4.25 kB

(-78.79%)
20.05 kB
Download
Upload new file

Expert mode

Offers more options to compress your image.

Reset parameters

Preprocessing

× 300 × 300

Size (px)

Defines the dimensions of the resulting image.

Trim

Removes any borders or edges of an image which did does not change in color or transparency.

Trim fuzz (%)

Defines a threshold for colors that are not-exactly the same, but should be treated as being the same.

Rotation

Rotates the image clockwise in 90 degree increments.

Brightness

Adjusts the brightness of the resulting image.

Contrast

Adjusts the contrast of the resulting image.

Blurring radius (px)

The Selective Blur is used to blur a background so it can be better compressed.

Contrary to other blur algorithms, the Selective Blur doesn't act on all pixels: blur is applied only if the difference between its value and the value of the surrounding pixels is less than a defined threshold value. So, contrasts are preserved because difference is high on contrast limits.

Blurring threshold (%)

The Selective Blur is used to blur a background so it can be better compressed.

Contrary to other blur algorithms, the Selective Blur doesn't act on all pixels: blur is applied only if the difference between its value and the value of the surrounding pixels is less than a defined threshold value. So, contrasts are preserved because difference is high on contrast limits.

Compression

Bit depth

8 bit
Creates a PNG with max. 256 colors and transparency.
24/32 bit
Creates a truecolor PNG with transparency if needed.

More in my article Finally understanding PNG (Color depths).

Colors

Specifies the amount of colors used to display the image.

More in my article Finally understanding PNG (Use 8-bit color depth).

Dithering

Dithering is the process by which we create illusions of the color that are not present actually. It is done by the random arrangement of pixels.

Extreme compression

Use more processor power to get a smaller image without affecting the image quality.
Be careful: This is time-intensive and is only used with 8 Bit PNGs.

Quality (%)

Posterizes the image. Posterization entails conversion of a continuous gradation of tone to several regions of fewer tones, with abrupt changes from one tone to another.

Compression view

Here you can see how efficient the compression of this PNG file is. Dark blue means this area could be well compressed and red means poorly compressed. You can use it for different things:

Dirty transparency detection
If you have fully transparent areas that are not shown in dark blue you know that there are visible pixels (perhaps with just 1% opacity) you should delete.
Broken repetition detection
If you have two objects in your PNG that should be absolutely equal but the second one is not shown in dark blue you know that they are not exatly the same. So you are losing much compression potential. Open the PNG in your graphics program to copy the object pixel-perfect. Remember the 32k sliding window imposed by the Deflate algorithm.

You don't like ads? Support Compress-Or-Die and become a patron who does not see ads. Your upload limit gets doubled too!

You want to learn about PNG?

Learn more about PNG compression in my article Finally understanding PNG.

;
News
Christoph Erdmann
Image compression with AI
2023-06-25

Wow! Google researchers have now proposed a new method that combines a standard autoencoder with a diffusion process to recover and add fine details discarded by the autoencoder. Interesting to see the possibilities AI opens up when compressing images.

How Should We Store AI Images? Google Researchers Propose an Image Compression Method Using Score-based Generative Models - MarkTechPost
MarkTechPostA year ago, generating realistic images with AI was a dream. We were impressed by seeing generated faces that resemble real ones, despite the majority of outputs having three eyes, two noses, etc. However, things changed quite rapidly with the release of diffusion models. Nowadays, it is difficult to distinguish an AI-generated image from a […]

Check out our Reddit channel if you want to comment on the news.