How to control the pipeline procession in the camera (ricoh theta v)


Hi, recently I wanna measure the live streaming performance. Is that possible to control the pipeline design in the Ricoh Theta V camera? For example, I wanna know how long does it take for the stitching, encoding, and transmission. I checked the Plug-in API, but don’t have good ideas about that.
In this topic, it seems like we can control the stitching manually by a plug-in.
Thanks in advance for any suggestion!


You can control the stitching with the plug-in for still images and video saved to file. I don’t think you can set it when the camera is live streaming.

The camera uses Android 7.1.1 as the OS. You should look for generic Android information and you may be able to apply it to a plug-in.


Really appreciate your reply! One more question: can I live streaming the original dual image data so that I can do the stitching, encoding by myself?


I do not think there is an easy way to live-stream the dual-fisheye video. You’d need to save it to video file and then stream the file.

You may be able to use getNumberOfCameras() and then try to use each of the dual-fisheye cameras separately. I don’t know if this is possible. I have not tried. No one has reported success with this technique. It’s likely that it will return only one logical camera (combined both lenses).

If you’re doing object recognition and your app can live with an image every 1 second, you can take a still image dual-fisheye every second over the network. It’ll likely be every 1.5 seconds.


Hi, right now, I wanna have a measurement study on the time consumption of the pipeline. However, since the real-time stitching is a black box for us. all I can get is the stitched frame during the live streaming. Thus I cannot measure the stitching and shooting time separately. Is the shooting time negligible compared with the stitching time during the live streaming? Thanks!