Wav Concat 🐑

Audio's version of StringConcat: take two clips, stitch them end to end, get one clip back. Where it earns its keep over just editing the waveform in an external tool is the set of small, genuinely useful controls it adds on top of the basic join - muting either side without rewiring the graph, and inserting silence at three specific points instead of one.

Where this fits in the pack. MieNodes ships TTS connectors (Bailian TTS is the one covered elsewhere in this batch) for generating speech from text. A single TTS call gives you one clip; the moment you want two lines spoken with a pause between them - a question and an answer, a title card followed by narration, two languages of the same line back to back - you need something to join them, and that's this node. It's also just generally useful audio glue any time you've got two AUDIO outputs in a graph that need to become one file.

Inputs that matter. audio1 is required - the base clip. audio2 is optional; if you leave it disconnected you're essentially using this node as a pass-through with spacer/mute controls on a single clip rather than a real concatenation. mute_audio1 and mute_audio2 (both default false) let you silence either side while keeping its timing - useful for A/B testing which half of a pairing actually needs to be there, or for building a template graph where you'll flip one side on or off per run without rewiring anything.

The three spacer inputs are where this node earns the "more than just glue" label: start_spacer adds silence before audio1 begins, middle_spacer adds a gap between the two clips, and end_spacer adds silence after audio2 finishes. All three default to 0, range 0–5 seconds in 0.1-second steps - enough resolution for a natural pause between two lines of dialogue without needing frame-level audio editing.

Output. concatenated_audio - a single AUDIO output, the two source clips (or one, with padding, if you left audio2 empty) joined in order with your spacers applied.

Install. Search "ComfyUI_MieNodes" in ComfyUI Manager, or cd ComfyUI/custom_nodes && git clone https://github.com/MieMieeeee/ComfyUI-MieNodes and restart ComfyUI. No model downloads for this node - it's waveform manipulation, not generation, so the only real dependency is whatever audio library the pack itself pulls in for reading and writing AUDIO data.

Troubleshooting. If the joined clip sounds wrong - a pop or a hard cut where the two clips meet - that's usually not this node's fault so much as a mismatch upstream: check both clips share the same sample rate before concatenating, since most audio nodes assume matching formats and won't resample for you silently. If you set a middle_spacer and don't hear any gap, double-check audio2 is actually connected - with nothing on that input there's no "middle" to space out, only a start and end around a single clip. And if muting a clip doesn't seem to shorten the overall output, that's expected: mute_audio1/mute_audio2 silence the audio content but preserve its duration, which is exactly the point if you're trying to keep downstream timing consistent while toggling content on and off - if you actually want the clip removed entirely, disconnect it instead of muting it.