Spatial Video recording for iPhone 15 Pro introduced in iOS 17.2 beta 2

AppleInsiderAppleInsider
Posted:
in iPhone edited November 2023

At a glance, there don't appear to be many new features in the second beta of iOS 17.2 -- with the exception of Spatial Video recording for Apple Vision Pro.

Settings toggle in iOS 17.2 beta 2 for Spatial Video recording
Settings toggle in iOS 17.2 beta 2 for Spatial Video recording



One of the features of the Apple Vision Pro is that it includes Apple's first three-dimensional camera, capable of taking spatial photographs and videos that can be viewed via the headset. Apple promised the feature would arrive "later" for iPhone 15 Pro, and the first vestiges of it are available in iOS 17.2 beta 2.

Videos are shot in 1080p, and Apple says that they will take about 130 megabytes per minute of video -- which is about 40% more than it takes for a minute of non-3D 1080p video. The files are stored in what Apple calls "Apple HEVC Stereo Video" format, with the "hvc1" codec.

Both left-eye and right-eye views are carried in each video sample of the track, and the format allows the primary eye to be either the right or left. Apple requires that a reader that does not support multiview video should be able to decode the base layer of the video track for presentation without a stereoscopic treatment for backward compatibility features.

How the tracks will differ when shot on the Apple Vision Pro, versus on an iPhone 15 Pro isn't clear. The camera separation on the Apple Vision Pro approximates that of the human eye gap, while the center of the lenses on an iPhone 15 Pro is about a half-inch.

AppleInsider will test this feature in the coming days. Ultimately, the final judgment will have to wait until the Apple Vision Pro ships in 2024.

Read on AppleInsider

Comments

  • Reply 1 of 7
    nightwatchnightwatch Posts: 50member
    I wonder if Apple could include Spatial Audio recording on the iPhone at some point, (in other words, a built in ambisonic mic with an SoC co-processor and a preference on how to decode it, unless it's Spatial Audio only - or at least make the files compatible with Logic Pro ). The mic array will have to somehow use an algorithm that would create an Ambisonics array out of their mics, because physically it's impossible to record Atmos/Spatial Audio ready audio without an ambisonic capsule array. I would think a special renderer on the SoC could handle this but I'm not exactly sure if it's possible. 

    If so, that's a feature that I'd buy a new iPhone for alone. Small market I know. 
    jellybellybyronlAlex1N
  • Reply 2 of 7
    eightzeroeightzero Posts: 3,069member
    I'm just gonna say what everyone is thinking: the big users of this tech will be the porn industry. 
    jellybellyScot1
  • Reply 3 of 7
    jellybellyjellybelly Posts: 111member
    I have to wonder if next year’s iPhone 16 Pro iPhones might have an extra ‘regular’ wide camera about 2.5” center to center away from current location towards the other end of the iPhone for landscape mode.  That’s the average distance between the center of each iris, or between each retina (2.5”= 63mm average).  That would be to shoot in landscape. Optionally it could be fudged for lower end of range for 2.2” or 67mm for portrait mode and located across the top.  

    Since this  is about a feature in the next release of iOS, maybe it’ll utilize two of the cameras that exist in the current iPhones, fudged down to that distance to get some depth with obvious improvement with the Vision Pro camera separation distance when it comes out. 

    Considering Apple’s computational use of two cameras in its current lineup, it seems that they can achieve some depth with the current camera separations. Hmmm. Perhaps they can use the three cameras in the Pro Max and get a depth perception that’s better than two.  Eg. Average the two cameras near the edge that are in vertical layout and compile that average with the inward camera. There’s not a lack of computation ability with current hardware.  The LIDAR depth map already useful for objects reasonably close can also be added into the computation.  

    This is getting interesting.  If the 3D effect is of decent quality this will really get people chomping at the bit to view their clips on the Vision Pro.  
    Apple: “Make a 20 minute appointment, with submission of your lens prescription if you have vision correction, and view the Vision Pro at an Apple Store and bring your iPhone clips to view in 3D.” 

    Heck yes, sign me up.  Then take my money. 
    byronlAlex1N
  • Reply 4 of 7
    jetpilotjetpilot Posts: 34member

    Videos are shot in 1080p, and Apple says that they will take about 130 megabytes per minute of video -- which is about 40% more than it takes for a minute of non-3D 1080p video. The files are stored in what Apple calls "Apple HEVC Stereo Video" format, with the "hvc1" codec. 

    The camera settings app on my iPhone 15 Pro Max states that 1080p HD at 30 fps uses about 65 MB per minute of video, so the spatial video recording using 130 MB per minute of video recording would exactly double the file size (which makes sense since it's recording data from two cameras). Not sure why the article states that it only increases the file size by 40%.
    appleinsideruser
  • Reply 5 of 7
    jellybellyjellybelly Posts: 111member
    jetpilot said:

    Videos are shot in 1080p, and Apple says that they will take about 130 megabytes per minute of video -- which is about 40% more than it takes for a minute of non-3D 1080p video. The files are stored in what Apple calls "Apple HEVC Stereo Video" format, with the "hvc1" codec. 

    The camera settings app on my iPhone 15 Pro Max states that 1080p HD at 30 fps uses about 65 MB per minute of video, so the spatial video recording using 130 MB per minute of video recording would exactly double the file size (which makes sense since it's recording data from two cameras). Not sure why the article states that it only increases the file size by 40%.
    jetpilot said:

    Videos are shot in 1080p, and Apple says that they will take about 130 megabytes per minute of video -- which is about 40% more than it takes for a minute of non-3D 1080p video. The files are stored in what Apple calls "Apple HEVC Stereo Video" format, with the "hvc1" codec. 

    The camera settings app on my iPhone 15 Pro Max states that 1080p HD at 30 fps uses about 65 MB per minute of video, so the spatial video recording using 130 MB per minute of video recording would exactly double the file size (which makes sense since it's recording data from two cameras). Not sure why the article states that it only increases the file size by 40%.
    They’re using HEVC Compression—a  better method for storage. I’m try an also a process optimized for Apple Silicon in hardwares. 
  • Reply 6 of 7
    badmonkbadmonk Posts: 1,295member
    eightzero said:
    I'm just gonna say what everyone is thinking: the big users of this tech will be the porn industry. 
    Agree—don’t underestimate the porn industry’s role in moving technological change forward for better or worse.  It will be interesting to see how Apple turns a blind eye to this since Steve Jobs was notoriously anti-porn.
  • Reply 7 of 7
    danoxdanox Posts: 2,875member
    eightzero said:
    I'm just gonna say what everyone is thinking: the big users of this tech will be the porn industry. 
    The only one in the gutter is you………
Sign In or Register to comment.