PixelCribs is a NFT project that consists of 420, randomly generated, pixel art and isometric cribs with unique characteristics. The essence of the project is that you can buy a piece of virtual land with a house in it, and of course this ownership can be verifiable on the blockchain.


I first came up with the idea of this project back in 2021 when the crypto markets, and specially NFTs were booming. The concept of NFTs immediately caught my attention, they are the combination between art and blockchain technology so I immediately started brainstorming how I could be part of this. I teamed up with a friend that would understand the concept and was interested in NFTs, I developed a strategy as to how we would bring this project to life, and together we manually created ALL assets.


The PixelCribs came to life for a simple reason, three forces were combined: Myself as the graphic designer and project manager, my friend as an asset creator who funnily enough is an architect and actually designed the house base/shape, and finally my brother who helped with the dev and coding aspect. This resulted in a well balanced mix of talents that created an incredible result.

There were many challenges and I was in charge of making sure that EVERYTHING, absolutely everything fitted together into place perfectly. Any small mistake in the production phase would delay us for days if not weeks. So my challenge was to improve the creation process so that we could only focus on creating great assets.

Main learnings:

Anatomy of a crib:

Below you can see the 15 attributes that constitute a crib. Each attribute category contains 10 assets on average. You can see that even if an asset is only a few pixels wide, the 192x192px canvas in preserved, and is exported in .PNG format in order to be able to merge them in the blending phase.


By using specialized software (Hashlips NFT engine), you can automatically mix all 15 random attributes into one crib. By stacking the 15 .PNG images from above, in a specific order, you get this beautiful crib as a result.

In this step I was in charge of setting the software up, establish different rarities for assets (e.g. sports car is rarer than a regular car) and finally in making a run of 10,000 randomly generated cribs, which we chose the best 420 of.

This was a very tedious process but resulted in less filler content.


(in logical order)

Example cribs: