LTX Studio: voices on Elements, from nowhere stated
Voices are attached to character Elements, which is the durable arrangement and the one that survives a long shoot. The page stops there: no catalogue, no sample window, and no statement that a recording is accepted at all. As of 2026-09-12.
| What the documentation settles | What it leaves to a take |
|---|---|
| Voices are attached to character Elements | Where a voice comes from before it is attached |
| Joint audio and video generation is documented on LTX-2.5 | Whether the joint route can be steered towards a chosen voice |
| An audio-to-video route runs in the other direction | Whether that route is how a production supplies a voice in practice |
Inclusion rule. Only statements found on the vendor page named in the caption, and only ones bearing on this one field. Nothing here is filled in from generated output. Order. In the order the statements arrive on the vendor's page.
1The audio-to-video route is the unstated answer
A platform that documents an existing track driving a generation has a plausible place for a supplied voice, and it is not this field. Nothing connects the two on the page, so the register keeps them apart: one is a documented route for audio, the other is a blank about where a character's voice originates.
Joining them would be a reasonable guess and still a guess. A production that needs a particular voice can test whether the reverse route accepts a dialogue recording, and that test is a finding about one attempt rather than a published behaviour.
2Element systems make the blank easier to live with
Where a voice belongs to a character object, a production only has to be lucky once. Whatever the first generation produces becomes the character, and the system keeps it. That is a very different failure mode from re-establishing a voice on every call and hoping it matches.
It is also the argument for recording, in a production's own notes, whatever created each Element. The platform holds the voice; nothing published says how to reconstruct it if the Element is lost.
3The others that publish nothing a production could hand over
Six more entries reach this column empty. Two of them also bind a voice to something durable, which makes the blank a gap in selection rather than in persistence.
- Kling AI — nothing documented as an input.
- Luma Ray — nothing documented as an input.
- Sora 2 — nothing documented as an input.
- sync-3 — nothing documented as an input.
- Veo — not documented by the vendor.
- Vidu — nothing documented as an input.
- Per-character bindingVoices are attached to character Elementstied to the element system
- Audio sourceJoint audio and video generation, plus audio-to-video, on LTX-2.5two directions
4Sources
Read from the product page at ltx.io on 2026-09-12. The same column across every entry is on voice source; everything this vendor publishes about speech is on LTX Studio. What counts as documented is on how read.