⚗️ Experimental R&D
IRIS is a generative audiovisual program that translates any song or audio clip into mesmerizing visual mandalas, in real-time.
[ The Great Preacher (13:10) – Arnaud Rebotini ]
[ TouchDesigner configuration ]
HOW IT WORKS
The program visualizes an audio clip in a method similar to a common audio equalizer display. The audio file is sampled at 250 different frequencies ranging from the low to high frequencies. The value of the sample is displayed as a pixel on a scale of black to white. This process is repeated 60 times per second, or 60hz to create a “live” visualization of the audio clip as it plays.
The rate of 60hz is chosen because it matches the maximum frame rate used in this project. If the clip was sampled at a higher frequency some of the sampled values would not be represented in a rendered frame of the output video, and that calculation would only add unnecessary computation and slow things down.
[ Example of a common Equalizer Display ]
The values of the EQ must be recorded for the entire song in order to accomplish a final image. IRIS accomplishes this by essentially capturing the running EQ to a rotating image layer. As the image layer rotates the values (brightness) of each of the samples in the EQ is added to the value of the image (which starts as black). As time progresses the IRIS fills up. The program actually takes into account the total length of the song, and rotates the underlying image according to the progression through the entire file. The result looks very similar to a typical audio spectrogram, but recorded in polar space.
The program has a ton of potential for visual entertainment; live or otherwise.
While the process of creating each iris is interesting, visually it can be unclear what exactly is happening. Emphasizing the current EQ state as it is being etched into the mandala helps to both clarify what is happening, and make the whole process more visually enticing. This emphasis is achieved with a few different effects.
Being able to see each song as a single visual object opens up some very interesting opportunities for accessible and objective critique. The structure of songs translates to different visual aesthetics. Short simple songs achieve a kind of geometric or even abstract simplicity, while longer longer more complex songs become incredibly detailed tapestries full of texture and nuance. Bold choices in intros, outros, and breaks become defining marks across the face of the underlying compositions. Even the word ‘composition’ comes to bear both its visual and musical meanings.
[ Savage – Megan Thee Stallion ]
[ Rockstar – DaBaby ]
[ I love You Always Forever – Donna Lewis ]
[ Fall Back 2U – Chromeo ]
[ Say So – Doja Cat ]
[ Thirst – SebastiAn ]
[ Drops of Jupiter – Train (Colorized) ]
[ Animation: Say So – Doja Cat ]