I’ve become obsessed with cameras, during our Tenerife trip, where I got to use a borrowed mirrorless camera for the first time. After that, I researched which camera I’d buy for myself for five months. While watching all the reviews, I was interested in the exposure settings for each picture presented. I’ve displayed many photographs in our Tenerife vlog with my wife. I’ve decided that I’d like to show those exposure settings in my other videos as they might inspire other photographers.
I’ve tried many image editors and tools, but not one was able to fulfill my demands on how I’d like to display those settings underneath each picture. Many tools supported some kind of framing, however it was always in a limited way.
Hello metaframer
I knew straight away that this would be a cool and small side-project that I could create. I could use my favorite technologies and have good content for my streams.
At first, I thought, that I would have to render the text right into the image. It will however be in some aspect a deformation to the quality of the original image. There was also a question of how should the text be displayed in a way that for any image size, the text will be of a similar or determinable size.
Instead of rendering pixels into every different format I’ve opted for creating an SVG file that sits alongside the original image. This way the text also scales with the image itself. The complexity was lowered instantly. Rendering SVG is just like rendering HTML. I’ve used handlebars template parser. There are more favorable template parsers in Rust like Askama. But I wanted to allow users to create their own templates so the templates are not part of the executables and they can be stored in the user’s .config folder. I had to test the support of my video editing software (I use Kdenlive BTW) and the decision for SVG was set in stone (a.k.a. README).
There were still complications along the way. SVG is still a much less capable format than HTML and positioning elements in a space-around fashion has to be calculated. There are also many different options and resolutions that I had to consider and I put all of them into the CLI arguments with clap. Clap is a fantastic command line argument parser for Rust. I had to do many calculations on the resolutions and whether the image was a portrait or not. In the end, It wasn’t as small of a side-project as I thought. I’m still very proud of it.
I got to experience the ease of publishing crates on crates.io with cargo.
This week, I’ve automatized the release process and published binaries for multiple platforms.
On to the next one
I can’t wait to use metaframer in my new videos. I’ve already ordered my first mirrorless camera and it is on the way. Be ready for new content. What should be my next project?