Frequently asked questions

Is there a batch function available?

As I am trying to get most out of the images it does not make much sense to me to work without a visual control mechanism as every image needs different parameters to get most out of it. So, no, at the moment there is no batch tool available.

Does a Wordpress plugin exist?

A WordPress plugin is on my todo list for a loooong time but unfortunately I can't say when it will see the light of day.

Does an offline version of COD exist?

No, sorry.

Why can I only upload images up to a size of 8 MiB?

I invest a lot of CPU power in most cases to get the most out of the images. And when many users are compressing images at the same time, it means a high CPU load on the servers. With large images, the CPU load increases exponentially. And CPU-strong servers cost money.

At the moment the users who support me via Patreon cover about the server costs. The 30-40 hours I otherwise put into the project are my free time. So it's purely a money problem that keeps me from doing it.

Why can't Photoshop open my compressed JPEG?

As far as I know, the problem only exists with Adobe Photoshop, which is why there is a setting in compress-or-die-expert-mode to make the file output photoshop compatible. The file then becomes minimally larger. So if you want to be on the safe side, activate this setting. However, I have never heard that the problem exists outside of Photoshop.

Your often mentioned values for JPEG quality do not match mine in Photoshop... why?

Photoshop actually uses its own quantization tables and its own scaling values for the different quality values. When I mention values, they actually always refer to the standard quantization tables, which are also used by very many programs, as well as the scaling function suggested by the JPEG specification.

Below is a table that shows how the values relate to each other (determined with Photoshop 2023, Version 24.1). The first value in the "Standard Quality" table corresponds to the luminance value and the second to the chrominance value.

Caution: From time to time Photoshop's quantization tables change slightly, so there may be slight variations.

Photoshop qualityPhotoshop Chroma Subsampling"Standard" quality
121×1~99, ~99
111×1~97, ~96
101×1~93, ~92
91×1~88, ~87
81×1~82, ~81
71×1~70, ~71
62×2~67, ~81
52×2~65, ~73
42×2~53, ~65
32×2~45, ~57
22×2~36, ~50
12×2~25, ~35
02×2~21, ~31
Photoshop "Save for web" qualityPhotoshop Chroma Subsampling"Standard" quality
1001×1~99, ~99
901×1~97, ~96
801×1~93, ~91
701×1~88, ~87
601×1~82, ~80
502×2~76, ~81
402×2~65, ~73
302×2~52, ~64
202×2~44, ~55
102×2~35, ~46
02×2~24, ~31


Christoph Erdmann

Your opinion is so important for me! Don't hesitate to drop me a line. I will do my best to implement your feedback.

Give feedback

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

Christoph Erdmann
Image compression with AI

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.