Finally, the moment has come! What initially started as a simple project aimed to replicate a popular drawing app (being a part of some workspace research), ended up being one of the most challenging projects we've ever had to deal with.
We decided to do a short write-up on the things that went wrong/right, as it might be useful to someone.
Why:
As mentioned earlier, this project was never meant to be more than a research project. It was originally inspired by another game we worked on, called Nonoi The Kākāpō. There was a small drawing game within it, which we always found the most fun to play and work on. The painting algorithm in it is done via a simple sprite colour change, which means one can't do any finger painting, so for our own drawing app we wanted to have functionality, meaning that we could not use sprites any longer.
Since Lighstromo is located in New Zealand, where the fauna is all avian (all birds!), it made sense to us to focus on birds. There is such a good variety of them, of all shapes and colours and sizes, with some of the most uniquely bizarre characteristics (like the flightless kiwi and kākāpō, or spur-winged plover who has spikes growing out of their wings?!). We also learned through our previous work that there are a lot of dangers, like pests, pets, people, of course, which can lead to some species of birds complete disappearance. That was what happened to moa and Haast's eagle. So, we additionally thought that promoting people's awareness of the bird life in New Zealand can help them appreciate these creatures more, and try to protect them from going extinct.
That was why we thought we would need to include some voice acting work, which will introduce each bird and their way of life. This, of course, required a lot of research on our end.
Those are the 2 websites we used the most when doing our research:
Timeline:
Roughly 2 years of part-time development with big breaks of 2-3 months. How quickly can it be done fulltime? 2-3 months total if started from scratch with a very small team.
Software:
The project was developed in Unity 2019.2.8f1.
We used Photoshop and Illustrator for creating the drawings, as well as UI items.
WavePad and Audacity were used for sound design.
Progress tracking:
For all our other projects, we normally use Trello. Trello is a great tool for smaller/ish teams, it has great visuals, very intuitive design. Interestingly enough, after a full year of development, we made a big switch to...a white wall with sticky notes! The wall was divided into two sections: TO DO and DONE. Each task has a note associated with it, and we were positioning the more important tasks at the top, while keeping the low importance ones at the bottom of the wall.
Now, this might seem like an incredibly simple system, but it improved the work morale greatly. It was a satisfying experience moving a note to the DONE side of the wall, and we really think 70% of the motivation to finish the project was generated by that simple technique. 10/10 recommend and will use again.
Design:
All technical design was done with pen and paper. Most of the coding issues were solved the same way - a good old sit down over a cup of coffee and a bag of crisps.
In terms of the visual design, we drew inspiration for the UI from Duolingo, which you can probably notice in the image below. The best way we can describe this style is 'clean and simple'.
The very first drawing we used in the app testing was inspired by old tattoo designs. We wanted something that was easy to paint, with areas of the drawings which were easy to fill. All birds were designed on paper first, then transferred to Illustrator to create the clean line-art style that you can see in the app, after that all necessary adjustment were done in Photoshop.
Below on the left, you can see the latest version of the first drawing mentioned above (we added Xmas decorations which will be removed some time in February 2021). In the middle, you can see the most prevalent style - somewhere between being too complex or too simple, with lots of medium-sized are to fill, while on the right side there is an example of a simplified drawing, more suitable for younger audiences:
What went wrong:
What didn't go wrong?! At some stage, the project seemed quite hopeless. Even getting the simple Pencil tool functionality out of the way was quite challenging.
Unity is a great engine, but probably was never designed to work as an engine for apps, let alone drawing apps. There are far better options for building design specific apps, but we decided to work to our strengths. We'd been working with Unity for about 3-4 by the time we started this project, so changing to something else was out of questions.
We tried multiple ways of enabling drawing, including using line renderer. Although line rendered is definitely something to explore further, it wouldn't work well with the Fill tool we had, so we opted to go with simple writing pixels into the texture, which works for both Pencil and Fill tools.
We also wanted to implement soft brushes. At some stage, we succeeded only to realize that softer strokes created a different level of complexity for the already overcomplicated Fill tool. It's easy to create a stroke of color A, and then recolor it to color B. Now, if we have a soft brush, and create a stroke of color A, and draw another stroke of color B on top of it, the soft edges will produce a variety of colors in between A and B (which can look like a completely different color: imagine mixing blue and red). This means the Fill tool will need to have some kind of a threshold of filling, so it can pick up those mixed colors, as well as half-transparent colors, yet at the same time the threshold had to be low enough to not fill in both colors A and B together. After testing this for about 3 weeks, we decided to revert the project to the old pencil-like brushes and a simpler Fill tool.
We wanted to have 24 images of birds at the beginning. We ended up with just 12, and even getting those 12 out should be considered a great success. Behind each image, there's a lot of research work that went into it and a lot of in-app testing.
Voice recording was another issue. Originally, we were planning to use a professional recording studio, as we did in the past, but then Covid-19 happened, and we had to buy our own equipment. Finding the right voice actor was a challenge, too. In the end, we opted to have mixed recordings, done by 3 different voice actors.
What went well:
UI sounds were all done by voice with the in-studio voice actor. It turns out that with a little bit of audio engineering, making UI sounds is simple as! They took almost no time to create, and were fun to do.
The main sound theme was originally written up in the studio. We later passed on the unfinished melody to our favorite sound design contractor who turned the simple tune into some light-hearted non-intrusive pure awesomeness! We are very happy with the way the melody turned out.
We switched from having to create two images to one. The way the fill tool used to work was by looking at the black pixels and considering them the stopping point. We didn't know how to work with transparency properly to begin with, so originally we had two images - a simplified pixel drawing at the bottom, which the user painting was written into, and a prettier line-art version at the top, which was not used for anything but hiding the uglier pixelated drawing underneath it. In the end, we learned how to simplify the whole drawing and filling functionality, so now we only use the pretty version of the drawing.
Publishing was, overall, not a bad experience. Although both Google Play Store and Apple store had their quirks, in the end, we managed to get reviewed and published in a relatively short amount of time. Google took about 5 working days to review our submission, while Apple took only one day. We would like to mention that you will need to have a place to store you privacy policies somewhere online, as both stores will ask you to provide those, so be ready for that.
Future of the app:
So, where to from now?
We are going to push some updates to the app in the future. The easiest one is removing the Christmas decorations from the drawings some time in February. That will also be a good time to potentially add more drawings.
We have been talking to some non-profit organizations about this app, which might lead to some interesting collaboration.
We are also thinking about reskinning the app, and publishing under a different name with a more adult theme. This will mean some UI rework, as some sections of the app will not be needed any longer, and of course, new art.
We will try to reimplement the soft brush functionality, although more R&D will have to be done in that area.
Get the app here:
Thank you for reading! Stay tuned for more updates on our work in the future :)
Comments