Skip to main content

How to Create a Karaoke Version of a Song

Turn a song you’re allowed to use into an instrumental WAV you can sing over. You’ll remove the vocals, check the chorus and the ending, then download. On-screen lyrics are a separate job.

What you get here: an instrumental WAV file. Not a lyrics video or a CDG.

Start extracting an instrumentalTrim your instrumental

Example: Turn a song into singing practice

A song leads to an accompaniment speaker and a person practicing singing, with a lyric sheet separate from the audio workflow.
Workflow illustration, not a song-quality test.

Make the backing track first. Add displayed lyrics with another tool.

  1. Song
  2. Backing track
  3. Singing practice

Choose accompaniment audio or a lyrics video

Audio first; lyrics separately: Song you may edit; Remove vocals; Instrumental WAV; Lyrics video / CDG: another tool
Audio first; lyrics separatelyAudio workflow only. Lyrics timing and video/CDG export are separate.

“Karaoke version” can mean three different things, and it helps to know which one you need before you start.

  • Backing track: the music without the lead vocal, as an audio file. You sing from memory or read the words from paper or another screen. This is what this guide makes.
  • Lyrics video: a backing track with the words shown on screen in time with the music, usually as a video file.
  • CDG: short for CD+Graphics, a format that pairs audio with simple on-screen lyric graphics, used by some karaoke players.

The figure shows a song passing through vocal removal to an Instrumental WAV, with a separate reminder about lyrics. Removing the vocals gives you an instrumental WAV. Getting lyrics on screen is a separate route: someone has to time the words to the music and then export a video or a CDG. StemNook covers the first route only.

If your singer needs the words on screen, the backing track is just the first step. And if you already have a backing track, you don’t need to remove anything: your next job is timing the lyrics.

Prepare an MP3 or WAV you may edit

A person chooses a short music file from their recording folder and checks its length before processing.
Prepare the audio firstChoose a permitted MP3 or WAV within the limits.

A few minutes of preparation saves a wasted attempt.

  • Rights: use a song you own or have permission to edit, such as your own recording or a licensed track.
  • Format and limits: MP3 or WAV, from 0.5 seconds to 5 minutes long and up to 30 MB. A WAV at 44.1 kHz, 16-bit stereo passes 30 MB at just under three minutes, so for a longer song, use an MP3 copy.
  • Length: for a song over 5 minutes, trim the part you want in an audio editor first.
  • Version: a clean studio version usually gives the separation an easier job than a live recording with crowd noise and room echo.
  • Your allowance: separation requires Google sign-in. Free accounts receive 600 seconds each UTC calendar month. Audio duration is rounded up to whole seconds when processing time is reserved; an exactly four-minute file reserves 240 seconds.

How to make a karaoke track, step by step

Five steps to a backing track: Google sign-in; Choose MP3 / WAV; Wait for two tracks; Check chorus and ending; Download Instrumental WAV
Five steps to a backing trackIllustrates the optional sign-in-first route. You can also choose a file first and sign in at Start. This is a workflow illustration, not a measured result.

These steps use StemNook’s AI separation, which runs in the cloud and returns two tracks: Vocals and Instrumental. For karaoke, the Instrumental is the one you keep.

  1. Choose your MP3 or WAV. Pick a file you have permission to edit that fits the limits above.
  2. Click Start separation. Sign in with Google if needed, then return here and start when you are ready.
  3. Start the separation and wait. Wait until both tracks appear.
  4. Listen to the chorus and the ending. Play those parts of the Instrumental before you download it. The next section shows what to listen for.
  5. Download the Instrumental WAV. Save it within 24 hours on a free account of the separation finishing.

Check the chorus and the ending

Check the chorus and ending: Chorus: sing along, hear overlap; Final 20 seconds: notes and echo; Compare the same original section
Check the chorus and endingUse these checks with your own permitted audio.

The chorus and the ending are where leftover vocals most often show up. Choruses tend to stack the lead with harmonies, and endings often hold ad-libs, final notes and echo.

A karaoke listening check

  1. Sing along to the chorus. Play the Instrumental’s chorus and sing with it. If you hear a second voice competing with yours, some of the vocal stayed in.
  2. Play the last 20 seconds. Listen for a held note, an ad-lib or an echo after the music should be bare.
  3. Compare with the original. Play the same moments of your original in your usual player and notice what was removed.
  4. Decide what’s good enough. A faint trace may be fine for singing at home. For a performance, try another version of the song if you have one.

For a clip shorter than 20 seconds, just listen from beginning to end. The figure for this step marks the two places to check on a song’s timeline: the chorus and the final seconds.

Download the karaoke track and trim it if needed

A backing-track file is saved to a local folder beside a clock marking the available access window.
Save the backing track promptlyDownload it during the 24-hour access window.

Download the Instrumental as soon as you’re happy with it. Free-account results are available here for 24 hours after separation finishes. If access has expired, first check the copy you already downloaded. Starting a new separation may use your monthly allowance.

The download is a WAV file at 44.1 kHz, 16-bit, stereo, which takes about 10.6 MB per minute, so a four-minute song comes to about 42 MB. Check your storage before downloading on a phone. There’s no MP3 download, so if your karaoke app needs MP3, convert the WAV with another tool.

Trimming the instrumental

Trimming is a separate step, not part of the separation. To shorten the track or cut a long intro, open the downloaded WAV in an audio editor. StemNook’s audio cutter can trim in your browser, but it accepts files up to 30 MB, and an instrumental WAV like this one passes that at just under three minutes. For longer instrumentals, use another audio editor.

Using the track

Play the WAV through speakers or headphones at a level where you can still hear yourself. While you’re learning the song, the Vocals track from the same separation can help: listen to it on its own for phrasing, then switch back to the Instrumental to sing. Keep both files together so you can switch easily.

Lyrics, CDG and public performance limits

An audio backing track and a separate lyric-video workspace are shown as two distinct tasks.
Lyrics are another stepUse another tool for words on screen.

Adding lyrics

StemNook doesn’t write, time or display lyrics, and it doesn’t make videos or CDG files. To get words on screen, you need lyric text you’re allowed to use and a tool for timing each line to the music, then an export as a video or a CDG. Plan that as its own task after you have the backing track.

A home singing practice scene is contrasted with an empty public stage to prompt checking the rights for the intended use.
Check rights for the settingHome practice and public sharing need different checks.

Personal, authorized and public use

Singing along at home to a song you own is different from sharing a karaoke track online or playing it at a public event. Removing the vocals doesn’t change who owns the song or give you permission to publish or perform it. Venues and platforms may have their own rules and licenses. Check the rules that apply where you are. This isn’t legal advice.

Karaoke creation questions

Why doesn’t my karaoke track have lyrics?

Because removing vocals only changes the audio. The Instrumental WAV is music without the singing. Lyrics on screen need separate timing work and a video or CDG export, which StemNook doesn’t do.

What’s the difference between a WAV backing track, a lyrics video and a CDG?

A WAV backing track is audio only. A lyrics video adds timed words on screen as a video file. A CDG pairs audio with simple lyric graphics for karaoke players that support it.

Why can I still hear vocals in the chorus?

Stacked harmonies and dense arrangements are harder to separate, so traces of the voice can stay in the Instrumental. Try another version of the song, or accept faint traces for home use.

Why is there a voice at the end of the track?

Endings often hold ad-libs, long notes and echo, and those can survive separation. Play the last 20 seconds before you download, and trim the end in an audio editor if needed.

When should I save my karaoke track?

Right after you’ve checked it. Free-account results are available for 24 hours after the separation finishes.

Can I trim the instrumental with StemNook?

As a separate step, if the file is small enough. The audio cutter accepts files up to 30 MB, which an instrumental WAV like this one passes at just under three minutes. For longer tracks, use another audio editor.

I already have the instrumental. Do I need to remove vocals?

If your existing backing track already suits your singing, you can skip vocal removal. If you want words on screen, lyric timing is a separate task with a different tool.

Can I change the key or tempo to suit my voice?

Not with StemNook. It separates the vocals from the music but doesn’t change pitch or speed. If a song sits too high or too low for you, you’ll need another tool for that, and key changes can add artifacts of their own.

Can I make a karaoke version of a song longer than 5 minutes?

Not in one go. Separation accepts files up to 5 minutes, so trim the part you want in an audio editor first and upload that.