AI startup Roll, is an AI-powered video production platform that is claiming that anyone can create professional, studio-quality remote videos in minutes, with just an iPhone using its software.
It was only a matter of time till AI-based software trickled its way down to the consumer and prosumer levels. AI is both intriguing and scary at the same time and it will be interesting to see where this technology takes us moving forward.
Roll is built on proprietary AI and cloud technologies, and it allows users to record remote video calls in broadcast-level quality, add AI-generated special effects, and publish that content 30x faster than traditional video production workflows. The Roll iPhone app is free to use and the overall AI-powered platform is available as a free trial with a paid subscription option.
So how does it work?
Roll pairs iPhone camera sensors with large AI models in the cloud to turn an iPhone into a pseudo-professional video production studio. Roll recreates the room in 3D in the cloud where it can move a virtual camera around. This then enables users to obtain pro image quality with bokeh, multi-cam shots, motion graphics, and AI-simulated sliders, dollies, and jibs, all with just an iPhone.
Roll’s ethos is that they want wants to democratize high-end video production and make it easy for anyone to create high-quality videos with the equipment they already have at home. Typically, high-budget shoots require studios, crews, and multiple cameras and equipment to get the image quality, multi-camera angles, and cinematic camera movements seen in broadcast television shows. In the last few years, companies and some TV shows have opted to use video conferencing tools for their main remote production needs, which has often led to compromising on quality for the benefits of flexibility and ease of use.
Roll is claiming that they can promise the best of both worlds, broadcast-quality production without any need for a studio or equipment, and at a fraction of the cost and time it takes to produce.
Using the latest in cloud computing and AI technologies, Roll automatically produces a final edited video for you, effectively removing the need for editing entirely and giving you a publish-ready video in minutes.
Who is it for?
Roll was designed for creators who care about producing high-quality long-lived videos that represent their brand and brings their story to life. This could be video podcasts, executive interviews, customer testimonials, webinars, fireside chats, virtual conferences, etc. or anywhere where you’d hire a professional video production company or invest in high-end camera equipment to produce professional videos.
Like most creative tools, they expect Roll to be used and purchased by the creative teams creating the content. Roll is focusing on the “prosumer” market like consumers, independent content creators, podcasters, and influencers who are creating their own content. They are also targeting a wide range of businesses, from small teams and SMBs creating their own marketing content, to in-house corporate marketing and video teams, to video production agencies, and eventually, large media companies whose bread and butter is high-quality video content.
Roll has 2 core products that work together seamlessly to create high-quality videos. They have the ability to carry context across software boundaries. By owning the end-to-end experience from capture to editing to publishing, they are able to create an end video result that incorporates all the context of capture into the most optimally produced and edited video.
- Roll iPhone App: captures and records high-quality video calls from your phone, and automatically uploads to the cloud for storage and processing
- Roll WebApp: SaaS platform to preview, access, share, download, edit, and collaborate on your recordings with your whole team
Roll Features
- Broadcast-Level Image Quality: Roll combines the best of portrait and wide scene captures, all in High Dynamic Range (HDR), and uses AI to apply cinematic bokeh (background blur), producing video quality equivalent to professional studios and cameras.
- Multicam Studio Shots: From a single phone, Roll captures and produce two different camera shots – a wide and a closeup – simultaneously. This allows users to create videos with professional multi-camera studio shots with just one phone.
- AI-Generated 3D Studio: Roll uses generative AI to recreate your room in 3D space so you can move a virtual camera around – simulating expensive studio camera movements, like panning from side to side, dolly, crane. And all done without moving your phone.
- AI Auto-Edits: No more time-consuming video editing. Roll creates an automatic, professionally produced and edited multi-camera video, minutes after you finish your shoot. All completely edited by Roll AI.
- Roll Editor: Roll automatically transcribes your video and populate it in our powerful and easy-to-use text-based editor. Cut lines, change camera angles, add camera movements and visual effects with one click with a click of a button. And collaborate with your team in real-time. (coming soon)
- Shoot once, Publish Everywhere: You can publish directly to any social media platform with automatic reformatting (TikTok portrait, YouTube 16×9, Instagram square) (coming soon)
According to Roll, its videos are sharper than anything on the market for iPhone capture. They claim that they are the only remote video calling solution that records in HEVC, getting twice the bitrate and higher image quality for the same file size. They also state that they are the only company using generative AI for simulating high video production effects.
Roll uses our own proprietary machine learning and AI models at every step of our process.
AI-Assisted Capture: Roll uses the iPhone as its main capture device, and their app records local video/audio and gathers metadata for use later in the video production process. In every scene of a recorded call, they record metadata about that video: face and body detection, recording conditions, lighting conditions, distance from camera, etc. They then automatically adjust the cameras and sensors on your phone to optimize for your environment, and they provide real-time feedback and instruction for composition and lighting, to ensure the highest-quality recording possible.
AI-Powered Video Processing: Based on the latest academic research, Roll has developed proprietary large computer vision AI models in the cloud for segmentation, depth analysis, and 3D inpainting. Using their data from capture, they can recreate the scene in 3D space and use generative AI to fill-in new pieces of the scene. This enables them to move the camera around the room in any way they like. According to Roll, they are able to simulate camera movements and effects, previously only available with sophisticated studio equipment or specialized VFX editors in Hollywood.
This technology provides foundational capabilities to simulate any visual effects a user would want. Early use cases they have tried are camera movements, and text/image overlays:
Roll has internally recorded over 22,000 video calls using its Roll Platform, creating its own dataset. They are not dependent on other people’s content or dataset for training their models. Any data that is used to train is fully anonymized and only used to improve spatial and depth measurements in service of the perception of high-camera quality gear.
Philosophy on Generative AI
Today, generative AI is too often associated with creating fake content out of thin air. That is not Roll’s philosophy. They do not generate fake pixels, people, or scenes. They are using generative AI purely as a tool for productivity as they want to democratize access to higher-quality video production. By using AI to generate effects that would otherwise require thousands of dollars of expensive studio equipment to create, Roll are unlocking high-end production capabilities for anyone with an iPhone.
Roll never generates content without the explicit knowledge and consent of the users. Users create their own videos using the Roll platform, and they use AI to enhance the production quality of those videos. According to Roll, fake content is never generated. When you see a person speaking, that is the real person from the real footage that was captured.
Roll’s AI is used to understand the 3D depth in the scene and is agnostic to the person sitting in the room. Its models are sort of an advanced AI version of iPhone LIDAR hardware, where we leverage data to understand the depth and shapes in a scene, but don’t try to generate or recreate those specific people or scenes in any way.
Roll Early Pricing Model
As I mentioned earlier, the Roll iPhone app is free to use and the overall AI-powered video production platform is available as a free trial with a paid subscription option.
Please note that this pricing represents early thoughts and will change as they have more data.
| FREETry before you buy | CREATORFor individuals | PRO For pro creators and small teams | ENTERPRISECustom plans | |
|---|---|---|---|---|
| Free | $49/user/mth | $199/user/mth | Contact Us | |
| Recording Hours | 2h (one-time) | 5 hrs / mth | 15 hrs / mth | Custom |
| Editors | Host-Only | 1 Additional Editors | 3 Additional Editors | Custom |
| Off Screen Producers | ✔️ | ✔️ | ✔️ | ✔️ |
| AI Features | ||||
| – Multicam Capture | ✔️ | ✔️ | ✔️ | ✔️ |
| – Smart Composition & Lighting Detection | ✔️ | ✔️ | ✔️ | ✔️ |
| – Cinematic Bokeh | ✔️ | ✔️ | ✔️ | ✔️ |
| – Auto-edits | ✔️ | ✔️ | ✔️ | ✔️ |
| – VFX Camera Effects (Pan, Dolly, Crane) | ✔️ | ✔️ | ✔️ | ✔️ |
| Remove watermark | ✔️ | ✔️ | ✔️ | |
| Admin Management* | ✔️ | |||
| Centralized Billing* | ✔️ | |||
| Dedicated Support* | ✔️ |
*Coming Soon
Thoughts
Now, while this all sounds great let’s be realistic here. It isn’t going to be able to replace trained professionals and specialist equipment. AI is certainly coming fast and we do need to be prepared for it, however, at least in my opinion, it certainly isn’t at a level yet that is going to take away the jobs of professionals in our industry.
The pricing of Roll also looks rather ambitious considering programs like DaVinci Resolve and major NLEs can be purchased or subscribed to for far less money.
Who is Roll?
Roll Camera Inc is a cloud-computing and AI-based video production platform that is claimed to be able to enable users to record remote video calls in broadcast-level quality using just iPhones and edit and publish that content 30x faster than traditional video production workflows. Roll is headquartered in the San Francisco Bay Area, with distributed teams globally.
Roll’s founding team were previously leaders at Box, GoPro, Technicolor, Epic Games, and Twitter.
