<aside> 🛠

Turn an Image List Back Into a Video

The finishing step for any frame-by-frame video pipeline: take an image list and reassemble it into a video. Set the output frame rate (or leave it to fall back to the source fps) and go. Completes the classic Extract → Process → Inject → Reassemble pattern — the video comes in as frames, gets improved one image at a time by any of the image tools in the studio, and comes back out as a finished video.

</aside>

What it does

Video Reassemble takes an image list (from any list-producing upstream node — a decoded video, a batch generation, an Extract → Inject pipeline, etc.) and encodes it back into a video. The images become sequential frames in list order. You set the Output frame rate; if the source video's fps is available upstream, it's used as the fallback default. The result: a single video file, ready to save, preview, or feed into video-level models (super-resolution, generation, etc.).

Problem it solves

• Close the video pipeline loop – The exit point for any workflow that processes video frame-by-frame using image tools
• Full-video enhancement with image models – Decode → apply image tool per frame → Reassemble = enhanced video (upscaled, denoised, colorized, restored, stylized)
• Frame-rate conversion – Reassemble at a different fps than the source to speed up, slow down, or standardize video output
• Custom slideshow / stop-motion output – Turn any curated image list into a video (product shots, generation sequences, timelapses)
• Deterministic output – Same list + same fps = same video every run
• Bridge back to video models – After per-frame image processing, reassemble to feed into video super-resolution (SparkVSR), video generation, or video segmentation nodes

Input/Output

<aside>

• Input: Image list
  • Format: A collection of images produced by any list-emitting upstream node (video-frame decoder, batch generator, List Inject output, curated sequence, etc.)
  • Order: Images are used as frames in list order — first item = first frame
  • Resolution: All frames should share the same dimensions for a clean output; mismatched sizes may require an intermediate resize step </aside>

<aside>

• Output: Single video

  • Format: Standard video (MP4, etc. — depending on the studio's default encoder)
  • Frame rate: As set on the node (falls back to source fps if left at default)
  • Duration: list length / frame rate seconds
  • Content: All list items encoded as consecutive frames

  

ebf8274b-d767-4c55-8bae-10e545d3fe16.mp4

f7d2110f-47c5-4b7b-80fb-c6bbf2258357.mp4

</aside>

Configuration Options

• Output frame rate (falls back to the source fps): Frames per second in the reassembled video.
  • Default: 24 (falls back to the source video's fps if that information is available upstream)
  • UI: Slider — drag to the desired fps; the current value is shown to the right
  • Common values:
    • 24 – Cinematic look (film standard)
    • 30 – Standard video (broadcast, social media)
    • 60 – Smooth motion (gaming, sports, high-motion content)
    • 12 – 15 – Stylized / stop-motion feel
  • Behavior: The video's playback duration is list length ÷ fps — so raising the fps without adding frames makes the video shorter and faster; lowering it makes the video longer and slower.

How to Use

<aside>

1. Prepare the image list → Set up whatever upstream flow produces the frames you want in the final video (decoded + processed video frames, curated image sequence, batch generation output, etc.)
2. Connect the list → Wire the image list into Video Reassemble
3. Set the frame rate → Drag the Output frame rate slider (or leave it to fall back to the source fps)
4. Run Flow → All frames in the list are encoded as a video at the chosen fps and emitted as the output </aside>

Use Case Tips

For Frame-by-Frame Video Enhancement (Canonical Pattern):