How to Make a Photo Slideshow With Voiceover

To make a photo slideshow with voiceover, build it in the opposite order to the one most people try: write the script, make the voice track, and only then lay the photos onto it. A photo can be stretched or trimmed by a second for free. A sentence cannot, so the voice has to set the clock.

This guide walks through a narrated slideshow end to end — how much script a slideshow needs, how to prepare iPhone photos so they stay sharp on screen, whether to record your own voice or generate one, and where exactly to cut between pictures.

Why the voice comes first

The usual approach is to arrange the photos, set each to a few seconds, and then try to talk over them. It works until the narration runs a little long on one photo and a little short on the next.

At that point there are two options, both bad. You rewrite the line to fit the picture, or you record it again faster. Either way you are editing the voice to suit the pictures, which is the expensive direction.

Turn it around and the problem disappears. With the narration already on the timeline, each photo simply starts where its line starts and ends in the pause after it. Changing a photo’s length is a drag of the mouse.

How much script a slideshow needs

Narration has a natural speed, and it is the number that ties the whole slideshow together.

The 150-words-a-minute rule

A relaxed narrator reads at roughly 150 words a minute, or about two and a half words a second. Brisk delivery runs nearer 170; slow, deliberate narration closer to 130.

So a one-minute slideshow wants about 150 words of script, and a three-minute one about 450. If the video must not overrun — a wedding toast, an entry with a time limit — write to the slow end and leave yourself room.

One line per photo

Write the script as a numbered list, one line per picture. Ten words a line gives each photo about four seconds, which sits comfortably inside the three-to-five-second range most slideshows use anyway.

This also makes the edit almost mechanical later. Line 7 belongs to photo 7, and when you cut a photo you know exactly which sentence goes with it.

Write for the ear, not the page

A script is heard once, at speaking speed, with no way to reread it. Keep sentences short, put one idea in each, and say names and places in full the first time.

Read it aloud before you record anything. Every place you stumble is a place a listener will too.

Prepare the photos

The pictures need less work than the voice, but two steps decide whether they look sharp or soft on a big screen.

Convert HEIC to JPG first

Photos straight from an iPhone are HEIC, and many video editors either refuse them or import them with the colour or rotation wrong. Converting to JPG first avoids all of that.

The fuller story — which editors take HEIC, and how to keep the import order intact — is in using iPhone photos in a video or slideshow.

Leave room for movement

A slideshow with narration usually has slow zooms and pans on each photo, often called the Ken Burns effect. That movement only stays sharp if the photo is bigger than the video frame.

A 1080p video is 1920 pixels wide. Give it a photo exactly that size and any zoom has to enlarge pixels, which turns soft on a television. Resize to about 2560 pixels on the long edge instead and the frame can move or zoom to 133% without ever stretching anything.

Our HEIC resizer caps the long edge across a whole batch in one pass. The full-resolution original is still larger than you need, and makes the editor slower for no visible gain.

Number them in story order

Rename the files 01, 02, 03 in the order of the script before you import them. Most editors sort by filename, and a numbered folder means photo 7 lands next to line 7 without dragging anything around.

Record or generate the voiceover

You need one clean voiceover file covering the whole script. There are two ways to get it.

Record it yourself

Your own voice carries a family video or a personal story in a way nothing else does. A phone’s voice memo app in a quiet room with soft furnishings is genuinely good enough, held a hand’s width from your mouth and slightly to one side.

Record the whole script in one take, then re-record only the lines you stumbled on. Leave a second of silence between lines — it costs nothing and makes every cut later easier.

Generate it with an AI voice

The alternative is to type the script and let an AI voice read it. An AI voice generator such as AnyVoice turns text into narration and can clone a voice from a short recording, so the track can even sound like you without you reading every line.

The real advantage is revision. When you change a sentence, you regenerate that line in seconds rather than setting up the microphone again, and the new take matches the old ones exactly.

AI voices read punctuation as timing, so use it deliberately. A comma gives a short breath, a full stop a longer one. Write numbers and unusual names the way they sound, and generate a test line before committing to the whole script.

Only clone a voice you have the right to use

Clone your own voice, or one whose owner has explicitly agreed. Using someone else’s voice without consent breaks most services’ rules and, in many places, the law — and a slideshow is no exception because it is personal.

Export as WAV if you can

Whichever route you take, the format of the voiceover file matters more than it looks.

WAV is uncompressed, so the editor gets every detail of the recording to work with. MP3 is fine for a draft or if it is all your tool offers. The finished video compresses its audio to AAC whatever you start from, so begin with the best copy available.

Time the photos to the voiceover

This is the part the earlier steps were for, and it goes quickly.

Lay the voiceover down first

Import the narration and put it on the timeline before any picture. Then mark the pause after each line. Many editors drop a marker at the playhead with a single key; in one that does not, splitting the audio at each pause does the same job.

Cut in the pause, never mid-word

Drag each photo so it ends on its marker. The change of picture then lands in the breath between sentences, which is where the ear expects something new.

A cut in the middle of a word feels wrong even when nobody can say why. A cut in the pause feels as if the voice asked for the next picture.

Let some photos breathe

Not every photo needs words. A sequence of two or three pictures with only music between sections gives the listener a rest and makes the next line land harder.

Plan for these in the script with a blank line, so the numbering still matches.

Music under the voiceover

Music carries the mood. The voice carries the meaning, so it has to win every time they overlap.

Duck the music under the narration

Set the music well below the voice — around 15 dB lower is a common starting point — and let it rise only in the gaps. Many editors can do this automatically, lowering the music whenever the voice track has sound on it.

Then listen on a phone speaker, not headphones. Small speakers are where music swamps speech first, and it is how most people will hear the result.

Use music you are allowed to use

A slideshow posted anywhere public can be muted or blocked for using commercial music. The editor’s own library or a royalty-free source avoids that, and both are easier than finding out after the upload.

Export and check

The safe export for nearly everything is an MP4 with H.264 video and AAC audio, at 1080p and 30 frames a second. Every phone, computer, television and social platform plays it.

Watch the whole thing once at full screen before sending it. Listen for a cut that lands mid-word, a photo that sits too long after its line, or music that rises before the voice has finished — those three account for most of what feels off in a narrated slideshow.

Which app to build it in

Almost any editor with a separate audio track will do.

On a Mac or iPhone, iMovie is free and handles a voice track, music and photo movement. On Windows, Clipchamp comes with Windows 11 and does the same. CapCut runs on phones and desktops alike.

The one tool that will not work is the Memories feature in the iPhone Photos app. It can set music, but it has no way to add a narration track of your own.

The bottom line

Build a photo slideshow with voiceover around the voice, not the pictures. Write about 150 words a minute, one line per photo, then record or generate the narration as a single file and put it on the timeline first.

Convert the photos to JPG, resize them to about 2560 pixels so the movement stays sharp, and cut each one in the pause after its line. Add music last, well under the voice, and check it all on a phone speaker before anyone else hears it.

Frequently asked questions

How do I add a voiceover to a photo slideshow?+

Write a script with one line per photo, record or generate the narration as a single audio file, put it on the timeline first, then place each photo so it changes in the pause after its line. Add music last, well under the voice.

How long should each photo stay on screen with narration?+

As long as its line takes to say. At a typical 150 words a minute, a 10-word line runs about four seconds. Without narration, three to five seconds a photo is the usual range.

How many words do I need for a one-minute slideshow?+

About 150 words at a natural pace, or 130 if you want room for pauses. That carries roughly 15 to 20 photos at three to four seconds each.

Can I use an AI voice instead of recording my own?+

Yes. An AI voice generator reads your script aloud from text, so you can rewrite a line and regenerate it in seconds instead of recording again. Only clone a voice that is yours or that you have explicit permission to use.

What audio format should the narration be?+

WAV if you can, MP3 if that is all the tool exports. WAV is uncompressed and gives the editor the best copy to work from; the final video compresses the audio to AAC either way.

How loud should background music be under a voiceover?+

Well below the voice — around 15 dB lower is a common starting point. Most editors can lower the music automatically while someone is speaking. Check the result on a phone speaker, where music drowns speech most easily.

Do I need to convert iPhone photos before making a slideshow?+

Usually, yes. Many editors import HEIC badly or not at all, so convert to JPG first. Resize to about 2560 pixels on the long edge for a 1080p video so slow zooms stay sharp.

Are my photos uploaded when I convert them here?+

No. The converter on this page runs in your browser, so your photos are decoded and re-encoded on your own device and never sent anywhere.