Z1 Video Stitch Quality Problems with Internal Stitching?

[craig]

I normally don’t use video on the Z1 as most businesses are using still images at the current time. However, as I am getting more questions about video, I am doing basic testing with video.

With Z1 firmware 2.00.1, I am getting significant stitching aberrations. In the portion below, you can see that the hanging branch is erroneously duplicated.

The bitrate is intentionally set to “low” to reduce the file size. However, I wouldn’t expect the bitrate adjustment to impact the stitching. I don’t recall the video stitching problem in my previous tests.

The video file from my camera is here.

I’m going to try another test with the saving the video as dual-fisheye and then stitching the video file with the desktop app.

Has anyone else encountered similar problems with in-camera video stitching?

---

Still Image

looks fantastic

Desktop Stitching

The desktop stitching is much better.

[jcasman]

I replicated your settings, @codetricity, and took 2 different types of videos and 1 image to test.

My Z1 is firmware 2.00.1

I used a tripod, and did my best not to move the camera.

I used the RICOH THETA mobile app to set the following settings:

1. In-camera stitching, 4k, Bitrate: High
2. Single picture
3. Without stitching, transfer as dual fisheye, stitch on desktop using RICOH THETA desktop app, Bitrate: High

And I made sure to point the side of camera at something with lines. In this case it was the shadows of telephone wires.

In-camera stitching (screenshot)

Original video

Single picture

Original image

Stitch on Desktop

Original video

Summary

“At a glance,” actually I thought all three looked very good. Upon closer inspection, the stitching for a still image and stitching on the desktop look fairly close to flawless. Stitching in-camera is excellent, but I found a broken shadow of a telephone wire.

[craig]

The first link is to a video you stitched on the desktop, not in-camera. The third link is a video stitched in-camera. Can you swap the links?

It would be a better test closer to the tree.

It would be interesting to repeat the test walking with the tripod. Put two markers, like rocks about 30 feet apart, preferably in that alley, and walk down it, from rock to rock. Then, repeat the process with dual-fisheye.