It would be fascinating if people suddenly started doing interesting things with NFTs from like 2019 presumed abandoned to junk status. A kind of unrugging.

“Hey remember that thing you minted, decided was a scam and forgot about? It can now be used to play this weird game.”

Onchain NFT metadata makes this a lot more feasible, i.e. Audioglyphs uses data from another NFT project to seed the generative music.

10,000 randomly generated, infinite audio NFTs derived from on-chain [Pixelglyph](https://opensea.io/collection/pixelglyphs) data.

Each Audioglyph synthesizes a unique, infinite stream of music directly in your browser at lossless quality.

## Future of Music
Audioglyphs are a glimpse into a future of music where instead of publishing statically recorded songs and limiting access, artists create infinite variations that are available to all to explore freely.

Instead of buying the rights to listen to music, listeners can truly own a unique version.

## Open Technology

We plan to open source the core technology Audioglyphs uses for managing and embedding audio processing on the web.

With the success of Audioglyphs, we plan to continue to develop tools for artists to create and publish generative music.

## Carbon neutral

Carbon emissions from minting will be offset with a donation supporting the Devarahipparigi Wind Power Project in Karnataka, India facilitated by Offsetra

Audioglyphs

Delver at contraptions.venkateshrao.com, running summerofprotocols.com, host of /gloom ☀️

Onchain NFT metadata makes this a lot more feasible, i.e. Audioglyphs uses data from another NFT project to seed the generative music. 

https://opensea.io/collection/audioglyphs

This was one of the first projects I bought into because of this very idea.

Now that I’m wiser, I see in the contract that tokenURI points to a heroku server for the metadata, and the Pixelglyph images are stored on ipfs.

USD Coin

Can’t say I’m surprised. L2’s in theory should make onchain metadata more accessible cost-wise.

Metadata is relatively inexpensive to store for generative collections since they’re arrays of text selectively served by the smart contract. Loot is a great visual example.

L2s are more helpful with writing data (lots of it and/or lots of times) and storing more complex media (in kilobytes, not yet megabytes)