TEST RECORD · ← all field reports

FIELD REPORTS/VIDEO/ISSUE #117

Turns one photo into a short video where the camera circles a frozen moment and returns to the first frame

A photo is one moment seen from one spot. This add-on for MiniMax's H3 video model walks the camera all the way around that moment and brings it back to where it started, and that clean loop is the part we kept rewatching.

MODELMiniMax-H3-360-Orbit-LoRA
PUBLISHEDOctober 1, 2026
READ TIME3 min
CATEGORYVIDEO

Field notes

01 What it does

You give it one photo, and it returns a short clip in which the camera travels a full circle around the subject while time stands still. A dancer stays mid pose, hair stays in the air, and only the viewpoint moves. The last frame is your original photo again, so the clip loops cleanly or joins onto another one without a visible jump.

That return trip is the hard part. Asked to start and end on the same image, the H3 video model on its own barely moves the camera; given only a starting image, it drifts off and never comes back. This add-on, a small set of extra weights known as a LoRA, teaches it to make the whole lap. If you read about Meridian in issue 115, which re-shoots a scene from a new angle, this one does a single job: the complete 360 degree turn.

It was trained on just 28 orbit videos rendered around 3D captures of people, so it is at its best with one person in the frame. Each clip runs about three seconds.

02 How to try it

The demo Space on Hugging Face is the quickest test. Click the ballerina example first: its result is already saved, so you see a finished orbit right away. Then upload your own photo of one person, ideally full body with some background around them, keep the square 768 canvas the add-on was trained on, and press Orbit 360°. Watch the back of the subject, the side the camera never saw, because that is where it has to invent detail. Generation runs on Hugging Face's shared free GPUs, so a free account gets a limited number of takes. Running it yourself means loading it on top of MiniMax H3 in a tool like ComfyUI, which needs a large NVIDIA card.

03 Caveat

The back of your subject is invented, not recovered, so clothes and hair on the far side can come out wrong, and small movements like blinks can still slip in. Subjects other than people are untested. The add-on also inherits MiniMax's H3 community license, which does not cover use in the United States, the European Union, the United Kingdom or South Korea.

04 What you can do with it

  • Turn a portrait into a frozen spin for a profile video or a reel.
  • Show an outfit or costume from every side using one full body photo.
  • Chain several loops into a longer shot, since each clip ends where it began.
  • Give a cast photo or a lookbook page a moving cover shot.

Try the demo

View model page

Read this issue on neuralexpedition.com →