Transitioning to Filmmaking Day 10: AI Creative Summit Day 1
Today was day 1 of the 2 day AI Creative Summit conference. I attended many sessions presented by some really talented creative professionals. My head almost exploded with so much information given during this great conference, but I feel so much more creative today then yesterday and realized that our creative possibilities are limitless with all these new tools and workflows.
Also, as an FYI, tomorrow, 9/15/23 at 11am PDT, I will also be presenting a breakdown of my short film, The Adversarial, showing you all the tools and workflows I used to get it made.
Keep reading for today's session details.
Presented by Gary Adcock and Jeff Greenberg
Real World VFX
Presented by Jeff Foster
Jeff covered how to retouch photos using Remini Web. He went into Photoshop's generative fill and scene expansion. For the expansion, he showed some great movie shots where he expanded the background in the shot using Generative Fill in Photoshop. He was able to accomplish this since it was a standing shot (the camera didn't move). It was very impressive to watch. He then moved on to Runway, which is an app I used in my short film, The Adversarial. He showed some Midjourney renders that were animated in Runway ML and then styled using Gen-1. He also spoke about Eleven Labs for creating synthesized voices and HeyGen for creating digital humans for presentations and for cloning yourself.
Demystifying Generative AI
Presented by Ashlee Martino-Tar from NVIDIA
Ashlee is a 3D artist experienced in 3D modeling. Her presentation by far was THE best presentation I've seen about generative AI so far (in any conference or meetup). She knows her stuff. She explained that Generative AI is not only about image creation but also about guided creation, refinement, finetuning and training. Her emphasis was on the artist and how important it is to know the language of art to get what you want out of generative AI. She did this by showing us an example of what her boss created with a simple prompt in Midjourney versus what she was capable of creating, since she has "the language of art" to pull from when creating text prompts. She gave plenty of examples of good text prompts, which include a subject, a token weight, combiners and negative prompts. She covered various tools including Controlnet, how to do character development, how to create materials for 3D models, how the same prompt will give you different results in Stable Diffusion, Midjourney, Adobe Firefly and Runway ML, using online services for generative AI versus local apps, and what local solutions are available today including Easy Diffusion, Invoke AI, Automatic 111 and Deforum Stable Diffusion.
She did provide this handy image (see below) showing all the generative AI apps in a graph comparing functionality versus ease of use.
She ended the presentation recommending that we follow:
Shaping Your Creativity: AI Tools for the Modern Photographer
Presented by Frederick Van Johnson
AI for Audio: From Smart EQing to Music Stems
Presented by Jeff Greenberg
Jeff covered a bunch of AI tools for music and audio creation. His presentation was super meaty and full of tasty treats! The tools he covered included:
Retiming of Music
Adobe Remix Tool (using the audio workspace you can customize variations and segments)
Like I mentioned at the top of this article, it was a lot of information to take in, but thoroughly enjoyed day 1 of this conference. Can't wait for tomorrow, as I'm presenting and also will be attending many other sessions.
Other things I worked on
Worked on the final touches for my presentation about my short film, The Aversarial, for the AI Creative Summit day 2.
Until the next entry!