Grab an early donut to celebrate National Donut Day tomorrow 🍩
And get ready for Monday, when Apple should finally be announcing their long-awaited VR/AR headset. Maybe they’ll also add a headset emoji in the next iOS.
btw – we’ve opened up our free community to new members. If you want to level up your video skills and connect with other creators, come join us!
🤯 A new app just dropped yesterday and it could make recording videos 10x easier and higher quality. It’s called Roll – it’s an iPhone app with a web interface. It uses your back-facing, higher-quality cameras to record video.
Then it uses AI to enhance the video quality, give you two angles (a wide shot and close-up), and enhance the audio. It’ll also automatically reframe and give you different format options.
But then the magic happens – it uses AI to recreate your background, so you have the ability to add realistic looking dolly slides or push-ins, all from your static iPhone camera. And this isn’t just for self-recording – it has video rooms and call features so you can have other people join and get the same quality experience.
Will definitely be digging more into this.
This is not a real dolly move
🏄♂️ Interesting essay from The Verge on how AI could be integrated into our web browsers and change how we surf the web, going as far as possibly changing a website or article you’re looking at based on your preferences. This is another reason I’m bullish on video in the generative search future – information delivered in another format that AI can’t duplicate or alter (yet).
🦑 This funky, squid-looking camera was making the rounds on social as the ‘future of photography – RIP cameras,’ where it used lensless sensors to recreate the scene in front of it. Shockingly, the internet was wrong and that’s not actually the case.
The ‘camera’ is grabbing all sorts of metadata – time, location, weather, points of interest. It then turns that into a text prompt and uses text-to-image AI to generate the ‘photo.’ The creator calls it Paragraphica.
Like Roll above, I think this is a peak into the future of how AI will be used to enhance the photos and videos we take.
🥽 Not to get buried in Apple’s AR announcements next week, Meta announced the Meta Quest 3, coming this fall. Coming in at $499, it’s way less expensive than the Quest Pro but pulls some features from it, like full-color passthrough which would open up a lot more augmented reality uses.
🎨 More awesome uses of AI being used as tools for a larger creative process. Midjrouney to generate scenes, Photoshop generative fill to expand them, and Blender to turn them into a 2.5D scene for a video game.
These tools are from our sponsors, which help make this newsletter and our videos possible. Even though it’s paid placement, they’re really great products – many of which we used before they became sponsors.
📮 Metricool ► Easy yet powerful social media scheduling and metrics dashboards.
📺 OpenReel ► No fuss video creation, great for marketing videos
🔒 Vestigit ► Secure and protect your video content
🪧 Adspective ► Use AI to dynamically place ads and products in existing videos and photos
🗄️ MASV ► Transfer large files and folders fast
Some more awesome gear that we saw at NAB.
Using AI as a Storyboard & Mood Board Generator
One of my favorite uses of AI, specifically Midjourney, is using it to convey ideas and concepts.
An obvious use is pre-production for planning a video shoot.
Now you could just type away what kind of scene you’re looking for, but there are also ways to narrow down the technical aspects of the image that’s generated, like previewing how a shot looks with a specific lens.
Midjourney – Camera lens experiments🧵
I experimented with 30 camera lenses so you don’t have to
I put the lenses up against the “photo” prompt in MJ
Here are the results 🤗Part 1/2
If you want the full PDF comment🔥📷🤖
&
I’ll send you a DM
(Must be following so I can DM) https://t.co/ismV6VXifn
— 4rtofficially1ntelligent (@4rtofficial)
May 29, 2023
Generating scenes
ok, confession, I used ChatGPT to generate all of the prompts in this thread.
I’ve tested hundreds of GPT prompts you can use to generate really good MJ prompts. I’ll be sharing my favs in my next YouTube vid (link in bio)
🧵 Some example results from one. Full prompts in ALT https://t.co/Yq6mHjzpXR
— Nick St. Pierre (@nickfloats)
Mar 22, 2023
And generating backgrounds
Creating factory lofts in the #Pixar animation style has become my ultimate obsession. love them so much!
Prompt in ALT.
#Midjourney #AI #AIart #AIartcommunity https://t.co/E3UWFVuMjP
— prof. Crispi (@profcrispi)
May 27, 2023
My next mission: use Midjourney to generate realistic backgrounds for YouTube videos shot on a green screen. It’ll be a combination of getting the lens perspective and field of view correct. Stay tuned.
How else I can help you:
Check out our YouTube channel with lots of free tutorials (and let me know what else you’d like covered)
Join our free community
Get our YouTube & Video course, which has had over 4000 students
If you have a specific project in mind and would like to potentially work together, just hit reply
