Nanophotonic media for artificial neural inference:We show optical waves passing through a nanophotonic medium can perform artificial neural computing. Complex information is encoded in the wavefront of an input light. The medium transforms the wavefront to realize sophisticated computing tasks such as image recognition. At the output, the optical energy is concentrated in well-defined locations, which, for example, can be interpreted as the identity of the object in the image. These computing media can be as small as tens of wavelengths and offer ultra-high computing density. [...]
Fig. 2. (a) NNM trained to recognize handwritten digits. The input wave encodes the image as the intensity distribution. On the right side of the NNM, the optical energy concentrates to different locations depending on the image's classification labels.
For a simpler and more accessible tool that turns light into numbers, see https://www.thingiverse.com/thing:1068443 which is a simple 3D-printed digital sundial (converts sunlight angle to a digital clock readout) complete with number transitions that are not unlike daliclock. As a bonus, the code to generate the part is available and directly encodes pixel masks, so you could make something that produced changing messages over time from sunlight.
That said, I should also mention that light in glass has amazing computational properties; people used lenses to perform approximate fourier transforms for 150 years now. This paper is effectively only theoretical (they simulated but did not construct the actual feed-forward neural network glass).
nevermind, you already posted the digital sundial.
This looks like something that can solve your song-ID problem!