Convert sound to video

"Convert sound to video" means four different jobs, and most of the pages ranking for it only do one, badly. Several of the big converters that come up for this search change the file container and hand back an MP4 with no picture at all, which YouTube still rejects. Pick your job first:

The preview under the file is not a mock-up. The same drawing code that renders the export draws the preview from a live analysis of the audio, at half resolution, so pressing play shows you the exact video, timing and colours you will download. Change anything while it plays.

Speed. A three minute song at 1080×1920 takes 15 to 40 seconds on a recent laptop and a minute or two on a phone, because the frames are handed to the hardware H.264 encoder your device already has. An hour of audio in the Still style is a few minutes. Keep the tab in the foreground; browsers throttle background tabs.

Try to upload an MP3 or WAV and YouTube Studio stops you with "The file you uploaded contains only audio and no video." YouTube only ingests files with a video track, and the same is true of Instagram, TikTok, X, LinkedIn and every course and podcast-hosting platform that accepts video. The fix is not a different audio format; it is wrapping the audio in a video container with a picture in it, which is exactly what this converter does.

Settings that work for YouTube

• Frame: 16:9 (1920×1080). YouTube shows a vertical video letterboxed on desktop, and a square one with bars.
• Style: Still for an album track, audiobook or sermon (smallest file, encodes fastest). Ring or Pulse with the cover art when you want the video to look alive in the player.
• Cover art: the album or show art, at least 1000×1000. Anything smaller is upscaled and looks soft on a TV.
• Title on the video: the track and artist. It is what viewers see when the video is embedded somewhere without a title bar.
• Quality: Full. YouTube re-encodes everything anyway, and starting from 1080p is what unlocks the 1080p option in the player.

Whole albums and long episodes

One video per track is the norm for music because each track gets its own title, thumbnail and search result. If you want a single video for the album, convert the mixed-down album file in the Still style and paste chapter timestamps into the description (a line per track, "0:00 Track name"); YouTube turns them into chapters automatically once there are at least three, starting at 0:00.

For a podcast that already has an RSS feed, YouTube Studio can import the feed under Create, then "Submit RSS feed"; it builds a static video from your show art for every episode, including future ones. It is the least effort for a back catalogue, at the cost of every episode looking identical. Use this converter for the episodes and clips you want to look like something.

Shorts

A Short is nothing more than a vertical video under three minutes. Render the same audio at 9:16, or cut a clip of the best minute, and upload it as normal; YouTube decides it is a Short from the shape and length.

An audiogram is the standard way to promote a podcast on feeds that autoplay video: a 30 to 90 second moment from the episode, the show art, a moving waveform so the frame is not static, and the quote as a title so someone scrolling with the sound off knows what they are missing. Here is the recipe, and the parts of it this page cannot do.

1. Pick the moment before you open anything. The clip should start on a complete sentence that works without context and end on the punchline. Listen back at 1.5x and note the timestamps; the trim handles here take a start and end in seconds.
2. Cut a clip with the handles under the preview. The waveform strip shows where the speech is, so you can land the cut in a pause.
3. 9:16 for Reels, TikTok and Shorts; 1:1 for LinkedIn and the Instagram grid. Render both, it is a few seconds each.
4. Bars or Wave, the show art as cover, the quote as the title (keep it under twelve words), the show name as the second line.
5. Keep the title in the middle of the frame. The layout already does: Reels and TikTok cover roughly the top 14% and bottom 35% of a 9:16 video with their interface, and the safe zone checker shows exactly where.

Captions are the part that matters most

Most feed video plays muted, so a podcast clip without captions is a picture of a waveform. This tool does not transcribe; it draws. Two honest options: the Faceless Audiogram wizard takes the same audio clip, transcribes it, burns word-timed captions in any of the caption themes and posts it to your channels on a schedule (20 credits a clip). Or drop the MP4 you just made into CapCut, which auto-captions for free, and export from there.

A song needs three videos: a 16:9 for the YouTube channel, a 9:16 teaser of the hook for Reels, TikTok and Shorts, and a Spotify Canvas. The first two come out of this page; the third is a silent loop and has its own maker.

• Ring around the cover art is the look most channels use for full tracks: the art stays the subject and the bars breathe with the mix. Pulse is the quieter version for ambient and lo-fi.
• Neon on black with your own accent colour matches most electronic artwork; Paper is for folk and acoustic covers; the two colour pickers take a hex straight from the artwork if you know it.
• Title as "Song title" and the second line as "Artist · Album (2026)". It is what shows in an embed.
• The teaser: cut the clip to the chorus (the first drop, the first line of the hook), 9:16, Bars, and put the lyric of that line as the title. Under 30 seconds performs better than a minute on every platform this year.

If you want the visuals to be a film rather than a visualizer, that is a generator's job: the AI music video path writes a shot list from the lyrics and renders animated scenes cut to the track.

Most converters cap at five or ten minutes because your file sits on their server while it renders. This one has no server, so the cap is your device: up to four hours in one go. The audio is streamed through the decoder a chunk at a time and never held in memory as a whole; what does accumulate is the finished MP4, which the tab keeps until you download it. The size line under the settings tells you what to expect before you press Convert.

• Still style is the right one for anything over 25 minutes: a static frame with a progress bar compresses to a fraction of an animated one, and it encodes in a few minutes rather than twenty.
• Fast quality (720p) halves the size again. For a lecture or an audiobook nobody is looking at the pixels.
• iPhone Voice Memos exports M4A (AAC). Those packets are copied through untouched, so the conversion is as fast as the drawing and the quality is exactly what you recorded. Share the memo to Files first, then pick it here.
• WAV and FLAC are fine as input and are encoded to AAC on the way out, so a 600 MB WAV produces a normal-sized video.
• On a phone, keep long files to Still or Fast; the memory a phone gives a tab is a fraction of a laptop's.

An MP4 with H.264 video and AAC audio, which is the one combination every platform's upload guidelines list first. Video is 30 fps, yuv420p, with a keyframe every two seconds so scrubbing is snappy, and the file is written fast-start so it begins playing before it has fully downloaded. Audio is 160 kbps AAC, or a lossless copy of the source packets when the source is already AAC. Full quality is 1080p on the long side (1080×1920, 1080×1080, 1920×1080, 1080×1350); Fast is 720p.

Sizes are ceilings computed from the encoder's bitrate targets; a still frame usually lands well under them because H.264 skips unchanged blocks. The page refuses to start a render it estimates over 900 MB because a browser tab cannot hold it, and suggests Still, Fast or a clip instead.

If you convert on a schedule, on a server, or forty tracks at once, ffmpeg is the tool and it is free. These three commands are the three jobs this page does. Install it with brew install ffmpeg on a Mac, winget install ffmpeg on Windows, or your distribution's package manager.

Honest comparison, because this page is not the right answer for everything. "Free" below means the export you get is free and usable; several hosted tools are free to try and charge to remove a watermark or a monthly minute cap, and those limits change often enough that you should check them on the day.

Two traps worth naming. A "video converter" that lists audio formats on the input side and MP4 on the output side is usually a container converter: it produces an MP4 with an audio track and no video track, and YouTube rejects it with the same message as the MP3. And any tool whose free tier says "watermark-free" only for static output will stamp the animated one; read the export screen before you spend ten minutes styling.

Where the file is going and what that platform accepts, so you pick the frame once. Checked September 2026; platforms move these.