Lesson 23 of 24
Sound effects for video: where to put them and how to find them
Sound effects for video, placed well: whooshes on movement, hits on text, room under it all, found by describing them and credited for you.
hopeengineer · Founder of Sharpy and Threadify18 October 2026 · 7 min read
The short answer
Sound effects for video do one job: they make what moves on screen feel physical. Put a whoosh where something moves fast, a hit where something lands, a soft room or street underneath everything, and leave the rest silent. In Sharpy you find a sound by describing it, place it so its loudest moment lands on the action, and let the sound pass give every graphic and transition its own recorded sound without crowding the voice. It all runs on your Mac, and any credit a sound needs goes into your video's description.
What does sound design add to a video?
Weight, and a sense of place. A title that slides in silently looks like a slide show; the same title with a short whoosh feels as if it moved through air. Sound design for video is mostly three layers:
| Layer | What it is | Where it goes |
|---|---|---|
| Movement | Whooshes, swipes, risers | Anything crossing the frame fast: a transition, a graphic sliding in |
| Accents | Impacts, pops, clicks, plucks, thuds | The moment something lands or appears: a word popping on, a box arriving |
| Place | Ambience, room, foley such as footsteps or a door | Underneath a whole scene, quietly, so it never feels like a vacuum |
The fourth ingredient is restraint. Every effect competes with the voice, so the best sound design for YouTube is mostly the sounds you leave out.
Where should sound effects go?
On the action, and only where the action is big enough to deserve one:
- Whooshes on transitions and big moves. A whoosh belongs to motion across the frame. A graphic that barely moves gets none.
- Hits on text and arrivals. A pop or click as a word appears; a heavier impact when a title lands.
- Risers into a reveal. A riser builds tension and stops dead at its peak, on the cut.
- Ambience under a scene. A room, a street, wind: a bed you feel rather than hear.
- Nothing on captions. Words already heard in the recording belong to the voice, not to a sound effect.
Keep them under the dialogue. Sharpy never puts more than five sounds in any five seconds, or more than two of one kind, and outside a fast montage never more than one accent every four seconds. Every event it leaves silent says why.
Why should a whoosh peak on the action, not start on it?
Because you hear a whoosh at its loudest moment, not its first. A whoosh swells before its peak: a professional one swells over a twentieth to a sixth of a second and then dies away, and its peak sits a third to a half of the way in. Start the file on the cut and the peak arrives late, so the sound trails the picture.
Sharpy places a whoosh, riser or impact so its loudest moment falls exactly on the instant you choose: a whoosh landed at 1.5 seconds is loudest at 1.5000 seconds. It reads a graphic's movement from what was drawn, where it moves fastest for a whoosh and where it lands for an impact, so a box eased in over half a second lands at exactly 0.500 seconds. An impact is allowed to be 15 milliseconds early to 30 late and never a frame early, because, as the broadcast recommendation ITU-R BT.1359 also finds, sound that arrives before its picture is noticed sooner than sound that arrives after. How loud the whoosh sits against the impact, and whether it builds into the landing, is left to you.
How do you add sound effects to a video in Sharpy?
By hand from the window, by asking the agent, or with the sound pass.
- Find the sound. Browse the sound effects shelf in the window's left panel, or ask the agent to find one by describing it (
find_sound): "a soft paper whoosh", "a wooden knock", or "more like this one" with a sound you already have. - Place it on the action. Drag it to the timeline, or ask the agent to land it (
edit_clipwithlands), giving the instant of the action rather than the start of the file. - Or let the sound pass do it. Ask for
place_soundswithshowfirst: it describes the whole plan and changes nothing, so you hear what would be added before it is. Then run it, and every graphic appearing, word popping in or transition crossing the frame gets a recorded sound for its role, at a sensible level, without crowding. The whole pass is one undoable step. - Check the credits. Each shelf item carries its credit line, written into the edit's record when you place it.
The sound pass knows sixteen roles, from move, swipe and arrival to typing, money, success, ambience and foley. A recording always beats a synthesised sound, the original file beats a preview, the length nearest the movement wins, and a role never repeats the sound it used last.
Where do YouTubers get their sound effects?
Mostly from free libraries, YouTube's own Audio Library and paid subscription libraries. The YouTube Audio Library in YouTube Studio has a Sound effects tab, and YouTube says its copyright-safe downloads are not claimed through Content ID. These are the free sound effects libraries Sharpy searches, with the licence it holds each to:
| Library | Licence in plain words | What Sharpy does |
|---|---|---|
| Freesound | Each sound is CC0 (no conditions), CC BY (credit the maker) or CC BY-NC (no commercial use) | Refuses non-commercial, Sampling+ and share-alike sounds; full-quality originals come in through your own login |
| Mixkit sound effects | Free under Mixkit's own sound effects licence | Fetches the original WAV, not the preview |
| Pixabay sound effects | Free, commercial use allowed, no credit required | Gives you the search link, because Pixabay forbids bots, and imports what you download |
| Wikimedia Commons | Each file carries its own licence | Refuses CC BY-SA, NC, ND and GFDL files |
| OpenGameArt | Several licences, chosen by each uploader | Takes CC0 only |
Each item's licence is read off the item itself when it is fetched, never off the search listing. Royalty-free music and sound effects goes through the same libraries for music.
What is on Sharpy's own sound shelf?
Three kinds of sound, all checked before they are offered:
- 239 recordings curated by role, each licence read off its own page, with alternate takes kept so a sound is not repeated.
- 32 designed sounds in eleven families, at 48 kHz and 24-bit: whooshes at 0.3, 0.6, 1.2 and 2.5 seconds, each in two takes crossing the stereo field opposite ways; risers in three lengths; downlifters; sub drops falling to 35 Hz with harmonics a phone speaker can play; braams; swells; stingers, glitches and shimmer; five interface sounds; and a tape stop.
- Fourteen accents in five families: impacts, thuds, pops, clicks and plucks, each ending in silence rather than stopping while it still sounds.
Every sound is measured for four kinds of damage: clipping, coarse digital steps, a tail cut off while it still sounds, and the dull top end of a lossy file. It is then judged against lossless professional recordings, with a person's ears last. The same recording is never offered twice, and ambience beds such as air and rumble are made to the length you place them for, without repeating.
Search works by meaning, not by tags someone typed: a sound described in words came up first 92.3% of the time. Very short sounds rank less well, so for a tiny click, audition the top few.
What should you check before you export?
- Listen on a phone speaker. Effects that sound fine on headphones can bury a word on a phone.
- Mute the effects once. If you miss nothing, there were too many.
- Check the voice is never covered. Effects sit under dialogue; audio mixing for video shows how to group them on an effects bus with one fader.
- Paste the credits. Every render writes the credits it owes into a text file beside the video, for the description. Sharpy never puts a credit on the picture.
Captions for deaf viewers describe only the effects that carry meaning: of Sharpy's own effects, a record scratch, a cash register and a chime. Whooshes are left out, as a captioner leaves out texture.
Frequently asked questions
01How do you add sound effects to a video?
Find a sound that suits the action, put its loudest moment on the frame where the action happens, and set it under the voice. In Sharpy, describe the sound to find it and ask the agent to land it on the instant you name.
02Are free sound effects safe to use on YouTube?
Yes, if the licence allows commercial use and you give any credit it asks for. CC0, Mixkit and Pixabay sounds owe no credit; CC BY sounds need the maker named in the description, which Sharpy writes for you.
03How many sound effects is too many?
When you notice them. A useful limit is five sounds in any five seconds and no more than two of the same kind, with long stretches of nothing but voice and room.
04Can I use my own sound library?
Yes. Search it by description in the same way, and Sharpy treats a file you bring in as yours, used under your own rights.
Sources
Checked on 18 October 2026.
Written by
hopeengineer
Software architect and founder of Sharpy and Threadify. Builds scalable systems and AI tools, and writes here about how Sharpy understands and edits video.
Sharpy opens on 1 November 2026.
The world's first context-aware agentic video editor, on your Mac. One plan, unlimited edits.