Save 4 hours! We auto-setup your workflow! Free!

Drop your workflow.json — we handle every dependency, custom node, and model. Just open the link and run.

Auto-Setup Workflow Json (Free) Now!
ComfyUI > Nodes > H3 Continuum > H3 Continuum Reference Audios

ComfyUI Node: H3 Continuum Reference Audios

Class Name

H3ContinuumReferenceAudios

Category
MiniMax H3/Continuum/Input
Author
ukr8b3g-cmyk (Account age: 400days)
Extension
H3 Continuum
Latest Updated
2026-11-04
Github Stars
0.31K

How to Install H3 Continuum

Install this extension via the ComfyUI Manager by searching for H3 Continuum
  • 1. Click the Manager button in the main menu
  • 2. Select Custom Nodes Manager button
  • 3. Enter H3 Continuum in the search bar
After installation, click the Restart button to restart ComfyUI. Then, manually refresh your browser to clear the cache and access the updated list of nodes.

Visit ComfyUI Online for ready-to-use ComfyUI environment

  • Free trial available
  • 16GB VRAM to 80GB VRAM GPU machines
  • 400+ preloaded models/nodes
  • Freedom to upload custom models/nodes
  • 200+ ready-to-run workflows
  • 100% private workspace with up to 200GB storage
  • Dedicated Support

Run ComfyUI Online

H3 Continuum Reference Audios Description

Bundles up to three ordered audio references for H3/Continuum sampling with VAE encoding.

H3 Continuum Reference Audios:

H3ContinuumReferenceAudios packages up to three standalone voice or sound references into one ordered reference-audio bundle for the MiniMax H3/Continuum V3.8 sampler. It preserves the exact order in which the references are provided, allowing each connected audio clip to correspond to the matching prompt tag, such as <Audio 1>, <Audio 2>, or <Audio 3>. The node is useful when you want generated audio to take inspiration from one or more existing voices, performances, or sounds while still allowing the sampler to create new final audio. Reference clips are prepared using the shared Core audio VAE, and the generated final audio is not replaced by any reference clip.

H3 Continuum Reference Audios Input Parameters:

reference_audio_1

This optional AUDIO input is the first standalone voice or sound reference. When connected, it becomes <Audio 1> for the sampler and is included as the first item in the reference-audio bundle. Its position is significant because the node preserves the input numbering for prompt-based audio references. There is no numeric minimum, maximum, or default value because this is an optional audio connection; leave it disconnected when you do not need a first reference. If you connect later reference inputs, this input must also be connected so the sequence remains continuous.

reference_audio_2

This optional AUDIO input is the second standalone voice or sound reference. When connected, it becomes <Audio 2> for the sampler and is placed after reference_audio_1 in the bundle. Use it when the generation should be able to distinguish a second reference from the first, such as when providing another voice or an additional sound characteristic. There is no numeric minimum, maximum, or default value because this is an optional audio connection. It must not be connected while reference_audio_1 is empty, because the node requires reference inputs to be connected in order without gaps.

reference_audio_3

This optional AUDIO input is the third standalone voice or sound reference. When connected, it becomes <Audio 3> for the sampler and is placed after the first and second references. It is useful when a workflow needs up to three separately identifiable audio references. There is no numeric minimum, maximum, or default value because this is an optional audio connection. It must not be connected unless reference_audio_1 and reference_audio_2 are also connected, since the node does not allow gaps in the ordered reference list.

reference_audio_vae

This optional VAE input supplies the shared Core Audio VAE used to encode every connected reference audio clip. The VAE is required whenever at least one reference audio input is connected, because it provides the audio encoding needed to prepare the references for the sampler. The same VAE is applied consistently to all connected references in the bundle. This parameter has no numeric minimum, maximum, or default value; it accepts a VAE connection rather than a number or selectable option. When no reference audio is connected, the VAE is not needed for the node to return an empty reference result.

H3 Continuum Reference Audios Output Parameters:

audio_references

This output provides a H3_CONTINUUM_AUDIO_REFERENCES bundle containing the connected reference audios in their exact input order, together with the shared encoding information required by the V3.8 sampler. The first connected clip corresponds to <Audio 1>, the second to <Audio 2>, and the third to <Audio 3>. The output is intended to be connected to a compatible H3/Continuum sampling workflow; it is not itself the generated final audio. If none of the three audio inputs is connected, the node produces no reference bundle, represented internally as None. The bundle does not replace the sampler's generated result; it only supplies optional reference material.

H3 Continuum Reference Audios Usage Tips:

  • Connect reference clips consecutively from reference_audio_1 onward. For example, when using two clips, connect reference_audio_1 and reference_audio_2 and leave reference_audio_3 empty.
  • Use separate, standalone audio clips when you want the sampler to distinguish different voices or sounds through separate <Audio N> prompt references.
  • Connect reference_audio_vae whenever at least one reference audio is connected. Use the shared Core Audio VAE intended for the surrounding H3/Continuum workflow so all references are encoded consistently.
  • Keep the input order intentional. The clip connected to reference_audio_1 is always treated as <Audio 1>; changing the order changes which clip each prompt tag refers to.
  • Use fewer references when a single clear voice or sound example is sufficient. Adding multiple clips is useful for distinct reference roles, but each clip should have a clear purpose in the prompt and workflow.
  • Leave all audio inputs disconnected when you want the sampler to run without reference-audio guidance. In that case, the node returns no reference bundle rather than fabricating or substituting an audio reference.
  • Remember that these inputs guide the generated result; they do not cause the sampler to copy a reference directly as the final audio.

H3 Continuum Reference Audios Common Errors and Solutions:

Reference Audio inputs must be connected in order without gaps so <Audio 1>, <Audio 2>, and <Audio 3> keep their exact Core numbering

  • Explanation: A later reference input is connected while an earlier input is empty, such as connecting reference_audio_2 without connecting reference_audio_1. Because input positions map directly to the sampler's <Audio N> tags, the node rejects this arrangement to prevent incorrect reference numbering.
  • Solution: Connect the audio clips consecutively starting with reference_audio_1. For two references, use inputs 1 and 2; for three references, use inputs 1, 2, and 3. Disconnect any later input that is not part of the continuous sequence.

Reference Audio VAE is required when audio is connected

  • Explanation: At least one reference audio has been provided, but no reference_audio_vae is connected. The node needs this shared Core Audio VAE to encode the reference clips before they can be used by the H3/Continuum sampler.
  • Solution: Connect a compatible Core Audio VAE to reference_audio_vae. If you do not want to use reference audio, disconnect all three reference-audio inputs instead.

No reference audio bundle is available

  • Explanation: None of reference_audio_1, reference_audio_2, or reference_audio_3 is connected, so the node returns no reference bundle. This is expected behavior when reference guidance is not needed, but a downstream workflow that requires reference audios may not have anything to process.
  • Solution: Connect at least reference_audio_1 and provide the required reference_audio_vae if you intend to use reference-audio guidance. If reference audio is optional in the downstream workflow, no correction is necessary.

Incompatible audio connection

  • Explanation: An input intended for an AUDIO connection has received a different data type or an audio object that the surrounding Core workflow cannot prepare as a standalone reference. The node interface accepts audio connections only for the three reference inputs.
  • Solution: Connect an audio output to the affected input and ensure the audio is produced by a compatible Core audio node. Use the VAE input only for the designated reference_audio_vae connection.

H3 Continuum Reference Audios Related Nodes

Go back to the extension to check out more related nodes.
H3 Continuum
RunComfy
Copyright 2025 RunComfy. All Rights Reserved.

RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.

H3 Continuum Reference Audios