

因为这种建造误差,这些设备的用户的唯一解决方法其实非常简单:使用带有耳机的产品,否则游戏便没有了任何乐趣。

乐趣作为一种测试元素

我们很难去测试这样的应用。因为必须将其放在口袋里,所以我们很难去调试系统或采用其它解决方法进行调试。然后我们决定耐心地记录来自加速计的数据——我们记录了各种参数,并将在之后阶段进行分析使用(当然也模拟了行走)。在开发的第一阶段完成这些工作其实已经足够了,但是在最终调试时我们需要离开办公室去娱乐其他人,让这些人能对我们的游戏内容感兴趣。

成功是有趣的!

在经过多次尝试与测试后,我们的产品渐渐成形,我们不仅对问题的解决方法感到自豪,同时也很满意游戏所呈献给玩家的全新体验。甚至这些体验比我最初所期待的有趣多了。

我们将测试版本提供给亲戚以及各种媒体,希望能够获得最有价值的反馈:

“我与好友一起玩了《Fun Steps》,最后我们都乐得在地板上直打转。”

“一开始我将其放在口袋里而散发出一些古怪的声音。这已经很有趣了,但是后来当我使用一个小小的蓝牙扬声器后,我发现真正的乐趣其实刚刚开始!”

“当我1岁半孩子坐在我的脖子上,然后我们开始四处走动时,这种体验真的非常有趣,就好似青蛙或小丑一般。他也一直咯吱咯吱地笑着。”

“在派对上我们呈现了几份《Fun Steps》,人们组织了一次特殊的‘乐趣跳跃’——不只有声音,还有各种类型的‘舞蹈’,并最终创造出了非常有趣的音效。”

“我是一个非常胖的人,所以走路对我来说就像是拖着一件6吨重的物体在行走,别人会认为我拥有一双机械腿。这真的很有趣。”

在经过各种问题以及开发阶段的多个不眠之夜后,这些反馈真的让我们又兴奋又自豪。人们也可以是出于爱去使用这一游戏,就像我们在《Fun Steps》的官方视频中所看到的那样,一个书呆子想要征服一个非常美丽的女生,这是很常见的情景,但是他所使用的不常见的方法便是,《Fun Steps》!

Fun Steps screenshot(from gamasutra)

Fun Steps screenshot(from gamasutra)

手机设备仍然拥有无限的探索潜能

技术的大爆炸推动着今天的手机设备能在眨眼间便出现一些新功能,所以这些设备总是拥有许多未被发现或使用的功能。完全使用这些功能并不是一天两天的事,这需要先驱者的开拓以及我们投入更多努力。

那些引导了手机革命的公司(游戏邦注:如苹果或谷歌)必须坦诚地接受过后这些改革努力——我们永远不知道哪些实验项目将带来新的“惊喜”。而所有的实验都有可能拓宽应用的潜力并挖掘出手机设备使用的各种可能性。

我曾收到一封来自美国俄亥俄州的用户的电子邮件,他叫Timothy,尝试着训练自己的小狗去识别《Fun Steps》中的不同命令,即当主人想要去散步,或者小狗需要执行某些任务,如慢步或接近主人等。这是一种完全不同的体验,我们甚至从未想过这种可能性。当然了,也许这些理念或实验不一定可行,但是我们却能从努力中有所收获。所以我们鼓励所有玩家去推翻现代手机设备的界限,要知道可能性始终是无穷巨大的。

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

Making accelerometer-based mobile games/toys – the Fun Steps experience

by Gyorgy Dragon

Introduction

You know the old story about the shoe-vendors, going to Africa? No? Here we go: “Two salespeople were sent to Africa to open up new shoe markets. Two days after arriving in Africa, one salesman called the office and said, “I’m coming back to the U.S. on the next flight because I can’t sell shoes here. Everybody goes around barefoot.” On the same day, the other salesman sent an E-mail to the factory saying: “The prospects in Africa are unlimited. Nobody wears shoes here!”

Translating this small story to the field of mobile devices means, that use of technology never seen before can result in huge success or a big failure – depending on the possibilities and the point of view from where we are looking at them.

The world of mobile development has plenty of developers: lots of companies are trying to make the ?ultimate” application, the ?million-selling” game, or just simply earn the day-by-day living. Coming out with new ideas is not only hard but dangerous, as nearly all ideas has been explored before – and maybe some of them just left behind having no sense at all.

We are a newly established company made by game industry veterans with plenty of experience in the entertainment industry. We decided to give mobile device users brand new experiences using the possibilities and the technologies incorporated in their devices. Our first family of products uses the accelerometer in an innovative way, giving newly discovered experiences to the users – for example that given feature of the mobiles is rarely used nowadays.  The name of the family of products is Fun Steps… and as a small demonstration here we go with the promotion video of it: (the important part begins at 00:30 ) http://www.youtube.com/watch?v=8ZLtpvMUc0Q

First step having fun

This app covers a motion detection toy that is used to replace the normal walking sound of a person with different sound effects from dino crushing to robot bashing. Our goal was to make human walking humorous or scary (like in the Police Academy movies using the incredible voice talents of Michael Winslow). The result is better, than expected: we explain it later!As the lead programmer of Fun Steps project I help you to understand the difficulties and the happiness to make a game based on motion detection and on technological principles. Maybe the idea is not very complicated, but the technology behind is worth the development.

Quick and accurate step detection – the first Challenge

You can download a lot of step counting applications from different stores, so anyone can think, that detecting a step is a well-documented feature. The reality is that this particular task made us sweat a lot, as motion detection in real time is a lot more complicated then counting the steps.

In the case of a step counting program it’s enough detect the fact of the step: it’s not necessary to obtain the exact timing of step. In our case the most important part of our application is that we start the sound as close as we can, when the user’s leg touches the ground.For realising our idea we could not use complex algorithms, because of the expected delay of the processing. So we decided to use a simple method to detect the step. It detects the change of the amplitudo that is relative to an absolute threshold. This selected approach caused some problems at the beginning, but after further refinement it can produce a lot of good results, but requires a sensitivity option on the interface. It can be adjusted by the user to balance the exact force that is detected by the system as a step, so the application is able to detect the steps of small, big, fat or skinny people using whatever shoe they use from softer ones to hardest ones.

Some problems to resolve on Android – the second challenge

For technical reasons the audio latency is bigger issue on Android, so the problem persisted heavily.Also a lot of trouble has been caused by the different quality accelerometer parts used by different manufacturers. We discovered that some devices’ well-functioning settings are continuously detecting movement on other devices – even in non-moving state! Then we developed a test application to merit the noise level of the accelerometers in a quiet state. The results where extreme: we experienced 200-300% difference between mobile devices.

We decided to throw away sensitivity for getting more rounded results on noisier devices.Another issue emerged also: it was hard to detect and when the measurements showed the facts it was hard to believe: some manufacturer’s devices speakers caused serious problems with the accelerometer of that device. And this is not a common mechanical feedback, but some electromagnetic distortion. For example some devices tested by us were able to detect an acceleration of 2G only because for their speaker’s voice. Without telling actual models and names I can tell, that some well-known brand’s products were also included producing that kind of phenomenon.

Because of this construction error the only solution for the users of those devices is really simple: use the product with a headset or the game will be of no use – we call it personal fun!

Fun, as a testing factor

It’s not easy to test such kind of application. As it’s necessary to put it on the pocket, it’s really hard to debug the system, or to experience with different solutions. We then decided to begin to walk up and down patiently to record the data from the accelerometer – we recorded the metrics to use and analyze them later on (and to simulate walking, of course). It was enough in the first phases of the development…… but during the final tuning we needed to walk great distances in our office making other people laugh and confusing our poor house animals and pets (not any of them has been harmed during development.

Success is fun!

After so many try-outs and testing the product took shape and all of us were very proud of it not just for the solution to the problem, but mainly because for the very new experience we created to the players. And that experience is a lot more fun as we originally expected.

We gave our test versions to our family members and to the members of the press and got some really incredible feedback:

“When I’m with my friend and we start Fun Steps we end laughing on the floor.”

“Firstly I used it in my pocket emitting some scary sounds. It was fun, but then I got a small Bluetooth speaker and the REAL fun begun!!!!”

“It’s so funny when I sit my 1.5 year old children in my neck and we begin to walk around, like a frog or a clown. He is so cute laughing!”

“in a party, where several copies of Fun Steps were present the people organized a special Fun Jumping compo – not only the voice but a ‘kind of choreography’ was important too resulting in very funny sound effects”

“I’m a big, fat guy, so walking like a 6 ton mech was no concern to people – they thought I have mechanical legs. It was fun”

All those feedbacks made us very-very excited and proud after all that problems and sleepless nights the development caused.Fun Steps is a very nice toy. And it can be used even for love purposes, as seen in the Official video of Fun Steps: Beautiful girl, nerdy guy trying to conquest her – quite usual situation, but quite unusual solution with the use of Fun Steps!. As we like our video, we call your attention to it the second time – the video can be watched at: http://www.youtube.com/watch?v=8ZLtpvMUc0Q

Mobile devices still having a lot more potential to explore

The technological explosion leading to today’s mobile devices happened very short time ago, so those devices still possess a lot of undiscovered, or not used features (don’t forget, user groups still find undocumented features for the Commodore-64 still in 2012!!!). To full spectrum use of those features is not happening from day to day – it needs a lot of pioneers and a lot of effort.

The companies, leading the mobile revolution (like Apple or Google) must be open minded for those revolutionary efforts – it’s never known which experimental project will be the new “wonder”. And any of those experiments can find a common use broadening the palette of the applications and the possibilities of the use of mobile devices.

An example for the stories we got from our player:I got an email from Ohio, US, where a user, called Timothy trained his dog to recognize several Fun Steps voices as different commands. For example if they went to walk, or the dog needed to do some task, like go slowly, or near to it’s owner etc… This is such a different experience, that neither we visioned it as a possibility.Of course, it must be some unsuccessful ideas or experiments around, but the success compensates us for all the efforts we made. We encourage all developers to pushing the boundaries of the modern mobile devices, because the possibilities are nearly limitless.If you have any questions regarding the article, contact us:info@pocketscientists.com(source:gamasutra)


上一篇:

下一篇: