The two LTX entries: the studio's menus and the model's prompt steps

LTX Studio lists camera motion presets under the prompt field: static, handheld, dolly, crane, dolly zoom and whip out. Its shot editor sets shot type, camera angle and framing for each shot. The open LTX model is read separately on this register: its guide makes camera movement a step of the prompt and names each edit in multi-shot prompts. As of 2026-09-25.

1Rows this register already read

Both LTX rows in full — the studio's presets and the open model's prompt guide — beside a third tool whose presets carry weights.

• Camera motion presets
  Camera motion presets under the prompt field: static, handheld, dolly, crane, dolly zoom and whip outLTX Studio blog, camera motion in the Gen Space / recorded 2026-09-25
• Precision in shot design
  Shot type, camera angle and framing are set for each shot in the shot editorLTX, Shot Video Editor platform page / recorded 2026-09-25
• How to use
  Presets are applied to a clip made from a first-frame image on the LTXV modelLTX Studio blog, camera motion in the Gen Space / recorded 2026-09-25
• Key elements
  Camera movement is a step of the prompt: how and when the camera moves, relative to the subjectLTX, prompting guide / recorded 2026-09-25
• Structuring your prompt
  Detail is matched to shot scale: close-ups need more than wide shotsLTX, prompting guide / recorded 2026-09-25
• Multi-shot prompts
  Multi-shot prompts name each edit, a hard cut, match cut or dissolveLTX, prompting guide / recorded 2026-09-25
• Advanced features
  The model supports several keyframes, not only a first frameLightricks, LTX-Video repository / recorded 2026-09-25
• Motion Controls
  Motion Control presets for Motion 2.0, from crane and dolly moves to orbits and bullet timeLeonardo.Ai Help Center, generating video / recorded 2026-09-25

2What the studio puts under the prompt field

LTX Studio's camera controls come as presets beneath the prompt: static, handheld, dolly in and out, crane up and down, dolly zoom and whip out, applied to a clip made from a first-frame image. Separately, its shot editor sets shot type, camera angle and framing for each shot in a sequence, which makes framing a property of the shot rather than of the passage of text.

A preset is a fixed move. Picking dolly zoom asks for one specific thing, and the register's reading records that list as the vocabulary the studio offers, not as a claim about how often it lands. The two points this row leaves empty — negative prompts and a motion setting — are silences from the same read.

3The open model under the same name, read on its own

The register keeps two LTX entries because the maker's two surfaces publish different things. The open model's prompting guide makes camera movement a step of the prompt: how and when the camera moves, relative to the subject. Detail is matched to shot scale, with close-ups needing more description than wide shots. Multi-shot prompts name each edit — a hard cut, a match cut or a dissolve — and the model is documented as taking several keyframes rather than a first frame alone.

Read the row above and this one together and a writer learns where each half lives: presets if the work happens in the studio, prompt steps and named edits if it happens against the open model. Folding them into one entry would lose exactly that, which is why they stay apart.

4Presets, prompt words and a third arrangement

Leonardo is on this page because its docs put a weight on the preset. Motion Control carries a value from minus two to two, one by default, so the same named move can be asked for strongly or faintly. Around that, its help lists camera words that can go in the prompt, with a warning that the effect may not be consistent.

Three arrangements now sit side by side on this page: a menu with no dial, a prompt step with named edits, and a menu with a dial next to a warning. The register keeps all three because the shape decides how a shot list is written — a preset is named, a step is described, a weighted preset is named and then tuned.

5What the counts around the names say

The material gathered around these names is lopsided in a way worth printing. A neighbouring model family is counted in thousands of posts, with the two lines quoted below, while the share table puts LTX at a small percentage of one subreddit's posts — and the collection carries no post count behind that share, so the cell says what the record holds rather than a figure that was never collected.

A small share is not a small thing; it is a smaller conversation. The register's answer to a camera question stays the same either way: the row read from the maker's docs, with the day it was read, and no line drawn from how often people talk about it.

6What the tools' own docs say

Inclusion rule. Tools this page is about, plus the pipeline this register reads for its own camera rows. A tool's cell carries the wording of the page read; the pipeline's cell is one sentence quoted from its published guides. A point a page does not carry is printed as not stated rather than filled from a showcase. Order. Alphabetical by named tool.

7What the gathered material counted

Two lines below, from a neighbouring family the collection counted in the thousands; the share table beneath them is the smaller number this page also carries.

Inclusion rule. Names carried in the gathered material and read on this page. A post arguing with itself can be counted in the positive and the negative column both. Order. Descending by posts.

Inclusion rule. Model families the collection put a share against, out of the posts it read in that subreddit. Order. Descending by the share as printed.

8When staying with the named tool is the simpler call

• The camera is a preset, not a paragraph — Picking from static, handheld, dolly, crane, dolly zoom and whip out needs no wording, which suits a shot list written by someone who is not a camera operator.
• Framing is set per shot — Shot type, camera angle and framing as shot-editor fields mean a sequence can be planned shot by shot rather than described in one passage.
• The open model documents the prompt route — Camera movement as a step of the prompt, edits named in multi-shot prompts and several keyframes are all written down for the model the studio is built on.

Searches for "ltx studio ai" land here; the studio and the open model beneath it are two rows on this register, and they do not read the same. Two more reads sit beside this one: Runway's camera work and OpenArt's director.