Drag & drop your images anywhere
Drag & drop your images anywhere
Online Image Tool supports multiple image formats using the most modern and best-in-class compression algorithms based on each individual image. In many cases, this translates to image compression of over 90% on standard images.
Want to learn more about compression? Please check out the following sections for more details:
PNG images are compressed using a chained series of compression algorithms, creating image files that are fully compatible with all web browsers, and reduced by up to 90% of the original size.PNG Compression Step 1 - PNG Quant(ization)
Images are compressed using a modified and optimized version of the pngquant algorithm leveraging a neural network capable of instantly recognizing the optimal parameters tailored to the particular image’s individual characteristics. This process produces a lossy compressed version of the original image, reducing the file size by up to 90% while maintaining resolution and no perceptible change to the image quality.PNG Compression Step 2 - Opti PNG
Once the image has passed the initial pngquant compression cycle, the image is further compressed using a modified version of the optipng compression algorithm producing a losslessly compressed version of the new image. This will typically result in a further 10% reduction in the size of the intermediary image.
PNG images are compressed using a chained series of compression algorithms, creating image files which are fully compatible with all web browsers, and reduced by up to 90% of the original size.
Compressing PNG images using Online Image Tool provides you with the following features:
PNG quantization works by reducing an image’s palette (the number of unique possible colors within a given image) from 16,777,216 to 256 (24-bits to 8-bits per pixel). Online Image Tool fully supports transparent PNG files.
Many images contain more than 256 colors, however, in most situations the human eye is not able to tell the subtle differences between very similar colors within the same image. Each minuscule difference in color requires data to define within the palette. This results in unnecessarily large file sizes to store what essentially amounts to imperceptible color differences.
PNG quantization identifies the “best” unique 256 colors to use as a palette for a given image and then “nudges” all other outlying pixels to fit the best and closest color within that reduced palette.
Can your eyes see the difference between these 2 PNG images?
For virtually all PNG images, especially photographs, the human eye will not be able to detect any difference in quantized images. however, there are caveats.
Images that contain heavy gradients that span a wide range of colors will create perceptible reductions in quality. This is due to the fact that the subtle changes in color gradients within the image (the intended effect of the gradient itself) will become apparent if the 256-color palette runs out of colors to hold.
The human eye has the incredible capability of distinguishing between 7,000,000 colors, however, our brains can only pick up on these subtle differences in color on pixels that are more or less adjacent to each other.
The example below illustrates the perceptible reduction in quality that color quantization will produce on an image with high and wide-ranging color gradients. Here, we can see it is apparent that limiting an image to 256 colors will not produce the desired results.
Quantization of images with heavy gradients may not always produce good results:
JPEG images are compressed using a WebAssembly build of the MozJPEG 3.0 library - a project from Mozilla considered the gold-standard of JPEG compression algorithm available today - creating image files which are fully compatible with all web browsers, and reduced by up to 90% of the original size.
Online Image Tool uses MozJPEG to smoothen values in the Scale quantization tables within a JPEG image.
The amount of smoothening used over these quantization tables carries a large influence on the degree of how much compression can be achieved and can be viewed as the “quality factor” of the image set as a percentage of 0% (worst) to 100% (best).
Online Image Tool chooses quantization values optimized to discard the high-frequency noisy details in which humans eyes are less sensitive to perceiving any loss of information.
Empirical data shows that setting the quality factor to 80% has produced the best balance between data compression and preservation of visual quality in JPEG images with virtually no perceptible loss in quality and up to 90% in file size reduction.
Can your eyes see the difference between these 2 JPEG images?
Compressing JPEG images using Online Image Tool provides you with the following features:
GIF animations are compressed using a tweaked version of Gifsicle configured with presets designed to optimize GIF file sizes.
Compression is achieved using the following 3 approaches:
Can your eyes see the difference between these 2 GIF animations?
Compressing GIF images using Online Image Tool provides you with the following features: