IAMCCS Shotboarder Aud+Vid Exporter PRO

If you're finishing IAMCCS Shotboard or AudioBoard projects, this is the node you save your movie with - and "save" is doing a lot of work. IAMCCS_ShotboarderAudVidExporterPRO (displayed as "IAMCCS Shotboarder Aud+Vid Exporter PRO") is a full professional muxing and delivery tool: it takes your generated VIDEO plus an audio source, muxes them with real ffmpeg codecs, embeds metadata, writes a sidecar, and can push the result through Nvidia RTX VSR on the way out. The default profile is prores_422_hq_mov for a reason - this is aimed at an editorial pipeline, not a quick web export.

How it works

It's an output node (OUTPUT_NODE = True) that delegates to ffmpeg under the hood - the pack even shims a bundled imageio-ffmpeg binary onto your PATH on import so the exporter has a working ffmpeg without you installing one. The two modes that matter:

• audio_source_mode = audio_input - mux an audio AUDIO input with the video.
• audio_source_mode = master_file_direct - read the master audio straight from a file (master_audio_file) and align it to the timeline using the optional editor_audio_edl_json edit list, which is how AudioBoard master audio lands on the edited video without re-rendering.

audio_sync is trim_to_video (default - audio is trimmed to the video length) or shortest. The 1.5.x release notes also describe roll de-duplication: overlapped pre/post-roll material is emitted once so nominal duration and master-audio sync survive editorial trims.

Codec and quality knobs

• video_profile - h264/h265 MP4 for delivery, then the editorial set: prores_422, prores_422_hq, prores_4444, dnxhr_hqx, v210 (10-bit uncompressed), ffv1 MKV for archival.
• audio_profile - pcm_s16le (default), pcm_s24le, pcm_s32le, flac, alac, aac_320/192, or copy_source (no re-encode).
• frame_rate_override, video_quality (CRF-style 0–51, default 18), embed_metadata, write_sidecar.

Then the RTX block: rtx_enabled, rtx_mode (VSR quality levels like "VSR Medium"), rtx_scale, rtx_megapixels, target rtx_width/rtx_height, ratio preset, and resize method. When enabled, the exported frame sequence passes through Nvidia VSR before muxing - realtime-ish hardware upscaling, not a diffusion upscaler, so think pixels rather than invented detail. The rtx_frames IMAGE output exposes the exact frame sequence used (post-RTX), and output_path/report strings tell you what got written.

The gotcha

The rtx_* transport fields are deliberately typed as STRING even though the UI shows selects - that keeps old workflows that saved a JSON string in those slots from failing validation. It means you shouldn't hand-drive them from upstream math; leave them to the UI. And remember this muxes - it doesn't generate. Feed it a VIDEO container from the Shotboard render backend, not a bare image batch.

Install the pack via ComfyUI Manager (search "IAMCCS") or git clone https://github.com/IAMCCS/IAMCCS-nodes.git into custom_nodes. RTX features need an Nvidia RTX GPU and the RTX nodes pack; everything else needs only ffmpeg, which the pack provides.