FIT supports all common web image formats for input and output:
Whether an output format is considered suitable for a client depends on its capabilities according to the Delivery Context and – of course – the format of the image being processed.
JPEG and PNG images are ubiquitous on the web and can usually be reduced in byte size by FIT.
For supporting clients WebP will often be the preferred format as it compresses very well both lossy and losslessly. SVG images are passed unchanged to supporting clients, to preserve the vector features whenever possible. They can be reduced in size if SVG minifying is enabled.
GIF is rarely used as output format. It is the oldest format and the more recent ones usually provide better results.
In addition to the above well-known formats, FIT offers ZorroSVG as efficient output format for partially transparent PNGs. A ZorroSVG image is an SVG image that includes a partially transparent PNG bitmap converted to JPEG. As JPEG lacks an alpha channel, areas supposed to be transparent will be masked by SVG alpha compositing. The resulting SVG is often much smaller than the corresponding PNG.
The image compression feature can considerably reduce the transfer size of image data without perceivable loss of quality. Multiple experiments are run to find the optimal output format for an image.
With image scaling enabled, FIT scales images according to the client’s display dimensions. To further reduce the transfer volume, images are compressed using a given quality setting.
If both features are disabled, FIT will simply pass images unchanged.