VR Post Production Workflow Part II:

Editing Social Videos vs. Immersive Documentaries

At Contrast VR, we divide our content into three categories: social videos, short 360 documentaries and immersive VR experiences. Considering the intended distribution channels for these pieces, we choose the best camera equipment to be used, which will in turn reflect the amount of time needed in post production.

Social videos are 1.5 to 2 minutes 360 videos that are created for our Facebook and YouTube 360 social platforms. They are shot using consumer friendly cameras, such as the Samsung Gear 360 or similar models (Kodak Pixpro, Nikon KeyMission, Xiaomi Yi, Gopro FUSION, etc.).

Short documentaries, on the other hand, are 4 to 6 minutes videos primarily made to be viewed on headsets (although we distribute them on social media channels as well). We shoot these on the GoPro Omni. These are more in-depth stories compared to the short social videos. Our immersive VR documentaries — the ones that are about 8 to 10 minutes long — are shot with cameras that capture stereoscopic images like the Nokia Ozo, for instance.

There are certain practices that are quite general when it comes to editing 360 videos. It is, for the most part, a lot more technical than creative. Of course, you can play with methods to build your story in an interesting way, but you can’t really switch shots every two seconds as you could in standard videos. That can cause extreme disorientation and nausea for the audience, especially if the video is being watched in the headset. One of the most important things in 360 video post production is to get the timing right — shots need to be long enough for people to be able to recognize the space they are being immersed in, but not so long that they get bored and lose interest in the story.

Before you start editing, it is necessary to understand how a 360 video translates into an editable file. A video in 360 format becomes a panoramic image in your editing software timeline. It is important to be aware of how the different areas of the shot represent different parts of the 360 environment. The first thing the viewer will see is the center of the shot. Check the image below:

A basic practice worth paying attention to is the “offset” effect in Adobe Premiere that allows you to ‘reorient’ your shot. Essentially, you can slide the image from left to right (or vice versa) in order to control what part of the 360 video is in the center viewpoint.

In this way, you can control the first thing a viewer will see after each edit point, guiding the viewer through the 360 space depending on the action of your scene. When you are directing or shooting a 360 piece, it is important to know how you will want to edit the video; shooting sequences with characters is a good way to keep viewers moving their point of view inside the 360 space.

1. Editing Monoscopic vs Stereoscopic videos

The first step to start editing is to set up your sequence settings. When creating a new sequence, make sure it is all set properly for 360 video by selecting the correct sequence preset you are going to work on (monoscopic/stereoscopic). This way you will be able to preview the shots not only in the flat format, but also in 360. In case you are working with spatial audio as well, you can adjust the sequence settings to ambisonic audio. See the image below:

Depending on the camera you use, you can capture monoscopic (flat) or stereoscopic (3D) images. If your video is being distributed only on social, it’s easier to use cameras like the Samsung Gear 360, for example, since Facebook and Twitter don’t support stereoscopic video.

On the other hand, if your video is going to live primarily on headset channels such as Oculus VR, Viveport, DayDream or even on YouTube, the best choice is to use a high-end camera that captures 3D images for a fully immersive experience (examples of cameras would be the GoPro Odyssey, Nokia OZO, Insta360Pro, Jaunt etc). While editing stereoscopic videos, you will see on the same canvas top and bottom images, which have a slightly different angulation that causes the 3D effect.

2.1 Editing 360 Social Videos:

The timing of a 360 social video is a little different than videos that are meant to be watched in a headset. Normally the majority of shots for social videos should range from 10 to 15 seconds and the total length of our videos is no longer than 2 minutes. Social media users tend to stop watching 360 videos within seconds, so if the beginning of the video is not visually grabbing and engaging, you will lose viewers fairly quickly.

Most users also don’t turn on sound while watching videos. The ideal is to embed the subtitles and create brief, precise title cards that will give the audience an idea of what the video is about, especially in the first 3 to 5 seconds.

The best thing to do is to edit the flat video and then watch it in 360 a few times in Premiere to adjust the timing of the shots, title cards and animations.

Of course, the audience is free to look around and find parts of the shots they find the most interesting, but good 360 stories include some sort of visual or audio cues to help guide the viewer around the 360 space. A good strategy is to ask the subjects, during production, to walk from point A to point B, so that in post production you will be able to create a sequence that lets you explore the 360 space without getting lost.

A useful exercise to make sure the viewers are not getting lost in the 360 space is to check Facebook. Its analytics tool provides heat maps of videos with over 20 thousand views, which is a great method for editors to check if people were looking at the spots of the 360 shot necessary to follow and understand the story.

2.2 Editing Immersive VR Documentaries:

Short form 360 documentaries and immersive VR experiences follow different rules than social video content in some ways. These are longer form documentaries that are meant to be watched in the headset and provide a truly immersive experience. Shots can be longer than 15–20 seconds, depending on the storyline and what is happening in the shot. The length needs to give the audience enough time to feel included in the scene, as well as simultaneously follow the character’s journey.

On the technical side, you will be working with much higher quality videos than with social videos. When doing so, it is usually a better bet to work on a rough cut using lower quality, rough stitched files (2K) and only fine stitch shots once you lock the picture. Once you reach a point where you know exactly what shots you will use on your final cut, and their time codes, you will be ready to work on fine stitches and render higher quality stitched footage [4K (4096x2048 mono/4096x4096 stereo) or 6K (6144x3072 mono/6144x6144 stereo).

If you decide to work with 6K footage, be aware that you will have to convert it to 4K later on in order to post your video on Facebook or play it in the Samsung Gear VR headset or Oculus. Samsung Gear VR, for example, doesn’t accept files larger than 4Gb, which is something you should take into account while exporting your 360 video.

Spatial audio is another important element of our immersive VR documentaries. It is essential to deliver a fully immersive experience to the audience not only through thoughtful and creative use of 360, but also by giving them a feel of the environment through a proper spatial sound mix that helps guide them through the scene. It is also interesting to add sound effects to the scenes in which they are suitable — to complement animations in a way that makes it look more realistic and engaging, for example. A good example of the use of spatial audio design with sound effects and animations is in our first original doc, I Am Rohingya. To learn more about spatial audio mix, stay tuned for part the Part III blog post on Post Production.