Earlier this year I wrote a post with tips on how to integrate music into Instagram Spark AR effects. Since then, I’ve developed many filters for artists and attempted to stay up to date with any major changes to the platform which may be interesting to the music business. One of these new features I’ve been meaning to talk about is the Audio Analyzer patch. This patch allows you to receive the real-time full-band signal power (or audio levels) of a playing audio track which you can use to dynamically adjust visuals in your filter.
So far, I’ve used this functionality twice: Once in an unpublished AC/DC prototype and again in a just published Foo Fighters filter. Let’s take a look at each and discuss how they came together.
Foo Fighters “Shame Shame” Filter
Foo Fighters are a long time client of mine and they approached me to turn Agustin Esquibel’s excellent motion graphic work for the upcoming record into an Instagram filter. If you look at the video which inspired this effect, you can begin to understand how I broke down this filter’s construction and connected it to the audio analyzer. On Agustin’s video, the band member is in black and white but the contrast of the picture is being adjusted by the audio. We can create a similar effect in Spark AR by patching the camera image into an Adjust Colors Shader, setting the saturation to
-1, and using the audio analyzer value to adjust the contrast. We’re also using the audio level to adjust the scale of the Foo Fighters logo. Separately, the eyeball effect was achieved by simply building off the provided Eye Color project template Spark AR provides and swapping out the included eyeball shader for an eyeball animation sequence Agustin put together for me. 💪🏻
AC/DC “Back In Black” Filter
When the audio analyzer was originally introduced, I put together this quick concept for AC/DC which not only uses analysis but also takes advantage of the Target Tracking functionality of Spark AR. Target Tracking allows you to start an effect as soon as the camera recognizes a particular image. In this case, the stark iconic album cover to Back in Black. To pull this off, I created separate 3D elements for each character in the AC/DC logo on the cover. As soon as the user points their camera at the album cover in their environment, the camera feed goes black, white, and dark. Then you hear those bells. 🔥🔔🔥 Once again, an audio analyzer is used to dynamically adjust the visuals, starting with the overall lighting of the scene. Louder signals make the scene brighter which allows the bells to adjust the mood. In addition, different signals bands from the audio are used to adjust the z index of each logo character, allowing them to jump off the cover.