ResolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More sounds simple. It’s this nice, neat number that helps us compare quality from cameras we haven’t gotten our hands on or seen sample footage from.
Even though we know a camera’s quality is measured by so much more than its resolution–I’ve seen loads of poor quality video from cameras with high resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More specs–we use this number a lot in our headlines.
But, using resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More as a baseline for quality is complicated, and it comes with a lot of caveats. You also have to think of VR headsets’ field of viewThe angle of space viewable from a given lens position., resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More loss due to stitching, streaming capabilities and more.
As the 360 world wraps itself around my head in the form of a Gear VR, Rift or Vive, I’ve been attempting to wrap my head around what resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More means in the world of spherical video.
Note: I’m not a resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More expert, but I’ve researched a lot, read a lot, and I have a very full piece of scratch paper in my notebook outlining the math 🙂 And, I’ve linked to original sources and recommended reading so you can dig into resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More even further. This is how I’ve come to understand it.
Here’s what I’ve learned so far.
How is resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More measured, and what does that mean for 360 video?
What does 4K, 5.2K or 8K even mean? This number refers to the horizontal pixel count of your camera’s CCD. In the case of 4K, that means around 4000 horizontal pixels, but not exactly 4000 pixels. The 4K umbrella includes Ultra HD (3840 x 2160), Ultra-Wide 4K (3840 x 1600) and DCI 4K (4096 x 2160).

Source: Wikimedia Commons
Already, that’s a bit confusing, but it’s made more confusing by the switch in standards from the vertical pixel counts of yesteryear (1080p, 720p or 480p). By those standards, 4K would actually be called 2160p. By the new standards, 1080p would actually be called 2K.
What makes 360 video resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More different from fixed frame resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More is actually its display. Instead of showing 4000 horizontal pixels in a tidy, tiny rectangular frame, you’re spreading those 4000 pixels in a complete sphere around the viewer.
Why does 4K look so different in a VR headset versus a 4K laptop screen or TV?
The viewer is only seeing around ¼ of those pixels at any given moment, depending on the field of viewThe angle of space viewable from a given lens position. of the VR headset. For example, the HTC Vive has about a 100-degree horizontal field of viewThe angle of space viewable from a given lens position.; Oculus Rift, 80; and the Samsung Gear VR, 96. The average human’s horizontal field of viewThe angle of space viewable from a given lens position. is about 120 degrees. Here’s a great resource explaining field of viewThe angle of space viewable from a given lens position. in more detail.

Source: 360Rize
So, if we average a headset’s horizontal field of viewThe angle of space viewable from a given lens position. to, say, 90 degrees, the viewer can only see one quarter of the 360 space at any given time. That means if a 360 camera could capture 1080p video, the viewer would actually be experiencing something more similar to a 640 x 320 80s-era computer monitor. 4K video would look like 1K, etc. To achieve true 4K resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More wherever the viewer looks, you’d actually need to capture 16K! Here’s a great chart from 360Rize.
These numbers will also change if the FOV changes. If our next-gen HMDs have a wider FOV of 120 degrees, 4K video would look like 1.3K, 6K would look like 2K, and so on.
How does stitching 360 video affect overall resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More?
Beyond dividing total resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More by field of viewThe angle of space viewable from a given lens position., when it comes to 360 video, you also have to consider the resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More you’ll lose when stitching. We don’t yet live in a perfect, seamless, stitch-free world.
For multi-camera rigs, you can’t just add up the resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More of each camera in your rig because you’re going to lose some of that to the overlap required to effectively stitch your footage. Of course, it probably isn’t worth having very obvious stitch linesThe seams in a 360° video where footage from one camera has been combined with another. to get a tiny boost in resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More, and you need enough overlap to have clean stitching.
What would we need to shoot to actually achieve human vision?
Another important factor to consider when it comes to consumption is pixel density.
A human with 20/20 vision can see roughly 60 pixels per degree. We can achieve 60 pixels per degree while looking at a 48-inch 1080p monitor from 6 feet away, or an iPhone 6 screen at 15 inches away. That means, even for someone with perfect vision, even if the images were higher resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More, the user wouldn’t be able to notice a difference. This article from RoadToVR explains it best:
“This means that if there an image with 3,600 pixels (60 x 60) and that image fell on a 1° x 1° area of the fovea, a person would not be able to tell it apart from an image with 8,100 pixels (90 x 90) that fell on a 1° x 1° area of the fovea.”
The HTC Vive, with 1080 horizontal pixels per eye and 100-degree horizontal field of viewThe angle of space viewable from a given lens position. gives a maximum pixel density of 11 pixels per degree.
Visbit explains how this pertains to spherical video resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More really well in this blog post. The takeaway is that your average 1080p 360 video actually serves up roughly 6 pixels per degree, or 20/200 vision. That’s like looking at your eye doctor’s chart and only being able to read the very first letter!
Even if we captured higher resolutionThe number of pixels in an image, typically presented as a ratio of the total pixels on x axis to the total pixels on th... More 360 video, how would we deliver it?
Of course, capturing higher and higher resolutions is only half the battle. And, honestly, it’s the winning half.
It’s a lot easier to capture 16K video than it is to actually deliver that experience. To resolve the deliver, we would need better bandwidth, new compressionA process of lowering overall image quality by reducing the amount of data in a file in order to make it more accessible... More methods, and view-optimized streaming. And, after that/along with that, better HMDs with wider FOVs and higher screen resolutions.
So, it’s all a moving target.
It’s complicated. And confusing.
It’s changing fast.
And it requires math.
And, I’m probably over-generalizing it here.
But, hopefully this overview gives you a better understanding of how 360 video resolutions actually play out in a headset, how they matter (and how they don’t matter) so you’ll have a clearer picture when you’re reading those resolutions in our headlines.
If you want to dig in, here’s some recommended reading:
- Pixel Counting in 360 Degree Videos: Reinventing Resolution
- 4K VR 360° Video: What is it and How Can I Produce it?
- 5 Things You Should Know About 360 Video Resolution
- Visbit opens beta access for its 4K+ 360 video streaming service
- Why do all the 360 VR videos today look so pixelated?
- Field of view face-off: Rift vs Vive vs Gear VR vs PSVR
- Understanding Pixel Density & Retinal Resolution, and Why It’s Important for AR/VR Headsets