Manga2Kindle – A Small Project to Automate Manga Conversion for Kindle

[unn0rm4l]

Hi everyone,

I like reading manga on my Kindle, but dealing with manual conversions, resizing, and sending files was getting tedious. So, I put together a Docker-based solution to automate the whole process. Figured I’d share it here in case anyone else finds it useful.

What It Does:
Syncs with your phone using Syncthing (works with Tachiyomi, Mihon, Yokai, etc.)

Detects CBZ manga files and converts them to a Kindle-friendly EPUB

Sends them directly to your Kindle (so you don’t have to do anything manually)

Has a web interface for tweaking image settings (contrast, black/white balance, etc.)

Automatically cleans up files after sending them

How to Use It:
If you already use Docker, you can run it like this:

docker run -p 32023:32023 -p 8384:8384 \
-v manga_sync:/sync:rw \
-v app_data:/app/data:rw \
unn0rm4luser/manga2kindle:latest
Then, open:

Web UI at localhost:32023 set up your Kindle email, and let it do its thing.

Syncthing UI at localhost:8384 to set up file syncing between your phone and the container

On your phone, install Syncthing and set it up to sync your manga app’s download folder with the container. Once it’s running, any new manga you download will be automatically processed and sent to your Kindle in a set interval.

Just Sharing & Looking for Feedback
I built this for myself, but if anyone wants to try it out, I’d love to hear how it works for you! Also, if anyone has a good image conversion setting for Re:Monster, let me know—I couldn’t get the black/white/shadow balance quite right.

🔗 GitHub: https://github.com/EugeneeT/manga2kindle

That’s it! Just wanted to put it out there. Let me know if you have any questions or suggestions! 😊

[axu2]

What's your opinion of Kindle Comic Converter? It has many tweaking image settings and some people use it in docker. Though obviously your app does a lot more that just convert files.

[unn0rm4l]

Originally Posted by axu2

What's your opinion of Kindle Comic Converter? It has many tweaking image settings and some people use it in docker. Though obviously your app does a lot more that just convert files.

i didn't really know this until i seen it here on this forum yesterday.

I looked for a few options online but couldn't really find anything thats automated so i tried building it myself, since i need to practice docker anyway.

Time is also the reason why i haven't found a good setting for ImageMagic yet and some mangas look way to light. I hope someone can play around a bit and share the right settings that there would be more shadow and more black in the lines and text. Re:Monster is a good example to find a good image setting.

[Quoth]

Why Docker?

[unn0rm4l]

Originally Posted by Quoth

Why Docker?

had to learn docker, and have it running on my home server, i am sure you can tweak it and use something else if you clone or fork the repository.

[Quoth]

Originally Posted by unn0rm4l

had to learn docker, and have it running on my home server, i am sure you can tweak it and use something else if you clone or fork the repository.

I mean why not just run it in a regular terminal?

I have had a server since maybe 1994 at home.

I've used VMs on laptop/workstation for the occasional non-native must-have program.

I've never seen the need to run non-native, sandboxes, docker or any such thing on a server, ever anywhere.

Certainly a Manga2Kindle utility doesn't need anything except a compatible OS and writing crossplatform (Mac/Linux/Windows) isn't hard using Java, QT, Lazarus or many other tools, for console or GUI.

[unn0rm4l]

Originally Posted by Quoth

I mean why not just run it in a regular terminal?

I have had a server since maybe 1994 at home.

I've used VMs on laptop/workstation for the occasional non-native must-have program.

I've never seen the need to run non-native, sandboxes, docker or any such thing on a server, ever anywhere.

Certainly a Manga2Kindle utility doesn't need anything except a compatible OS and writing crossplatform (Mac/Linux/Windows) isn't hard using Java, QT, Lazarus or many other tools, for console or GUI.

That’s a fair point, but for me, it's more about learning. I'm still a beginner, and my go-to language has been JavaScript. I wanted to learn more about Docker, so I used it as a learning tool.

It also makes it easier to manage—Docker Desktop or Portainer gives me a clear view of running containers, and other users can just pull and run them without much setup. Running everything directly in the terminal feels a bit intimidating and not very beginner-friendly.

Right now, I’m practicing with a small N100 server. My first project was a Manga2Kindle Docker container to get familiar with containers. Now, I’m refactoring my AWS Lambda function (which fetches TV show data from TheMovieDB) into a PHP application with a local MySQL database inside a Docker container—mainly to learn PHP and MySQL.

[axu2]

Originally Posted by unn0rm4l

Time is also the reason why i haven't found a good setting for ImageMagic yet and some mangas look way to light. I hope someone can play around a bit and share the right settings that there would be more shadow and more black in the lines and text. Re:Monster is a good example to find a good image setting.

You can check how kcc handles image darkening and black levels. By default it does a lot of image optimizations for eink.

[unn0rm4l]

i have a look at it, thank you

[unn0rm4l]

Originally Posted by axu2

You can check how kcc handles image darkening and black levels. By default it does a lot of image optimizations for eink.

i just downloaded it, i have a look or maybe fork it and put them both app's together
But what do you guys use for the actual transformation, i use ImageMagic but at first i tried with Sharp, that made my container to heavy tho.

[axu2]

Originally Posted by unn0rm4l

i just downloaded it, i have a look or maybe fork it and put them both app's together
But what do you guys use for the actual transformation, i use ImageMagic but at first i tried with Sharp, that made my container to heavy tho.

Everything is in image.py