I’m not an active person on social media but I intend to do things on them, wonderful. Link to heading

Gif where my cousin and I are dancing for a sample video for a Spark AR filter.

Papi Gavi’s filter for Instagram.

Putting us in the picture Link to heading

I have recently delved back into the world of augmented reality with Spark AR, but this time I have gone deeper than my previous adventure with this programme. After this second round, I must emphasise one thing: augmented reality is a much more extensive and complicated discipline than I first thought.

To put us in perspective, it took me about 8 years to open an account on Instagram since its debut in 2010 and I ended up leaving it in oblivion after a year because I considered it as a social network focused exclusively on lifestyle. Apparently, between the fact that I had little to show the world and the little interest I had in this type of platform, I didn’t find out about the new features that were implemented on the social network over time, and I think I missed out on one of the most interesting functionalities that have appeared on the platform. In August 2019, Spark AR, the framework for creating filters for Instagram and Facebook, was openly launched.

Augmented reality is something that caught my attention years ago, but which I forgot about relatively quickly as it was not accessible to me. My first memory of it goes back to 2003 playing with the PlayStation 2 EyeToy, which I could play with exclusively during my brief visits to the supermarket on the PlayStation 2 retail display.

Some 20 years have passed since that EyeToy, and these kinds of technologies are more present and accessible nowadays, so much so that we can even carry them in our pockets. Google Maps gives us the option to use this technology to guide us to our destination on a walking route, or Pokémon Go gives us the opportunity to catch a Pikachu when we go outside.

The Spark AR and I, a story of a crazy interest Link to heading

Although it’s a topic I’m curious about, it wasn’t until February 2021 that I started creating filters for Instagram and Facebook with Spark AR. According to the projects I maintain in a test account, I worked with version 106 of Spark AR Studio, which meant I was 105 versions late 🥲. At the time of writing this post they are already on version 141.

In the screenshot below, we can see that first contact with Spark AR, which consisted of five filters:

  • Two were from the streamer Ibai, which I honestly made to get visibility. On the one hand a face mask of facial tattoos (Post Malone style) and on the other hand a 3D model of a wrestler to which I added his face and gave him movement in Mixamo.
  • A Hulk made by the user Lil_CJ which I made twerk dance.
  • A King from the game Tekken of dubious provenance to which I also implemented a dance that I don’t remember.
  • And finally, a Patamon made by user rafael.r92f, which is placed on top of the user’s head, like in the Digimon series.

A screenshot of the panel of the different Spark AR filters posted on my test account

Capture of the five filters I made in my first contact with Spark AR.

As I am a nobody, any number higher than 10 seems to me a savagery, and what to say about the Hulk filter that received more than 1 million views until the end of its days, a tremendous madness. I must confess that I didn’t really care about the numbers from the beginning as my main objective was to learn, but I won’t deny that I really enjoyed watching how the numbers moved and seeing the reach they were having:

A screenshot of the countries that used the Hulk twerking filter the most - India, Indonesia and Brazil.

Countries that used the Hulk Twerking filter the most.

The funniest thing of all is that 60% of the people using the Hulk twerking were from India, and it was quite amusing to see the public reels that people shared interacting with the filter. What you usually saw was people dancing with the character to Bollywood-style Hindi music in the background, and it’s a shame I didn’t save a single one. 😥

All very nice, but what would you have liked to know beforehand? Link to heading

In my case I am a user who still has a lot to learn in Spark AR, but I must confess that I have a slight advantage over someone who starts from 0 in the digital field. However, there are certain things that I would have liked to know before and that I think may be of value to people who are starting out. What would be ideal for making good filters?

1. Knowledge of image editing, video editing and 3D modelling is essential. Having knowledge in Photoshop, Illustrator, After Effects, Blender or Cinema4D, from my point of view, is 50% of the construction of a filter. This does not mean that one must be an expert in all these tools, but we must be clear about what we need and how we will do it. Knowing this, the Internet is the best university to research and start doing things. This is the life of the autodidact.

Spark AR is a tool that will allow us to create a virtual experience, but it is nothing more than a place where we are going to give interactivity to certain elements, and these elements must come from somewhere. In a way, Spark AR is a game board, and we can’t play a game without pieces. We have to know what tokens we need and we have to know where to get them from to put them on the game board.

2. Spark AR can be like web development, there can be different paths to the same result.
There is not just one way of doing things, and some will be more optimal than others. We will have to choose the most efficient or the one that best suits our needs.

We learn by making mistakes, it is in our nature. We stumble over the same stone until we know how to escape from it.

3. Having a webcam is an extra that can be a great help. To make a filter you need a computer and a phone, but during the development process, a webcam can be a great ally because we can visualise our work in real time. This makes the creation process much easier.

4. This does not generate money..
Creating filters DOESN’T MAKE MONEY like a mobile app can. In an app, you can have different ways to generate revenue, such as setting a selling price per app, adding micropayments or even implementing ads. Not here.

This doesn’t mean that a person can’t make money with filters, as there are companies or brands looking to have their own filter and there is a big market out there.

5. Templates are great, but it is better to avoid them at the beginning. In Spark AR we have templates that allow us to save time in the process of creating filters, but at the beginning it is better to avoid them as they make things much easier and it is better to learn progressively about each element.

After having a run-in using “Blank Project”, from my point of view, we can make the leap to them to save time.

And finally…
EXTRA: Each platform has its own software.
Spark AR for Instagram and Facebook are not the only ones in this market, we also have Lens Studio for Snapchat and Effect House for TikTok.

In my case, I haven’t explored the rest for long enough, but I have to confess that I would have liked to try Effect House due to the boom that TikTok’s platform is having, but they still only have support for macOS (it’s July 2022). A pity because not all of us have a Macbook. 😢

Below are links to these resources and I sincerely encourage you to take a minute to research this world which still has a long way to go. 😁