Impact of AI on the CGI Industry – End of the artist or more intelligent process?

 

 If AI is everywhere, why shouldn’t it also be present in art? Creating art is, indeed, a traditionally manual process. But, I invite you to reflect on the fact that AI in 3D can open a door of possibilities for the artist, as well as bring more agility and efficiency to the process.

 

For the past few months, I have been experimenting with an artificial intelligence tool within Discord (a popular communication channel) called Midjourney.

 

As Artists we are always engaged in process improvement. After all, if something is good, it can always be better! It is our hope that what we share here can inspire you to go further in your creativity as a 3D artist.

 

 

Generally speaking, the period of time it takes to complete a piece of art varies from days to weeks, depending on the complexity of the representation. Now, imagine expediting the process to a few minutes or even seconds, with a highly satisfactory result.

 As an example, processes such as composition, lighting, rendering, and photography can be optimized with Midjourney. This tool works with a database joining photos and other art, merging two or more artistic foundations to deliver scenarios with an impressive and detailed richness.

In the process of experimentation, many artists were asked about their perceptions of AI in 3D. They had the following two queries:

 – How can AI help the artist in his creation process? – Will the tools that currently exist replace the artist’s work?

 Midjourney may perform activities with greater agility than the artist’s hand, but the greater challenge is to replace the human vision. There is still one human being operating the tool: the artist himself. We are going to share this experience to show you what AI can do for artistic design.

 

 

At first, Midjourney may seem confusing. Being a beta version, it has evolved very quickly and its database has grown and improved thanks to the artists themselves. For those of you who want to venture into the platform, just access the ‘newbies’ bar. It is an intuitive tool and makes adapting to the interface very simple.

 On the tool’s website, there are brief explanations of how to interact. (/imagine) You can replace “imagine” with specific keywords to attract the image you want. The possibilities for exploration are endless. What is it that you want? Your technical background can help you be more precise in your search.

 

 

Midjourney resources were used to explore AI in artistic composition. During the experimentation process, I wanted an ultra-realistic render. And so, specific keywords were targeted: Ultra Realistic Rendering, 3ds Max, V-ray, Jungle, Jedi Temple, Cliffs, Waterfalls, Depth of Field, Eye Level Shot, Matte Painting Trees, Aliens, Titas Ženevičius (name of reference artist), Swamp, Daylight, Concept Art, Sunlight, Low Sun.

 The idea was to create a kind of alien-like, futuristic scenario that resembled planet Earth. Note that in the keywords, references such as ‘software’ and ‘renderer’ were used to signal to AI to assimilate the software.

 In general, there are no rules for this research, as this is a personal matter for each artist. The biggest focus can now be technical in the composition of scenes, reference artists, etc. The choices you make are also part of the process of understanding and experimenting with the platform.

 Of course, I wouldn’t be able to do a rendering at the same speed as the AI. This tool provided a background (a composition reference), saving you, the artist, time in creating your own image. Midjourney simply expedites the process.

 The idea is to take the reference image and to improve upon anything that doesn’t make sense. For example, it could have a foreground or more water falling. In short, the image can be improved to meet artistic expectations.

 

 

From that image, artistic creation began. The 3ds Max® render was used to help with the initial shapes for a recreation very similar to the photo produced by Midjourney. This is a technique in which the photo is embedded within the viewport in 3D, allowing free work, based on the artist’s eye.

 

Modeling refinement in Z-brush

 

The idea was to bring improvements to the image, making it as realistic as possible. So for the next step, ZBrush, a digital sculpting software, was used to refine the shapes already created in 3ds Max® to create a closer approximation with the texture of a stone.

 

The composition process

 

After this refinement, the 3D model of the cliff is brought back to 3ds Max®. At this time, the composition, texture, light, camera, waterfall simulations, and rendering work begins.

 

The post-production process

 

Finally, the image can be completed with a very simple artistic touch to bring more realistic details and special effects like smoke and splash.

 

Optimization of time

 

Midjourney optimizes the artist’s research time, allowing him/her to focus on details and fine finishes in the creation process. Here at Elephant Skin, the idea is to test these technologies, discuss the untapped potential, and make AI an ally in processes such as branding, moodboards, etc.

 Experimentation is the best way for these novelties to be incorporated to support the work of artists and bring greater efficiency to the processes. In addition to Midjourney, we have also experimented with Dall.E, mainly in the context of creating characters and figures.

 AI in 3D proves to be a powerful resource. The idea is to continue experimenting with the many potentialities offered by modern technology.

 The artists will not be replaced by this intelligence and will never cease to be important in the creation process, but we cannot ignore the breadth of possibilities that make creations increasingly detailed and realistic.

 

What do you think about using this type of tool in your artistic creation process?

 

 Written by:

 

Gabriel Ceriani 3D Artist
Credits: Victor Luidgi, Elephants Skin’s Team:

Communications Manager – Roberta Lemos | Communication Intern – João Victor Campos

 

Author – Jornalist Daiana Barasa and Juliana Rodrigues | Naiá