Crafting NFTs: An Artist’s Perspective
Insights gained from the launch of Orcanauts, an impact-driven NFT collection on Solana.
This past season, I had the opportunity to create the artwork for the Orcanauts, which we released as 10,000 adorable NFTs into the wilds of Solana. As part of the Orca community, I’m incredibly proud that we collectively raised over $900,000 USD in mint revenue to fund Aflatoun International’s development of a DeFi education curriculum for youth.
Although I have ample experience in digital design, this was my first splash into the NFT world and my first time creating generative art. As such, some parts were smooth sailing… and plenty others took a stormy turn! Hopefully sharing my process and what I learned will help out any other fellow beginners.
In this post, I’ll dive into the process of creating the generative art behind the Orcanauts. To see what it was like to develop this NFT project from an engineering perspective, check out @jnwng’s 3-part series!
The Template
Generative art can be created using a template that combines variable elements. Having designed other templates, I knew it was important to figure out early on what constraints would affect the overall composition of the images we would mint as NFTs.
Our vision for an Orcanaut was a friendly whale that would make for an adorable profile picture (PFP), and I used this as the starting point for sketching ideas. This influenced the Orcanaut’s wide, open face (for cute expressions) and bold outlines (for visibility).
In hindsight, I realized many other PFPs feature a centered character, which reduces the chance of cropping issues and helps keep the composition balanced. Something to consider next time!
Creating and Handing-off the Artwork
As far as actually creating each image, I worked across a few apps. Sketches started in Procreate on my iPad, which I brought into Illustrator to draw the finished vector art. For a while I stayed in Illustrator, but for better hand-off and visibility to the team, I ended up moving everything to Figma.
It wasn’t a clean switch to Figma because Illustrator is a more robust vector artwork tool. This was one of the annoying parts of the workflow because I had to manually transfer artwork, and not all features translated well. Some things to consider if you’re making this jump too:
- Clipping masks in Illustrator become regular paths in Figma.
- Strokes in Illustrator with variable widths are expanded into shapes in Figma
- And, this one might be obvious, but any advanced features like envelope distort or 3D effects will not translate directly to Figma. Some of these became raster images, like the SOL beach ball.
On the positive side, I was then able to use Figma for what it does best: collaboration and components.
For reviews, @oritheorca and I were mostly async, using Discord and Figma. Ori reviewed each trait for a “first pass” to check that the illustration visualized the right object and was stylistically in the right direction. Then, I revised them for a “second pass” to correct any mistakes, make the traits cuter, or address any visual weirdness. I tried to batch the reviews to be at least 15–20 traits at a time so that it would be easier for us to resolve comments and track progress for each one. There were some small tweaks later on, but we found this review cycle worked really smoothly during the production phase.
For file set up, I made each trait into a component, positioned in their frames where they would be in the final image. At first I thought making each trait a component might not be necessary, but this came in super handy later on for late game changes and creating promotional materials. Using the consistent frame meant there was no manual positioning needed on the engineering side.
Finding your Frame
I set the artwork size to 900px square to begin with. Since I was working with vectors, I knew we could resize it later if needed, but some quick research showed that image sizes for generative NFTs spanned from 300px to over 4000px square. We ended up sticking close to the original, with a final image size of 888x888, which works for most PFPs.
While most of the transition from Figma to the final image went as expected, we did have a hiccup with layer blending modes. The grin trait was originally set up as an overlay layer so the mouth color would match the body color. However, we realized late into the review that it wasn’t showing up correctly. Although I’m sure there’s an engineering solution to creating the effect, it would have required earlier art and engineering collaboration. In this case, it was simpler to manage the blend mode on the art side, so I created separate grin images.
Late Stage Learning and Changes
In the prior steps, I was typically working with one Orcanaut at a time, turning layers on and off to see various combinations. This is a fine approach to checking the positioning of elements and to see if any layer overlaps look bad (like the astronaut helmet + snorkel mask), but it’s a poor tool for understanding how the collection would look as a whole.
It’s important to look at the NFTs altogether because they would later be displayed as a collection on secondary market sites, and if someone owned multiple Orcanauts, they would likely see a grid of them in their wallet too. Plus, part of the fun of being in an NFT community is sharing them in chat channels and seeing everyone’s different NFTs!
@jnwng patiently generated and regenerated batches of the Orcanauts so that I could review them as a collection. After seeing early batches like this one, a few things became obvious:
- The basic background was repetitive. The original yellow ripple background (Basic, Honey) was designed to be a simple but pleasing pairing with the whale, but it literally showed up thousands of times! In a grid, it was overpowering and made all the ‘nauts feel similar even if they had totally different faces and accessories. This inspired the color variations for the basic background.
- Similarly, the whale’s body color was repetitive. We started with one body color to reserve body color as a trait for future features, but again it didn’t look good viewed in a collection. We decided to have three body colors, which I like to think of as being similar to the starter Pokemon options.
- Certain combinations didn’t make sense and some layers needed special treatment. For example, the snorkel needed to show up above the hats, unlike other eye traits.
- Lastly, reviewing the collections with @jnwng gave me a better understanding of trait rarity and how much artwork actually needed to be created for a project like this. In the beginning, I had calculated the total possible Orcanauts that could be created from the traits we planned, which was tenfold the number we needed. However, accounting for rarity meant that we really did need all the traits we planned.
From the art perspective, this collection view also allowed me to make general compositional changes–the balancing act I mentioned at the top of this post. Despite having the general template in mind and manually testing combinations, seeing them all together allowed me to:
- Realize that the whale could be more centered in the frame overall. Initially I left generous spacing above the whale in case of large hats, but it turned out I didn’t make the hats that large 👒. This meant everything looked too close to the bottom edge in the collection, so I shifted everything up.
- Spot any style anomalies. For example, the coral accessory started with no stroke, but all the other foreground traits (everything except the backgrounds) had strokes.
- Dial in other system details. I went back to make stroke widths consistent; double check color consistency; and adjust any strange overlaps or tangents.
After taking these insights to the drawing board, the improvement was as obvious as it was immediate:
This project was a joy to work on, from collaborating with a stellar team to taking on the art and design challenge. The most meaningful aspect was envisioning how the project could do good — and then following through to fund Aflatoun International’s DeFi curriculum. It’s been a personal goal of mine to use art as a means to support a cause I believe in, so I’m heartened by the many crypto folks willing to support an artist and a charitable cause with their funds.
Although the focus here was the art and design side of creating the Orcanauts, it doesn’t escape me that @jnwng’s technical expertise made the workflow easier all around. Even if the engineering side isn’t your focus, his deep dive on Medium will help round out your understanding of how to produce the final NFTs!
To hear more from Cori, be sure to follow her on Twitter: @corcorarium
And for all things Orca and Orcanautical, head over to @orca_so
Disclaimer: The content of this communication is not financial advice and should not be relied on by any persons as financial advice. This communication has not been provided in consideration of any recipient’s financial needs. We have not conducted any financial assessment based on the personal circumstances of any recipients. All persons are encouraged to seek their own independent financial advice prior to taking any further action.
