Over the last few months, we’ve been steadily rolling out features for YouTube 3D. Early this year, we announced support for the LG Optimus 3D and NVIDIA 3D Vision, and this month we announced support for Samsung 3D TVs. And we’ll keep rolling out to more platforms to make it easier for everyone to get access to user-generated 3D content.
Many readers have been asking how to upload content. 3D uploads to YouTube will work across all supported devices, from red/blue glasses to the latest 3D televisions.
There are two ways to upload content to YouTube in 3D. If you’re just uploading one or two videos, upload content in any supported format and then edit the video settings to tell us that it’s 3D. From a video’s watch page, select “edit info”, then “3D Video”. This will only work on videos you’ve uploaded yourself.
If you’re using the API to upload 3D videos, you’ll want to format your video in a certain way. YouTube supports automatic detection of 3D content with certain file formats. If you upload with a format that supports automatic detection, we’ll automatically set the correct settings so that your video shows up automatically in 3D and shows up properly on phones, TVs, and PCs.
In both cases, YouTube requires that content be uploaded in Side by Side L/R with each frame squashed 50% and preferably with a 16:9 aspect ratio. The following formats are automatically detected when uploaded:
- WebM with the StereoMode setting set to
STEREOMODE_TYPE_LEFT_RIGHT. Note that other modes, including alternate track, are not supported.
- H264 AVC with Supplementary Enhancement Information, Frame Packing Arrangement set. If you are usingx264, the argument
--frame-packing=3will set this correctly. In Handbrake you can set this in “x264 Advanced Option String”.
Be sure to get the squashing and positioning exactly correct. 3D content won’t look right if video formatting is off by even a pixel or two.