AI in Graphic Design Process - Design Development - Motion Graphics

AI in its full power

Starting with motion sequence I first wanted to try harnessing AI's power as much as possible. For this reason I have decided to first start trying storyboarding with Krock.

It mimicked traditional techniques of storyboarding, using AI-generated images based on prompts to map out the story on a board next to each other. While this was nice and I didnt have to manually download and put images together, I felt it could benefit even more from intergrating some AI brainstorming, rather than having to lean on external tools. 



 

After establishing storyboard I turned to Midjourney, where I started generating my initial still images. I was evaluating them based on the suitability for the advertisement, and its potential to be animated.






Having those still images, I moved into Pika which allows for Image to Video along with prompt guidance. But despite its potential, transitions often glitched or failed altogether. Even repeating the same commands led to different results. Pika's restrictions, like short duration and low resolution, were somewhat countered by TensorPix, which upgraded the visuals.








Having somewhat usable animations, I moved to after effects where I tried to stitch everything together. But a significant challenge arose: the lack of cohesion and thematic unity between the sequences. The inherent intricacies of video—demanding a cohesive narrative, fluid transitions, and continuity—weren't fully realised in the AI-generated outputs. Adjustments and tweaks could only do so much. So even though I have already incorporated sound effects and created voice over through elevenlabs, I decided to abandon this project and rather focus on other design territories that could benefit from AI. 




Comments

Post a Comment

Popular posts from this blog

AI in Graphic Design Process - Finalisation and Delivery

Image
Preparing for Hand-In When it comes to wrapping up a design project, the final retouches and delivery phases are crucial. For this final project it meant preparing for final submission of all my finalised designs, its mockups, screen cast and all working files.  While AI tools showed promise, they couldn't match the precision required for layout creation and I crafted the brand guidelines manually in a reader-friendly horizontal format. They detail how brand assets, such as colours, fonts, and logos, should be used in various applications. Separately, to offer a tangible representation of how the designs would manifest in real-world settings, I created detailed mockups, which when coupled with a screencast of my website, provided a comprehensive visual narrative of the project's outcome.  A cornerstone of this project was its organization. From the outset, I was resolute about maintaining a structured approach. My final designs are neatly housed within descriptively la...
Read more

AI in Graphic Design Process - Ideation

Image
Brainstorming The initial phase of brainstorming led me to ChatGPT. By feeding the AI the entire organized brief, it managed to produce some intriguing concepts, such as rocket launch and the planet infusion idea. While these served as an excellent foundation, the AI's capacity to switch rapidly between varied concepts - a critical facet of brainstorming - was found wanting. Nevertheless, I found value in these AI-generated ideas, using them as foundational stones for further development.   Mindmapping For the mind mapping phase, I turned to an AI tool named MyMap (MyMap.ai, 2023), which promotes AI-driven conversations in a 2D map format with branching capabilities. But instead of fluidly branching from my inputs, it offered verbose, AI-generated points. This produced a cluttered mental landscape, prompting a return to the time-tested method of manual mind mapping. Additionally they offer complete mindmap generation based on input keyword but these results were more or le...
Read more

AI in Graphic Design Process - Presentation and Revision

Image
 Human feedback When diving deep into the world of AI-driven design, I've been amazed at the capabilities of machines. But as I had dabbled with AI-generated presentations and even simulated client feedback in previous step of this graphic design process, they had their limitations and it became clear that genuine human feedback was irreplaceable, especially when refining design concepts. So, for this critical stage, I decided to go old school and I chose to gather direct feedback from real people that would into the imaginary company's target audience - millennials and Gen Z, aged between 18 and 35, of all genders - adventurous, open-minded, enjoy trying new experiences, and are health-conscious . My first step was to craft a questionnaire. With the help of ChatGPT, I generated few initial drafts. After some tweaks and combination, I had a 10-question survey in hand. It was a mix of open-ended queries, rating scales, and multiple-choice questions, all des...
Read more