FAL Vidu Q1 Reference To Video (Soze)
The node with a movement-amplitude dial, and not much else in the way
- image_1
- image_2
- image_3
- image_4
- image_5
- image_6
- image_7
- image_batch
- video
- video_url
- status
Vidu is Shengshu's video model, and Q1's party trick is a control most video models don't give you: a movement_amplitude dial. You get to tell the model how much motion you actually want, from small (locked-off, gentle) to large (camera swooping, things moving). That single control makes it much easier to get a specific kind of clip instead of hoping the model's default energy matches your intent.
The inputs
Standard pack citation format here: @Image1, @Image2, etc., with up to seven reference images in image_1–image_7 (or image_batch, capped at 7). The prompt is capped at 1500 characters - shorter than some of the other nodes, so be economical. aspect_ratio is limited to 16:9, 9:16, and 1:1.
Then the dials that make Vidu Q1 distinctive:
movement_amplitude-auto,small,medium, orlarge. Start withautoto see what the model wants to do, then pin it down. A talking head wantssmall; an action beat wantslarge.bgm- a plain boolean, default off. Flip it on and the clip comes back with background music. It's a binary - you don't get to pick the track, you get whatever the model decides fits.seed- 0 = random; set one for reproducible iteration.
No resolution picker, no audio sophistication, no duration option at all - Vidu decides the length. It's a focused node.
How it works
Same FAL machinery as the rest of the Soze pack: reference images upload to FAL's CDN, the node POSTs to fal-ai/vidu/q1/reference-to-video, waits synchronously, then downloads the mp4 to your output folder as a VIDEO object. Outputs are video, video_url, and status. Nothing runs on your GPU.
The movement_amplitude value is passed straight through to the API, so it's worth treating as the primary creative control - most reference-to-video models bury this decision in the prompt and hope. Here it's explicit, which is the whole reason to reach for Vidu over the generic options.
Install and key
From the Soze pack ("Quality of Life Nodes for ComfyUI"). ComfyUI Manager → search "Soze", or:
cd ComfyUI/custom_nodes
git clone https://github.com/SozeInc/ComfyUI_Soze.git
pip install -r ComfyUI_Soze/requirements.txt
Restart, then set FAL_KEY as an environment variable or in ComfyUI_Soze/config.ini:
[API]
FAL_KEY=your_key_here
Key at fal.ai/dashboard/keys; the config file is gitignored, don't commit it.
Common issues
- Too much or too little motion. That's exactly what
movement_amplitudeis for - don't fight it in the prompt, just turn the dial. - Skips with no output. Needs at least one reference image; it's reference-to-video only.
- Length feels fixed. Q1 doesn't expose a duration control, so if the clip length doesn't suit you, this node isn't the place to fix it - frame your references for the timing you want.
Inputs (13)
| Name | Type | Default | Description |
|---|---|---|---|
| prompt | STRING | Max 1500 characters. | |
| aspect_ratio | COMBO | 16:9 | 3 options: 16:9, 9:16, 1:1 |
| movement_amplitude | COMBO | auto | 4 options: auto, small, medium, large |
| bgm | BOOLEAN | false | Add background music. |
| seedopt | INT | 00–18446744073709550000 | 0 = random. |
| image_1opt | IMAGE | Reference image 1 (cite as @Image1 in the prompt). | |
| image_2opt | IMAGE | Reference image 2 (cite as @Image2 in the prompt). | |
| image_3opt | IMAGE | Reference image 3 (cite as @Image3 in the prompt). | |
| image_4opt | IMAGE | Reference image 4 (cite as @Image4 in the prompt). | |
| image_5opt | IMAGE | Reference image 5 (cite as @Image5 in the prompt). | |
| image_6opt | IMAGE | Reference image 6 (cite as @Image6 in the prompt). | |
| image_7opt | IMAGE | Reference image 7 (cite as @Image7 in the prompt). | |
| image_batchopt | IMAGE | Optional IMAGE batch (capped at 7). Ignored if any image_N slot is connected. |
Outputs (3)
| Name | Type | Description |
|---|---|---|
| video | VIDEO | — |
| video_url | STRING | — |
| status | STRING | — |