Speed, Zooms & Transitions
The motion toolkit — per-range speed changes, aimable zooms, automatic punch-ins on emphasis, beat-aligned cuts, and one-request transitions.
Published 2026-07-24 · Updated 2026-07-24
Speed Control: 0.25x to 4x
Any range of your video can play at any speed from 0.25x to 4x. Speed changes are per-range spans, not all-or-nothing: say "slow the replay from 0:42 to 0:48 to half speed" and only that stretch slows down. Voices keep their natural pitch at every speed — sped-up speech sounds faster, not chipmunk-higher — and music, captions, and effects stay in sync with the new timing.
Speed is a pacing tool, not the only one. For talking footage, removing dead air usually tightens a video more than a blanket speed-up — see Cuts & Cleanup — and the two combine well: cut the rambling, then speed the B-roll.
Slow Motion, Honestly
Slowing footage works by duplicating frames — Valmera does not synthesize new in-between frames with AI. In practice that means slowdowns to around 0.6x look natural, while anything below 0.6x visibly steps: the motion gets a stuttering, frame-held quality. If you want a dramatic slow-mo moment, a moderate slowdown reads best; the agent will not pretend a 0.25x stretch is buttery when it is not.
Zooms & Targeted Zooms
Four zoom styles, each aimable at any point of the frame — not just the center — with punches up to 2.5x:
Punch
An instant jump-in for emphasis — the short-form staple. Up to 2.5x.
Ease
A gentle, gradual move that adds life to static shots without drawing attention.
Push-in
A steady move toward the subject — punchlines, reveals, key statements.
Pull-out
A move away from the subject to release tension or close a section.
Aiming matters: "zoom in on the product on the table, not my face" is a valid instruction. The agent looks at the actual frames before and after each edit, so the zoom lands where you pointed it.
Auto Punch-Ins on Emphasized Words
One request — "add punch-ins where I emphasize things" — and the agent analyzes the stress in your delivery, finds the words you actually lean on, and places zoom punches there. It is the difference between zooms on a timer and zooms that follow your performance. Works best on talking footage with real vocal variation; you can always add or remove individual punch-ins afterwards.
Beat-Aligned Cuts
For montages and music-driven edits, ask for cuts on the beat and the agent snaps cut points to the musical tempo of your track — the edit starts moving with the music instead of against it. One honest behavior to know: if no reliable tempo can be detected, the agent declines and says so. It will not invent a beat grid for ambient audio and pretend the cuts are synced.
Transitions: 7 Styles, One Request
Seven transition styles: dip to black, dip to white, whip left, whip right, zoom punch, glitch, and flash. You pick one style and it applies at every cut — "use the glitch transition" styles the whole video in one request, which is exactly what keeps a fast edit coherent.
Fades: separate from transitions, the video can fade in from black at the start and fade out at the end — a soft open and close on top of whichever cut style you chose.
Example Requests
"Slow the replay from 0:42 to 0:48 to half speed"
A per-range span — only those six seconds slow to 0.5x, pitch preserved.
"Speed up the unboxing section 2x but keep my voice natural"
The pitch preservation is the default; captions and music stay in sync.
"Add punch-ins on my most emphasized words"
Audio-stress detection places zoom punches where your delivery actually peaks.
"Cut this montage on the beat of the music"
Cut points snap to the tempo of the track — or the agent tells you no reliable beat was found.
"Use whip transitions at every cut, and fade in from black at the start"
One global transition style plus an opening fade, in a single request.
Honest Limits
- Slow motion duplicates frames. No AI frame synthesis — below 0.6x the motion visibly steps.
- One global transition style. You cannot assign a different transition to each cut.
- No crossfades or dissolves. Two clips cannot blend into each other; dip to black or dip to white is the closest equivalent.
- Beat-sync needs a real beat. If the track has no reliable tempo, the agent declines instead of guessing.
Frequently Asked Questions
Can Valmera slow down or speed up just part of a video?
Yes. Speed changes apply to any range you describe — the whole video or a single moment — anywhere from 0.25x to 4x. The rest of the video plays untouched, and music, captions, and effects stay in sync with the new timing.
Does changing speed make voices sound weird?
No. Voices keep their natural pitch when you speed up or slow down — a 1.5x section sounds faster, not higher. That applies across the whole 0.25x–4x range.
How smooth is slow motion?
Honest answer: Valmera slows footage by duplicating frames — it does not synthesize new in-between frames with AI. Down to around 0.6x this looks natural; below that the motion visibly steps. For dramatic slow-mo moments, moderate slowdowns read best.
Can Valmera automatically zoom in on important moments?
Yes. Ask for punch-ins on your most emphasized words and the agent detects the audio stress in your delivery and places zoom punches there. You can also aim a zoom manually at any point of the frame — punch, ease, push-in, or pull-out, up to 2.5x.
What transitions does Valmera support?
Seven styles: dip to black, dip to white, whip left, whip right, zoom punch, glitch, and flash. One global style applies at every cut — you cannot mix styles per cut, and there are no crossfade dissolves. The agent will say so rather than fake one.
Can cuts land on the beat of my music?
Yes. Ask for beat-aligned cuts and the agent snaps cut points to the musical beat of your track. If no reliable tempo can be detected in the audio, it declines honestly and tells you — it will not guess at a beat that is not there.