Yacht
Bee
Queenstown

My recent book Pro CSS3 Animation walked the reader through a simple version of a 3D flip image gallery. After publication, I wanted to take it further: thus, this article.

The UI challenge addressed here is the same focused on in many of my other CSS 3D works, such as the Origami UI: we live in a world of increasingly diversified screens, with smaller sizes rapidly becoming the norm. That demands a reconsideration of space: rather than placing captions below photographs, we might want to place them behind the image, to be revealed with a mouseover or tap action.

Creating The Basic Gallery

The markup for the image flip gallery couldn’t be much simpler:

<div class="flip-3d">
	<figure>
		<img src="yacht.jpg" alt>
		<figcaption>Yacht</figcaption>
	</figure>
</div>
<div class="flip-3d">
	<figure>
		<img src="bee.jpg" alt>
		<figcaption>Bee</figcaption>
	</figure>
</div>
<div class="flip-3d">
	<figure>
		<img src="queenstown.jpg" alt>
		<figcaption>Queens-town</figcaption>
	</figure>
</div>

Note the outer div for each <figure>, which will be used to create a 3D context for the gallery content.

To gain the 3D effect shown in the example, we need to do three things:

  • Make each figcaption cover the same area as the responsive image it describes.
  • Reverse that caption, placing it on the other side of the image.
  • Create a dropshadow that will move with the image, without adding any extra markup.

The Initial CSS

First, the basics: we need to manipulate the figure elements in 3D space. We achieve that by placing a perspective property on the figure’s common parent element:

div.flip-3d {
	perspective: 1200px;
	width: 26%;
	margin: 3%; 
	display: inline-block;
}

The <div> elements are placed in the same row with inline-block and provided with even distribution using a combination of width and margin. ( might also be used to achieve this).

To keep the transition of the figure elements as efficient as possible, we define in advance that we will only be tracking the element’s transformation. We also add the frequently-forgotten transform-style property, which will come in handy later.

div.flip-3d figure {
	position: relative;
	transform-style: preserve-3d;
	transition: 1s transform;
	font-size: 1.6rem;
}
div.flip-3d figure img {
	width: 100%;
	height: auto;
}

(A @media query that simplifies the gallery on smaller displays and narrowed browser windows is not shown here).

Positioning The Captions

The captions are placed perfectly on top of the images by using position: absolute (possible only because the <figure> elements are position: relative). Using half a turn (i.e. a 180° rotation) and a tiny translate value places the captions “behind” the images, so long as the transform-style property is applied correctly to the <figure> element. (You might want to take a look at what happens when you remove the property).

div.flip-3d figure figcaption {
	position: absolute;
	width: 100%;
	height: 100%;
	top: 0;
	transform: rotateY(.5turn) translateZ(1px);
	background: rgba(255,255,255,0.9);
	text-align: center;
	padding-top: 45%;
	opacity: 0.6;
	transition: 1s .5s opacity;
}

The text is placed in the center of the caption using a small cheat, and provided with a white background that is half-transparent. It’s completely possible to create a completely opaque <figcaption>, turning the image into a 3d postcard, but I prefer the visual effect of fading in the background. Note that this fade-in will be delayed by a half second so that it starts halfway through the rotation of the figure, which we have already set to last one second.

Setting The Rotations

Now that the gallery has been set up, creating the animations is simplicity itself:

div.flip-3d:hover figure { 
	transform: rotateY(.5turn);
}
div.flip-3d:hover figure figcaption {
	opacity: 1;
}

Adding The Shadows

I was intrigued by the possibility of adding shadows below the images, a goal that was complicated by three issues:

  • The shadows should be created without adding any extra markup;
  • They had to be wide and thin: in other words, not the actual size or shape of the image – eliminating box-shadow and drop-shadow.
  • The shadows had to rotate in their own plane while moving with the image.

The solution I came up with was to use generated content filled with a radial gradient, creating the impression of a shadow “linked” to its originating element:

div.flip-3d figure:after {
	content: " ";
	display: block;
	height: 8vw;
	width: 100%;
	transform: rotateX(90deg);
	background-image: radial-gradient(ellipse closest-side, rgba(0, 0, 0, 0.05) 0%, rgba(0, 0, 0, 0) 100%); 
}

It’s probably easiest to see the final effect in an illustration from a different angle:

Illustration of flip image with projected shadow

That's it! I hope that this might provide a kernel of inspiration for your own uses of CSS 3D.

Photos by Cuba Gallery