Content
Who knows the difference between Head-Locked and non-Head-Locked? Or what are the advantages of using the Head-Locked method and Ambisonics? In this article, you will find out!
Strictly speaking, the focus here is on trying things out. Sometimes it’s not that easy to get everything to work on VR glasses. My wdm_placeholder_number_1gives you an idea that the devil is in the details.
With some videos, it’s just not so clear whether the sound is really 3D audio or not. Therefore I created a dummy video to find this out. So the video is not very pretty, but it serves its purpose.
Below are the links to download and try it yourself. Make sure to download the correct file format for optimal VR playback.
I am happy about suggestions for improvement because there are now already several versions internally. But with the comparison head tracking vs. head-locked mna can certainly improve et what.
A test dummy, in this case, isn’t a stand-in app or a simulated user. It’s a short 360° video built for one job: making it obvious, by ear, whether spatial audio is actually working.
Stereo, head-locked, and head-tracked Ambisonics all sound different once you put on a headset, but from watching a video you can rarely tell which one you’re getting. So instead of describing the difference in the abstract, I built a dummy that walks you through it: the same voice and the same piece of music, played back once as head-locked and once as head-tracked Ambisonics, back to back.
Use it to check your own player or headset the same way I use it to check mine: if head-tracking is working, the sound should stay anchored in the room when you turn your head. If it doesn’t, something in your playback chain stripped the spatial information out.
Simplified you could say that our good old stereo in 360 videos is head-locked audio. However, it is only in combination with Spatial Audio that we really talk about Head-Locked.
With head-locked, the audio signals are routed to discrete channels and are not placed in the scene. In stereo, channel 1 is for the left ear and channel 2 is for the right ear.
Thus the audio content moves with the listener, i.e. if the listener turns his head to the left, the sound turns to the left as well. So you always have to be a little careful how you phrase this. If the sound rotates with the listener, it actually remains unchanged, it is independent of the head movement.
The head-locked method is particularly suitable for voiceovers or, even better, for music. Usually, music is only available as stereo audio. However, there are much better variants for including music in 360 videos.
I personally avoid the head-locked track for various reasons. Head-locked audio is supported on various VR platforms and devices.
Here, head tracking adapts the sound to the head movement in real-time. This is also referred to as scene-based. The sound sources are mentally anchored in the room. If you turn your head, the sounds stay where they are in the room, but the character of the sound changes for our ears.
There are several 3D audio formats that make this possible, as I’ve already explained in the Pros and cons for 360 videos . In the world of VR videos, Ambisonics has established itself as a quasi-standard.
I have already written quite a bit about Ambisonics on my blog. The post: wdm_placeholder_number_9describes the pros and cons and gives a small overview of the workflow.
Whether it is first order or higher, be for this test once not too important. To fully experience Ambisonics audio, users need to insert their phone into the VR headset.
A dedicated test file like this saves you the guesswork of troubleshooting a real project: instead of uploading a finished video to YouTube or a VR headset and hoping the spatial audio survived, you check it against the dummy first and know immediately whether the problem is in your file or in the platform.
A few situations where that’s useful:
The Test Dummy can be accessed through various VR apps on Facebook and YouTube, allowing users to download, share, and play VR videos on different devices.
Attention: On Youtube the video only runs with the Firefox browser.
To view the VR video, users can tap the headset icon in the YouTube app.
This version uses pure head-tracked Ambisonics, without any head-locked stereo layer.
If you want, you can also download the video and load it on your VR headset.
Download link: FileDownload
Alternatively, you can search for the Test Dummy video on various VR platforms like YouTube and Vimeo.
In order to get a better idea of the structure and process of the dummy, the script is presented below. Suggestions for improvement are also welcome here. It is aimed at developers and beginners in the field of spatial audio. Since I experience again and again that the difference between head-locked and head-tracked is not clear.
Ambisonics sound is heard at 0° azimuth and 0° elevation, also what would be center with speakers. Time: approx. 0-15 seconds
LOUDSPEAKER: “Hello and welcome to the 360 video demo test, if you turn your head you can still hear the audio from 0°. If you move your head to the left, you will hear the voice turning to the right. This simulates head tracking.”
Ambisonics sound rotates around the head, users are told to keep their heads still. Clockwise rotation around azimuth axis. Time: approx. 0-15 seconds
SPEAKER: “Rotate your head backwards by 0°. As we proceed, we mix in music and let the audio content circle around your head. The rotation is horizontal, without altitude information.
Only the first channel of Ambisonics will be used.
LOUDSPEAKER: “Let’s forget about 3D audio for a moment. I am now a voice in your head. You can turn your head, but my voice remains the same. It is monophonic. This is useful for making it clear to users that a voice is not part of the scene, which is called non-diegetic. This is used for voice-overs like narrators, who are not invisible but tell more about the scene.
Music is heard that is obviously stereo. Time: SPEAKER: “Let me play my favorite record. What you just heard is called head-locked audio. Move your head and you’ll notice that the music moves with you – it’s not placed in the scene. Head-locked audio is an optional stereo stream that can be used for voice-over, or better yet, stereo music.”
Sound comes from the left.
SPEAKER: “We’re not limited to mono, so I can sit next to your ear. Now I’m on the left and whispering to you
Sound comes from the right.
SPEAKER: “Or I can jump to the right and talk to you. Keep moving your head and I’ll always stay on the right.”
A voice is heard again in the center, and the music is heard as Head-Locked.
SPEAKER: “Now you’ve learned the difference between diegetic and non-diegetic. Or, head-tracked audio versus head-locked audio. Let’s try both at the same time. You can listen to music that is not part of the scene as background music – that’s where it’s stored in the headphones. While you can hear my voice as part of the scene as if a person is talking to you.
Now the other way around: voice in the head and music in the scene.
SPEAKER: “Now let’s do it the other way around: I’m not a voice in your head, so I’m not part of the scene. But you can locate the music from the front. So move your head and the music will move around your head, while my voice is independent of where you’re looking. That’s typical of music scenes where you can hear a musician playing while a recording of an interview is playing.”
SPEAKER: “This was the demo test for 360 videos. We hope you learned something about the potential of head-tracking and head-locked audio. But this is just the beginning, contact Martin Rieger to create next-level experiences.”
This version compares both directly: head-locked stereo against head-tracked Ambisonics, so you can hear the difference yourself.
Now that the sound check is done and you have an idea of Ambisonics and other virtual reality audio formats, it’s time to create quality content. I’m happy to help you with that, just contact me:
Are you working on a VR experience and the audio still feels flat?
Get in touch →