Skip to content

Latest commit

 

History

History
355 lines (270 loc) · 13.8 KB

README.md

File metadata and controls

355 lines (270 loc) · 13.8 KB

gatsby-plugin-sharp

Exposes several image processing functions built on the Sharp image processing library. This is a low-level helper plugin generally used by other Gatsby plugins. You generally shouldn't be using this directly but might find it helpful if doing very custom image processing.

It aims to provide excellent out-of-the box settings for processing common web image formats.

For JPEGs it generates progressive images with a default quality level of 50.

For PNGs it uses pngquant to compress images. By default it uses a quality setting of [50-75]. The pngCompressionSpeed value is a speed/quality trade-off from 1 (brute-force) to 10 (fastest). Speed 10 has 5% lower quality, but is 8 times faster than the default (4). In most cases you should stick with the default, but if you have very large numbers of PNGs then it can significantly reduce build times.

Install

npm install gatsby-plugin-sharp

How to use

module.exports = {
  plugins: [
    {
      resolve: `gatsby-plugin-sharp`,
      options: {
        // Defaults used for gatsbyImageData and StaticImage
        defaults: {},
        // Relates to "options.failOn" in https://sharp.pixelplumbing.com/api-constructor#parameters
        failOn: `warning`,
      },
    },
  ]
}

Options

  • defaults: Default values used for gatsbyImageData and StaticImage from gatsby-plugin-image. Available options are: formats,placeholder,quality,breakpoints,backgroundColor,tracedSVGOptions,blurredOptions,jpgOptions,pngOptions,webpOptions,avifOptions. For details of these, see the reference guide.
  • failOn: default = warning. By default, builds will fail if sharp finds an image with corrupted pixel values. When setting failOn to none the image will return undefined instead. You can customize this option, see options.failOn. Images with corrupt image headers/metadata will always fail, regardless of this setting.

Methods

resize

Parameters

  • width (int, default: 400)
  • height (int)
  • quality (int, default: 50)
  • jpegQuality (int)
  • pngQuality (int)
  • webpQuality (int)
  • jpegProgressive (bool, default: true)
  • pngCompressionLevel (int, default: 9)
  • base64(bool, default: false)

Returns

  • src (string)
  • width (int)
  • height (int)
  • aspectRatio (float)

fixed

Automatically create sizes for different resolutions — we do 1x, 1.5x, and 2x.

Parameters

  • width (int, default: 400)
  • height (int)
  • quality (int, default: 50)
  • jpegQuality (int)
  • pngQuality (int)
  • webpQuality (int)

Returns

  • base64 (string)
  • aspectRatio (float)
  • width (float)
  • height (float)
  • src (string)
  • srcSet (string)

fluid

Create fluid sizes (in width) for the image. If the max width of the container for the rendered markdown file is 800px, the sizes would then be: 200px, 400px, 800px, 1200px, 1600px – enough to provide close to the optimal image size for every device size / screen resolution.

If you want more control over which sizes are output you can use the srcSetBreakpoints parameter. For example, if you want images that are 200, 340, 520, and 890 wide you can add srcSetBreakpoints: [ 200, 340, 520, 890 ] as a parameter. You will also get maxWidth as a breakpoint (which is 800 by default), so you will actually get [ 200, 340, 520, 800, 890 ] as breakpoints.

On top of that, fluid returns everything else (namely aspectRatio and a base64 image to use as a placeholder) you need to implement the "blur up" technique popularized by Medium and Facebook (and also available as a Gatsby plugin for Markdown content as gatsby-remark-images).

When both a maxWidth and maxHeight are provided, sharp will resize the image using COVER as a fit strategy by default. You can choose between COVER, CONTAIN, FILL, INSIDE, and OUTSIDE as a fit strategy. See the fit parameter below for more details.

Parameters

  • maxWidth (int, default: 800)
  • maxHeight (int)
  • quality (int, default: 50)
  • jpegQuality (int)
  • pngQuality (int)
  • webpQuality (int)
  • srcSetBreakpoints (array of int, default: [])
  • background (string, default: 'rgba(0,0,0,1)')

Returns

  • base64 (string)
  • aspectRatio (float)
  • src (string)
  • srcSet (string)
  • srcSetType (string)
  • sizes (string)
  • originalImg (string)

Shared Options

In addition to their individual parameters, all methods above share the following:

  • grayscale (bool, default: false)
  • duotone (bool|obj, default: false)
  • toFormat (string, default: '')
  • toFormatBase64 (string, default: '')
  • base64Width (int, default: 20)
  • cropFocus (string, default: 'ATTENTION')
  • fit (string, default: 'COVER')
  • pngCompressionSpeed (int, default: 4)
  • rotate (int, default: 0)

toFormat

Convert the source image to one of the following available options: NO_CHANGE, JPG, PNG, WEBP.

toFormatBase64

base64 image uses the image format from the source, or the value of toFormat. This setting allows a different image format instead, available options are: JPG, PNG, WEBP.

WEBP allows for a smaller data size, allowing you to reduce your HTML size when transferring over the network, or to use a larger base64 placeholder width default for improved image placeholder quality.

Not all browsers support WEBP. It would be wasteful to include a fallback image format in this case. Consider also adding a backgroundColor placeholder as a fallback instead.

The plugin config option forceBase64Format performs the equivalent functionality by default to all your base64 placeholders. toFormatBase64 has a higher priority for base64 images that need to selectively use a different format.

base64Width

The width in pixels for your base64 placeholder to use. The height will also be adjusted based on the aspect ratio of the image. Use this to increase the image quality by allowing more pixels to be used at the expense of increasing the file size of the data to be transferred.

The default for Gatsby is 20px. This keeps the data size low enough to embed into the HTML document for immediate display on DOM loaded and avoids an additional network request.

Facebook and Medium are both known to use 42px width for their image placeholders. However Medium presently uses a solid background color placeholder to load the page as fast as possible, followed by an image placeholder requested over the network instead of embedding it with base64.

The plugin config has an equivalent option, allowing you to change the default for all base64 placeholders. This parameter option has a higher priority over the plugin config option.

cropFocus

Change the cropping focus. Available options: CENTER, NORTH, NORTHEAST, EAST, SOUTHEAST, SOUTH, SOUTHWEST, WEST, NORTHWEST, ENTROPY, ATTENTION. See Sharp's resize.

fit

Select the fit strategy for sharp to use when resizing images. Available options are: COVER, CONTAIN, FILL, INSIDE, OUTSIDE. See Sharp's resize.

Note: The fit strategies CONTAIN and FILL will not work when cropFocus is set to ENTROPY or ATTENTION.

The following image shows the effects of each fit option. You can see that the INSIDE option results in one dimension being smaller than requested, while the OUTSIDE option results in one dimension being larger than requested. Sharp transform fit options

pngCompressionSpeed

Change the speed/quality tradeoff for PNG compression from 1 (brute-force) to 10 (fastest). See pngquant's options.

rotate

Rotate the image (after cropping). See Sharp's rotate.

grayscale

Uses Sharp's grayscale to convert the source image to 8-bit grayscale, 256 shades of gray, e.g.

allImageSharp {
  edges {
    node {
      ... on ImageSharp {
        resize(width: 150, height: 150, grayscale: true) {
          src
        }
      }
    }
  }
}

duotone

Applys a "duotone" effect (see I, II, III) to the source image if given two hex colors shadow and highlight defining start and end color of the duotone gradient, e.g.

fixed(
  width: 800,
  duotone: {
    highlight: "#f00e2e",
    shadow: "#192550"
  }
) {
  src
  srcSet
  base64
}

the source image colors will be converted to match a gradient color chosen based on each pixel's relative luminance.
Logic is borrowed from react-duotone.

You can pass a third optional parameter, opacity:

fluid(
  width: 800,
  duotone: {
    highlight: "#f00e2e",
    shadow: "#192550",
    opacity: 50
  }
) {
  src
  srcSet
  base64
}

If set, a semi-transparent version of duotone'd image will be composited over the original image, allowing the original image and its colors to partially "shine through". Heads up: If the original image contains an alpha channel it will be flattened before creating the composite.

This works by adding an alpha channel to the duotone'd image - then we let Sharp do its magic via overlayWith; quoting the Sharp documentation:

If the overlay image contains an alpha channel then composition with premultiplication will occur.

Setting sharp's level of sensitivity to invalid images

By default, the build will fail when sharp encounters an error while processing an image. You can change parts of this behavior by changing the failOn setting to none. In that case sharp will then ignore any errors relating to the pixel values/file structure of your file. However, if your image has corrupt image headers/metadata the build will still fail. It is important to note that any images that would have otherwise failed will not be accessible via childImageSharp until the underlying issue with the image is addressed.

EXIF and ICC metadata

By default, gatsby-plugin-sharp strips all EXIF, ICC and other metadata present in your source file. This is the recommended default as it leads to smaller file sizes.

However, in situations where you wish to preserve EXIF metadata or ICC profiles (example: you are building a photography portfolio and wish to conserve the color profile or the copyright information of the photos you've exported from Adobe Lightroom or Phase One's Capture One), you can set the stripMetadata plugin option to false in gatsby-config.js.

It is important to note that if stripMetadata is set to false, all metadata information will be preserved from the source image, including but not limited to the latitude/longitude information of where the picture was taken (if present). If you wish to strip this information from the source file, you can either leave stripMetadata to its default of true, or manually pre-process your images with a tool such as ExifTool.

Troubleshooting

Incompatible library version: sharp.node requires version X or later, but Z provides version Y

This means that there are multiple incompatible versions of the sharp package installed in node_modules. The complete error typically looks like this:

Something went wrong installing the "sharp" module

dlopen(/Users/misiek/dev/gatsby-starter-blog/node_modules/sharp/build/Release/sharp.node, 1): Library not loaded: @rpath/libglib-2.0.dylib
  Referenced from: /Users/misiek/dev/gatsby-starter-blog/node_modules/sharp/build/Release/sharp.node
  Reason: Incompatible library version: sharp.node requires version 6001.0.0 or later, but libglib-2.0.dylib provides version 5801.0.0

To fix this, you'll need to update all Gatsby plugins in the current project that depend on the sharp package. Here's a list of official plugins that you might need to update in case your projects uses them:

  • gatsby-plugin-sharp
  • gatsby-plugin-manifest
  • gatsby-remark-images-contentful
  • gatsby-source-contentful
  • gatsby-transformer-sharp
  • gatsby-transformer-sqip

To update these packages, run:

npm install gatsby-plugin-sharp gatsby-plugin-manifest gatsby-remark-images-contentful gatsby-source-contentful gatsby-transformer-sharp gatsby-transformer-sqip

If updating these doesn't fix the issue, your project probably uses other plugins from the community that depend on a different version of sharp. Try running npm list sharp or yarn why sharp to see all packages in the current project that use sharp and try updating them as well.