Jun 10, 2025 1:59:19 PM

How to Record Spatial Audio for 360 Video in 2025: Key Lessons Learned

Ruben Åeng, Lead Audio Engineer at Nomono

At Nomono, we’re passionate about immersive storytelling. In 2019, we set out to revolutionize how Spatial Audio is recorded – making it easier, faster, and more accessible for creators working with 360 video, or just immersive audio in general. Our Sound Capsule simplifies spatial audio capture down to a single button press, intelligently tracking lav mics and objects in space so that audio aligns naturally with what the viewer sees.

Voices come from the direction where the person is actually speaking. EQ, denoising, leveling, and loudness normalization are handled automatically. And you get an Ambisonic ambience recording to mix immersive background sound to taste - all from a device that fits in your hand and weighs less than most traditional Ambisonic microphones.

We’ve always believed this tool could transform immersive production. And now we’ve put it to the test in our own creative project: a 360° documentary about the legendary Goatboat – an ancient boatbuilding tradition from Nordmøre, Norway, dating back to the Viking era. The full documentary launches at the end of June 2025, but for now, you can check out the teaser and follow our journey into immersive content creation.

Below, we’re sharing some behind-the-scenes insight into what it really takes to create 360 video with Spatial Audio - from recording and editing, to muxing and publishing. Spoiler: it’s still a bumpy ride, but we’re getting there.

Disclaimer - Before you watch the video: A Note on Audio Playback in 360 Video

Due to limitations in YouTube’s 360° audio support, this teaser features First-Order Ambisonics (FOA) only. That means it doesn’t include headlocked stereo audio – commonly used for voiceovers, music, or fixed-position sound effects.
There is also a temporary bug with Youtube and 360 videos that makes the audio play back wrong if you go to Youtube to watch this video. Please watch it here on this page! If you are on a phone, it will not play back correctly here either, and you should watch the Binaural version at the bottom of this blogpost.
For the most immersive experience, we recommend viewing the teaser in VR using Meta TV or YouTube VR.

1. Recording a 360 Video with Ambisonics Audio

Recording was one of the most seamless parts of the project, thanks to the combination of the Nomono Sound Capsule and the Insta360 X4. Setup was fast: no separate audio bags, no bulky gear – just a single system that handled everything.

As 360 cameras continue to improve, tools like the X4 have made this space far more accessible. That said, we quickly realized that 8K resolution doesn’t go as far as you’d expect once it’s stretched across a full sphere. If you’re using auto settings, your footage may suffer. Lesson learned: manual exposure is the way to go, and we’ll be applying that in our next shoot for sharper, better-looking results.

2. Post-Production: 360 Video Meets Ambisonics Audio

Post-production is where things start to get complicated. We shot the film in February 2025, and since then, the tools have already improved. Insta360 Studio now supports importing from various 360 cameras and includes support for 1st Order Ambisonics audio – removing some of the friction in muxing (more on that later).

We edited the full documentary in DaVinci Resolve, which has solid support for Ambisonics mixing. However, it still requires you to manually label Ambisonic files on import, and doesn’t support muxing audio and video into a single deliverable on export. That means extra steps and third-party tools are still needed to finish your immersive project.

Also Davinci Resolve crashes if you try to import an EBU ADM file on mac, but works well on PC. However importing Dolby Atmos ADM works great on both PC and Mac. So if youre working on a project for Apple Vision Pro youre in luck.

There’s hope, though. Google is actively developing the Eclipsa audio format, which could make this whole workflow more streamlined. With major players like Google stepping in, we’re confident the landscape will evolve quickly. Immersive content is still in its early days, and the pace of change is accelerating.

3. Muxing: The Metadata Maze

One of the biggest challenges was muxing – combining your video with 3rd Order Ambisonics and a headlocked stereo track (used for voiceovers, music, or fixed effects). This process is tedious, fragile, and requires unsupported legacy tools. But once you have it up and running, it works great.

Currently, only one tool supports this: the now-discontinued Facebook 360 Encoder, which was part of the Facebook Spatial Workstation. It hasn’t been updated since 2018, but thanks to a modified version maintained by the late Angelo Farina (to whom we owe deep gratitude), we managed to get it working - after two months and five different Macs.

If you’re okay with 1st Order Ambisonics and no headlocked stereo (not ideal for music or narration), you have more options - but they’re still clunky. Adobe Premiere supports it, but only 1st order ambisonics and at a rather expensive cost. Or you can try using the old Spatial Media Metadata Injector, though installing it requires some technical finesse.

Imagine a world where you could export fully muxed, metadata-rich 360 video straight out of DaVinci Resolve or Premiere. We’re actively advocating for this and are in discussions with companies like Blackmagic and Google to help make it happen.

4. Publishing Immersive Content

Playback should be the easy part - but unfortunately, it’s another friction point.

If you’re watching on a Meta Quest or Apple Vision Pro, immersive content playback is generally solid. But for those without VR headsets, things get tricky.

Facebook offers headlocked stereo and 1st Order Ambisonics support - but only in the mobile app. Playback from a desktop browser doesn’t work properly.

YouTube also has some issues. In the mobile app the head tracking doesnt seem to work. And browser playback often results in incorrect channel mapping and broken spatial orientation. Oddly, embedding a YouTube video on a webpage (like this one) fixes these issues (if you are viewing from a desktop browser). We’ve reported the bugs to YouTube, and they’ve even filed an internal ticket - but the problem remains unresolved for now. Perhaps the release of Eclipsa will solve this.

For now, we recommend watching our teaser in a VR headset if possible for the best experience. You´ll find it on Youtube VR and Meta TV. If not, you’ll get the most accurate playback by watching the embedded video right here on our blog.

This article was written on June 10th of 25. By the time you read it, things might have changed.

Final Thoughts

Immersive content creation is still in its early days, and while tools like the Nomono Sound Capsule make recording easier than ever, the post-production and especially publishing pipeline still needs serious improvement.

But we’re getting there. Every update, every workaround, every piece of content we publish brings us - and the wider community – one step closer to a truly accessible workflow for immersive media.

We’ll keep sharing what we learn. And we’ll keep pushing for better tools, better standards, and a better experience for creators and audiences alike.

Until then, enjoy the teaser for our 360 documentary The Goatboat – and stay tuned for the full release at the end of June 2025.

Here is the Binaural version of the video if you are struggeling with the headtracked version: