Cosa rende un gioco "buono"? [chiuso]

Sto lavorando a un gameplay 3D utilizzando OpenGL e C ++ e sto pensando di concentrarmi sul dare al gioco una grafica moderna e accattivante, ma più ci penso e più mi rendo conto di non sapere davvero cosa rende la grafica "bene". Certo, posso andare a giocare ad alcuni giochi AAA ben noti e crogiolarmi nella grafica incredibilmente messa insieme, ma non so davvero come la grafica abbia un bell'aspetto. (ecco perché considero i giochi un'arte!)

Questo è quello che posso pensare ora:

• Texture di alta qualità
• Modelli di alta qualità
• Un buon modello di illuminazione
• Bumpmapping + mappatura della specularità
• Interfaccia utente di alta qualità, se applicabile
• Una ricchezza di posteffetti non esagerati

Lo sto chiedendo qui nella speranza di uno sviluppatore di giochi esperto che abbia prodotto giochi e sappia come funzionano dentro e fuori, può spiegare alcune tecniche che garantiscono che la grafica di un gioco abbia un aspetto "buono", e alcuni suggerimenti stravaganti poco noti. Sarebbe fantastico.

Grazie!

Ciò può sembrare sorprendente, ma in realtà non lo è: più si aumenta la risoluzione grafica, il conteggio dei poligoni, la profondità del colore e quant'altro, più è facile che il tuo gioco appaia terribile. Il componente cruciale per godere di un'opera d'arte audiovisiva è la mente : la tua mente sa che stai guardando uno schermo piatto. Sei disposto ad accettarlo e non esserne infastidito, e invece costruisci l'immagine mentale del mondo che viene rappresentato nella tua mente.

Ora, se le immagini sono altamente stilizzate, abbozzate o pixelate, la mente è sintonizzata su questo e su fattori che vengono definiti "sfondo" - non ostacola la costruzione mentale della scena. Ecco perché un gioco come l'originale Monkey Island, con la sua risoluzione di 320x200, può cavarsela con un'animazione della porta che consiste di solo due fotogrammi: accettiamo già che la scena è una taverna pirata e non abbiamo problemi ad accettare che la porta solo ha aperto.

Ma se hai un'alta risoluzione e vuoi optare per il fotorealismo, le aspettative mentali aumentano drammaticamente. Non è più ovvio che ci troviamo di fronte alla narrativa visiva e la mente fa molto meno del lavoro di ricostruzione interna. Più realistico è l'input, meno l'immaginazione è al lavoro riempiendo i dettagli. L'effetto collaterale (forse non intenzionale) è che tu come designer devi ora andare e produrre tutti i dettagli nella grafica reale e nella modellazione 3D. Hai esternalizzato con successo il lavoro fantasioso dal tuo consumatore e hai firmato il tuo team di progettazione grafica per l'attività.

Uno stile visivo inadatto e il fascino possono essere enormi problema. Se hai mai sentito che un gioco assomiglia più a un foglio di calcolo Excel che a un gioco per computer, sai che il design visivo è andato storto.

Di conseguenza, non dovresti cercare di puntare al realismo quanto a uno stile visivo coerente. Il realismo non ti dà l'immersione. Al contrario, rende molto più difficile raggiungere l'immersione. Al contrario, avere uno stile visivo unico e memorabile potrebbe essere molto più semplice da produrre e potrebbe rendere il tuo gioco riconoscibile, memorabile e divertente.

Ti stai concentrando molto sugli aspetti tecnici, ma sembra che ti dimentichi totalmente di qualcosa che è molto più importante di qualsiasi funzionalità del motore grafico:

Il design!

• Modelli e trame che hanno un aspetto unico, ma seguono ancora temi coerenti
• Scenari che sembrano diversi, ma che si adattano ancora insieme
• Tavolozze di colori ben scelti che supportano l'umore del gioco
• Illuminazione utilizzata per impostare e migliorare l'umore di ogni scenario

The player won't care about hyperlinear-ultra-multipass-highdef-anisotropic-super-occlusion-dynamic-parallel-megashader-rendering, when you use it for ugly, uncreative models in a boring environment.

On the other hand, look at all the independent game studios which create truly beautiful games with minimalistic graphic engines.

Cohesion? I think that's most crucial. A certain technology certainly isn't going to make a game look good, there are countless examples of games with, in its time, state of the art tech, but they didn't look good.

When you say "high quality", what defines a quality model is subjective.

If you had written "high resolution textures" or "high polygon-count models" I would say no, that's not necessary, and not a sign of quality.

Most of this will come down to artistic ability, working within your constraints and still making something look good is truly art.

Some of the most gorgeous games do not adhere to any singular principle of what's 'good' and 'not good'. From what I've seen in both AAA and indie game space is an extreme amount of focus placed on style, consistency and theme. To elaborate:

This covers anything from the color palette chosen, level of detail, the choice between richly rendered 3-d assets to painstaking pixel-art. When you see an out-of-context screenshot from games you find visually striking, you can often recognize the game immediately. This is because the artists responsible took the time not only to define a visual style in which to create everything, but have refined it to a point where it can be nothing else besides what it is. On the flip side, games that have a muddy or nonspecific visual style do not strike the same chords with us visually.

Examples of striking visual style:

• Braid (2-d pixel/painted art and backdrops. Visible brushwork on both canvas and characters. Fluid yet simple animation. Saturated colors that tend to bleed into one another, contributing to the 'dreamy' feeling of the levels.)
• Borderlands (3-d with painted textures, hard silhouettes/outlines, exaggerated proportions, grimy and dusty metallic surfaces, and lots of harsh lighting)
• Fez (2-d pixel art. A very mild color palette that balances vibrant color with pastels and muted primaries. Geometric blocks/tiles, and lots of symmetry)
• World of Warcraft (Simplistic but extremely refined 3-d models with richly painted textures. A clear distinction between color palettes in reference to different zones, cities, racial consistencies and themes of building construction and apparel/accessories. Large feet/shoulders, distinctive posture and excessive attention paid to communicating through body language)

In order to achieve these kind of results, an artist needs to be able to envision not only the end-vision of the style in question, but how it can be used to complement gameplay and add to the user experience. The less artists at your disposal, the more diverse a skillset they will require. A single artist creating the visual assets for a 3d title will need to have a firm grasp of the following and more:

• Construction and animation of 3-d models
• 2d image projection/texture mapping creation
• An understanding of architectural style and settings
• The ability to build a 'scene' from foreground to background and everything in-between
• An understanding of setting, mood and how to extract the desired emotional impact of both the world and the characters that live in it.

I am only scratching the surface here, but that's because to attempt to apply something so varied as art into silos of 'good' and 'bad' is complicated.

You need skilled artists.

All the technical stuff can help a bit to make the work of the artists better appreciable, but the truth is that you can get a perfectly capable rendering engine for free or very cheap, and a better engine will only increase the quality marginally.

To begin, I think it's important to point out the obvious. A good looking game does not necessarily mean it has good graphics. And one doesn't have to look any further than Nintendo to see how it is done.

Take Wii Sports for example. One of the most successful games of our generation and it has horrible graphics on a purely technical level. However, the look of the game is by all accounts a tremendous success. So what did it do right?

In my personal opinion, it's because the target demographic interested in the game bought into its presentation. They expected a certain product when they picked up the title, and the title delivered on their expectations if not surpassing them. To me, that's the fundamental point. It has to touch the player somehow and how to go about achieving this will depend on who you are targeting.

That's the reason when I picked up Playstation Allstars for my PS3 recently, I was disappointed by the "graphics". I was expecting it to look more like Smash Bros, but instead everything looked quite cold and bland in comparison. Not only did it not meet my graphical expectations, it did not excite any new emotions, so I just ended up being disappointed. (To be fair, the gameplay is fun).

I would also use this argument with Diablo 3. Technically a fabulous looking game in nearly every aspect. However, it still disappointed many core fans because it changed the dark tones that were so prevalent in the first and second iteration.

Good graphics is just like any other art form. It's more than the sum of its parts and has to touch on the player's emotions at that specific point in time. Thus, I'd argue that there is no secret sauce to it because our tastes are constantly changing over time. What looks good today may not look so good tomorrow, yet looks good again the day after that.

A game with lambert-shaded polygons or wireframe models can look good! All kinds of games have looked good, historically, and still do.

An old arcade machine running Battlezone looks good.

The main thing is for everything to be robust, highly playable and not exhibit any flicker or tearing (use double buffering or blitting of whole frames from an off-screen buffer, ideally synchronized to the vertical retrace of the display device).

Basically you have 2 things:

1) Art

2) Effects

Art is all about just putting the world exactly as the artist envisioned it. This means having a killer artist who can design a cohesive world with fantastic looking textures. As long as you have enough texture memory, you can make a scene exactly as the artist envisions it / authors it in a 3d modelling package.

Effects. These are the real-time effects you lay on top of the already awesome art, like your standard particle engines and bloom effects and better looking execution of the artists vision using things like order-independent transparency.

Consistency.

It's not the only thing that matters, of course, but it's very jarring if there's not a consistent point-of-view throughout a game. People have made great games with all level of graphic quality. Look at Kingdom of Loathing; a great web-based game that is consistent in its graphic quality (low) and humor (even lower). World of Warcraft has a graphic style that has been consistent since the old RTS games (at least from Warcraft II, the original had a very different style). Braid had that beautiful hand-drawn painting style.

In other words, people may buy into your vision or they may not, but there should be a consistent vision.

1) The models

2) Clearly and decidedly SHADING. Programing crazy shaders will give you exeptionnal graphics.