Recording a 360 Video with a Ricoh THETA X
A 360 video recorded while you walk through a building is enough for the platform to rebuild the place in 3D. The quality of that result depends far more on how you hold the camera and how you walk than on the camera itself. This article gives the settings to use on a Ricoh THETA X and the habits that make a recording easy to process.
Equipment
- A Ricoh THETA X. A standard unit is all you need.
- A monopod or extension stick with a standard 1/4 inch screw, such as the Ricoh THETA Stick TM-2.
- The lens protector made for the camera (Ricoh TL-3). The two lenses are exposed, and a scratch shows on every frame.
- A Bluetooth remote control compatible with the camera, to start and stop recording without touching it.
- Optionally, a helmet mount if you need both hands free on site.
Recording does not need an internet connection. The videos are added to your project afterwards.
Camera settings
Set the video mode once on the camera and keep it for every recording:
| Setting | Value |
|---|---|
| Resolution | 5.7K (5760 x 2880) |
| Frame rate | 10 fps |
| Bit rate | 60 Mbps |
| Top/bottom correction | On |
| Wireless LAN | Off |
The camera itself requires the wireless LAN to be off at 5.7K 10 fps, which means the smartphone app cannot drive the recording. Use the remote control or the button on the camera.
Ten frames per second may look low, but at this resolution and bit rate each frame is sharp and well exposed, and the files stay reasonable to upload. A higher frame rate or the 8K mode does not improve the result. It only makes the files heavier.
Keep each video under 9 minutes. Between 3 and 7 minutes per sequence is the comfortable range, see the section on splitting the site below.
Holding the camera
- Hold the camera vertical and steady, above your head, about 2 m from the floor. Extend the monopod straight up.
- Do not spin the stick while you walk. The camera should keep the same orientation from the start to the end of a sequence.
- Stay out of the center of the image. If your head or your body fills the middle of the frame, the processing loses its reference on the scene. Above the head is the right place, not in front of the face.
Walking
- Walk at 2 to 3 km/h, roughly half your normal pace. When in doubt, slower is always better than faster: a slow walk means less motion blur.
- Press record, wait 2 to 3 seconds, then start walking. During the recording, avoid standing still for more than 2 to 3 seconds.
- Move smoothly. No sudden turns, no jerky moves, no running, no stairs taken two at a time.
- Close your loops. Whenever the layout allows it, come back to a spot you already filmed rather than ending at a dead end. A route that returns to its starting point is much easier to process than an open line.
- Avoid long straight lines at a constant heading. Add gentle sideways variation, show the same surfaces from several angles, and pass close to distinctive elements: corners, columns, equipment, signs, textured surfaces.
Split the site into sequences
Do not record a whole building in one go. Make one video per room, corridor, floor or zone, of 3 to 7 minutes each, and never more than 9 minutes. Start each new sequence inside an area that the previous one already covered, so that consecutive videos share the same visual content. That overlap is what lets the zones be joined together.
Difficult areas
Some places give the camera very little to hold on to: white corridors, large empty rooms, glass walls, repetitive industrial structures. There, slow down further, walk closer to the walls or to the objects, and give your path a slight curve instead of a straight line.
In low light, apply the same rule and slow down even more. Motion blur is the first thing that degrades a recording.
Give the reconstruction its scale
When no laser scan and no 3D model of the site exist, printed markers give the reconstruction a reliable scale. Print 4 to 6 ArUco markers with an outer square of 18 cm, spread them across the area, and walk within 5 m of each one during the recording.
When a laser scan or a 3D model of the site already exists, make sure the video also covers areas present in that data. The trajectory can then be aligned on it.
After the recording
Copy the video files from the camera to your computer without editing them: no trimming, no re-encoding, no stabilization pass, no compression. Then add them to your project from the Files tab, like any other file (see Add Files to a Project).