Pinned Tweet
Been working on optimising the renderer behind my latest fully on-chain works. The result is What Matters Refined.
The updated renderer uses an adapted DynamicBuffer by David Huber (@cxkoda ) and Simon Fremaux (@dievardump ), alongside streaming Base64 encoding and more efficient memory reuse.
For the heaviest 67-layer work, the new renderer benchmarks at around 381 million gas for tokenURI, compared with the previous reported live reading of 3.38 billion. That's almost 89% less read gas.
The individual PNG layers and traits are reused and remain fully on-chain.
The original What Matters collection stays on-chain as a record of that exploration. It helped reveal how much these works could contain, and where existing retrieval tools started to struggle.
For Refined, I'm keeping the final tokenURI below 2 MB as a practical compatibility budget, with the aim of making the traits, thumbnail and animation easier for marketplaces and other tools to retrieve.
Big thanks to @EtoVass and @tokenfox1 for pointing me toward dynamic buffers and helping me understand where the memory was going.
Explore the collection and its smart contract using the links below ⤵️
What Matters #3 — Someone to Return To
An exploration of human connection, belonging, and the quiet comfort of knowing there's someone waiting for you.
Phase 3: Abstraction
800 × 800 px
54 fully on-chain layers
artonchain.net/what-matters#…
MasterOnChain retweeted
#2 — Room to Be
48 on-chain layers
35 visible / 13 hidden
Its complete tokenURI requires ~1.457B gas and produces a 1.83 MB Data URI.
That is already far beyond common public RPC limits — yet this one is still successfully indexed and displayed by marketplaces.
MasterOnChain retweeted
#1 — Where I Begin
67 on-chain layers
59 visible / 8 hidden
Rendering the complete tokenURI requires ~3.379B gas and produces a 2.32 MB Data URI — beyond what many public RPCs can currently execute, causing some marketplaces and wallets to fail to retrieve it.
GM. Today I'm revealing What Matters — my latest fully on-chain layered collection on Ethereum.
The works test how far a fully on-chain layered NFT can go.
Each image is assembled from dozens of individual on-chain layers, revealing the practical limits of how today's infrastructure retrieves and presents them.
🧵↓
The collection extends the four-state layered system I developed for Rooted In Code and Inflection Points.
In the final abstract state, each holder's address deterministically recomposes the work, adding to its visual provenance.
Here is how the same work changes across different wallet addresses:
xitter.cf/masteronchain/status/2…
Replying to @masteronchain
Each address-derived state becomes another chapter in the work's on-chain visual history.
Artist → Auction Contract → Collector → Vault
Explore the evolving Visual Provenance ↓
ric.artonchain.net/inflectio…
What Matters
Ethereum L1
Maximum supply: 42
Explore every layer in the Layer Explorer and bid on the available works ↓
artonchain.net/what-matters
This space has some genuinely good people in it.
Received these three works by @COMPUTERBLOOD and his daughter @GL1TT3RGH0ST just for placing a bid on one of their auctions.
How cute is that?
Thank you fam! 🩶🩸
🌬🍂🐦⬛🌘☁️
Fading Forest now has its own page
Seven layered, animated artworks, fully on-chain on Ethereum.
Auctions have moved from Manifold to PND, with #2 now available. Bids start at 0.02 ETH.
Explore the series ↓
Inflection Point #4 is now live on PND ↓
Fully on-chain layered art on Ethereum.
48-hour auction, starting with the first bid:
pnd.fyi/0x0fF1e77BAAD8c5923c…
🟠 Inflection Points auction listed
Structural Breakdown
Token #4
Reserve: 0.01 ETH
Duration: 2 days
artonchain.net/inflection-po…











