Image Sequences

An image sequence is a series of numbered image files, such as render_0001.png, render_0002.png, render_0003.png. Each file is one frame. WATCHOUT optimizes the frames together into a single video asset that you place on the timeline and play like any other video. Image sequences come mostly from CGI and 3D render pipelines. Use them when you need uncompromised quality at the highest fidelity: lossless frames, up to 16-bit color, with no compression artifacts.

When to Use Image Sequences

Image sequences are the preferred delivery format when:

  • Highest fidelity matters. Lossless frames convert straight to a Raw playout codec with no quality loss. There are no codec compression artifacts, and 16-bit frames avoid banding on smooth gradients.
  • 3D rendering. Most render engines (Cinema 4D, Blender, After Effects) output numbered frames. Importing them directly avoids a separate encoding step.
  • Interrupted renders. If a render crashes, completed frames are kept. A video file may be corrupt after a crash.
  • Rendering on site. Re-render only the frames you change. With the Asset Watcher, WATCHOUT updates just those frames, not the whole sequence (see Updating Frames below).

If you do not need lossless, high-bit-depth frames, deliver a compressed video (ProRes or HAP) instead. It uses less storage and optimizes faster.

How Detection Works

When you add an image sequence, the Asset Manager scans the selected folder and runs a detection algorithm:

  1. Extension filter. Only files with a supported image extension are considered (see the table below).
  2. Frame number parsing. The filename is read in reverse to find the trailing digits. For example, render_0042.tiff gives frame number 42. The parser uses the last group of digits before the extension.
  3. Prefix grouping. Files are grouped by prefix (everything before the trailing digits). The group with the most files becomes the sequence.
  4. Contiguity check. The sequence must have no gaps. A missing frame in the range fails the import.

Sequences must be contiguous. A missing frame number in the range fails the import. Verify the render output is complete first. A sequence of frames 1, 2, 4, 5 (frame 3 missing) is rejected.

Supported Frame Formats

WATCHOUT reads these frame formats. For the highest fidelity, use a lossless format at 16-bit (PNG or TIFF).

FormatExtensionsCompressionBit depthAlpha
PNG.pngLossless8 or 16-bitYes
TIFF.tiff, .tifLossless (lossy variants exist)8 or 16-bitOptional
TGA.tgaLossless8-bitYes
BMP.bmpUncompressed8-bitNo
WebP.webpLossless or lossy8-bitYes
JPEG.jpg, .jpegLossy8-bitNo
GIF.gifLossy, 256 colors8-bitNo
DDS.ddsLossy block compression, like HAP8-bitVaries

WATCHOUT optimizes at up to 16-bit RGB or RGBA. Higher-precision sources, including 32-bit float formats such as OpenEXR, are not supported. All files in a sequence must use the same format and resolution.

The Alpha column shows which formats can carry transparency. How alpha passes through to the output codec, and when it is discarded, is covered in Formats and Codecs.

Adding an Image Sequence

  1. Right-click in the Assets window, open the New/Add submenu, and choose Add Image Sequence.
  2. A folder browser opens. Select the folder that contains the numbered image files.
  3. The Asset Manager scans the folder, detects the longest contiguous sequence, and creates a single asset.

The new asset appears with a video-style icon. Its type is Video, not Image, because it holds multiple frames.

Updating Frames

When you iterate in production, you usually re-render only a few frames. Re-export those frames into a watched folder, and the Asset Watcher updates only the changed frames. It does not re-optimize the whole sequence.

The Asset Watcher detects each changed frame file and re-optimizes only those frames. Unchanged frames are kept, and only the changed frames' data transfers to the Runners. This makes on-site render farms practical: re-render a small frame range, and the show updates without re-processing thousands of frames.

This partial update runs through the Asset Watcher. Re-importing the folder by hand re-optimizes the whole sequence.

Frame Rate and Duration

Image sequences are imported at 60 fps (frames per second), whether you add them manually or through the Asset Watcher. There is no per-folder frame-rate setting. For a sequence at another rate, such as 50 fps, set the correct rate by reinterpreting the frame rate in the Create Version dialog (a metadata-only change). Otherwise the sequence plays at the wrong speed and duration.

The duration is the frame count divided by the frame rate:

> Duration = number of frames / frame rate

A 1200-frame sequence at 30 fps is a 40-second asset. The same sequence at 60 fps is a 20-second asset.

Optimization

An image sequence runs through the same optimization pipeline as a video file (see Asset Manager). The sequence-specific part is the source codec: the frames are read in order, and their pixel format (8-bit RGB, 8-bit RGBA, 16-bit RGB, and so on) sets the source codec. All frames are optimized together into one video asset for playback on the Runners.

The default output codec depends on the source pixel format:

Source pixel formatDefault output codec
8-bit RGBRaw 8-bit RGB (pass-through)
8-bit RGBARaw 8-bit RGBA (pass-through)
16-bit RGBRaw 16-bit RGBA
16-bit RGBARaw 16-bit RGBA (pass-through)
8-bit grayscaleRaw 8-bit RGB
8-bit grayscale + alphaRaw 8-bit RGBA

The default output codec is configurable. See Formats and Codecs for the full codec mapping.

A low-resolution preview is also generated for Producer, the same as for any video asset (see Asset Manager).

Image sequences have the lowest optimization priority. They are queued after all other asset types. Plan for long optimization times with large sequences. For faster processing, encode the sequence to a video file in your compositing tool first.

Playback Behavior

Once imported and optimized, an image sequence behaves identically to a video asset:

  • Place it on the timeline by dragging from the Assets window.
  • Set in/out points, loop mode, and duration in the cue properties.
  • Apply effects and transitions as with any video cue.
  • Use as a Dynamic Asset version for content that changes between shows.

Performance Considerations

A 4K 16-bit sequence at 60 fps can reach hundreds of gigabytes of source data. Keep in mind:

  • Storage. The source images need significant disk space during import and optimization. The optimized output is usually much smaller. Chunks default to 512 MB.
  • Optimization time. Encoding thousands of frames takes longer than re-encoding one video file. Plan for longer times with large sequences.
  • Color space. Image sequences are assumed to be sRGB / Rec.709 (SDR). The optimizer handles the color transfer during encoding. For HDR, deliver HEVC 10-bit video instead (see Formats and Codecs).
  • License and demo mode. Without a license (demo mode), output is limited to 8-bit Raw codecs (RGB, RGBA, grayscale, grayscale + alpha). A 16-bit sequence (for example, 16-bit PNG or TIFF) needs a valid WATCHOUT license to optimize.

Best Practices

  • Use consistent numbering with leading zeros (for example, frame_0001.png through frame_2400.png). Detection relies on the trailing digits in filenames.
  • Verify contiguity before importing. Missing frames fail the import. Check that the render output is complete.
  • Use 16-bit PNG or TIFF for the highest fidelity. Lossless 16-bit frames convert to Raw with no compression loss and no banding.
  • Set the frame rate during import. The 60 fps default may not match the show. Reinterpret the fps in the Create Version dialog so the duration and playback speed are right.
  • Pre-encode for faster turnaround. When optimization time matters, encode to ProRes or HAP in your compositing tool and import as a video.
  • Keep source files available until optimization finishes. The optimizer reads frames from disk during processing.