An homage to John Howe’s Door of Night, which was originally done for Tolkien’s The Book of Lost Tales. The original illustration was one of my earliest inspirations that got me into fantasy and digital art.
Featured on the Tolkien Gateway
Originally started as a project for a class I took in 2017, I had to put this project on the side throughout many years, iterating over it slowly and redesigning it over time.
I decided to change the composition to emphasize the sense of scale and awe (check out the original), focusing on the main fantasy elements while also guiding the eye towards Eärendil’s ship. I diverged from the original description of the doors and took some liberties to give it a more modern look, while keeping its original spirit.
Wings and clouds detail
- Materials are mostly all procedural with Maya shading networks.
- Dragons were modeled in ZBrush, and scales were generated in Houdini with Redshift proxy instancing.
- The door’s lintel was generated through procedural cracks.
- The scene has ~200 lights total.
- The door’s horns and sigil was modeled in ZBrush – the sigil represents Arda Marred, or the broken earth. It was initially modeled as terrain and then sculpted on top.
- Close clouds were first iterated with WDAS’ Moana cloud to define lookdev, and then modeled and generated in Houdini from scratch. Background clouds and sky are a combination of Noah Bradley’s Sky reference picture pack, and the WDAS Moana Cloud that I left as easter egg.
Made with Maya, Redshift, Houdini, ZBrush, Substance Painter, comped with Photoshop.
Various mood tests to see if it could be darker and grittier. Ended up going for the more high fantasy approach
Part of the challenge of this piece was its scale: the original work had too many dragons. I rebuilt them many times without success, and it was becoming hard to see the next steps. Realizing that this was a tribute and not a reimplementation opened my eyes, letting me focus on the gate and the two main dragons. It allowed me to move forward and finish an otherwise gargantuan scene.
In fact, I think this was my limit time-wise. Five years for a personal project like this is too much in terms of opportunity cost – I could have done many other small to mid-size projects in the meantime. My recommendation to any artists looking for their big next project is to reconsider first, and then iterate on the scope constantly. Focusing will not only improve the piece visually, but it will let you iterate and improve your skills faster.
Paintover experiments to figure out door details.
The scales are unique and procedurally generated
For more WIP images, check out my ArtStation post