How to Prepare Your Song Before Generating a Music Video
Export the finished, full-length master as MP3 or WAV (M4A, AAC, OGG and FLAC also work), under 50 MB, straight out of your DAW or distributor. Never upload a re-encode of a re-encode, and never a recording of a speaker. The Melodious director listens to that file first, reading the tempo, the sections, the vocal and the lyric lines when there are some, then plans the video against what it hears, so the file is the plan. Three things move the result most: contrast between quiet and loud passages, an audible vocal, and uploading the version you will actually release. Then know which part of the song you want a video of, because the director's first question is which part to make, from a short part up to the full song.
Plans from $29/month, cancel anytime. Starter makes a 30-second video with lip sync a month.
What does preparing your song actually involve?
The short answer: Export the finished, full-length master as MP3 or WAV (M4A, AAC, OGG and FLAC also work), under 50 MB, straight out of your DAW or distributor. Never upload a re-encode of a re-encode, and never a recording of a speaker. The Melodious director listens to that file first, reading the tempo, the sections, the vocal and the lyric lines when there are some, then plans the video against what it hears, so the file is the plan. Three things move the result most: contrast between quiet and loud passages, an audible vocal, and uploading the version you will actually release. Then know which part of the song you want a video of, because the director's first question is which part to make, from a short part up to the full song.
Most guides to making a music video start after the song is already uploaded. That skips the step where first attempts most often go quietly wrong.
The whole process of turning your song into a music video hangs off one input: the audio file. Everything downstream, meaning what the three ideas are about, where the shots change and what the plan thinks your chorus is, is built from what the director hears in it. The rest of the pipeline is covered in how to make an AI music video; this post is only about the file you hand it. Give it two minutes of care and you skip a redo.
Why the file matters
In the studio you add a song and a sentence about your idea, and the director listens before it asks you anything. Everything after that is built on what it heard: the questions, the three ideas, the plan and the storyboard.
Two consequences that matter for how you prepare:
- The audio is doing work from the first minute. The listen carries a chunk of the plan, so the quality of what it has to hear matters.
- Nothing expensive runs without you. The ideas, the plan and your messages to the director use a few credits each as you go. The storyboard and the video are separate paid steps, each a card with its exact price that you tap to approve.
So a sloppy file is not just a worse read, it is a round you want to do again. Prep is cheap insurance.
Export the right file: format, length, and what quietly degrades it
Melodious accepts MP3, WAV, FLAC, M4A, AAC and OGG, up to 50 MB. Within that list, the format is close to a non-issue. A good MP3 and a WAV both give the director plenty to work with, and agonizing over which one is a way of avoiding the decisions that actually matter.
What does matter:
- Export the finished, full-length track. Not a rough bounce, not a trimmed preview, not a two-minute radio edit unless that is the release. A partial file is read as though it were the whole song, so you get a plan for a structure your record does not have.
- Export the audio, not a recording of the audio. A screen capture of a player, a phone pointed at a monitor, a voice memo of a rehearsal: all of these add room noise and clipping, and often lose the first half-second. The low end takes the worst of it.
- Do not re-encode a lossy file. MP3, AAC and OGG all discard information to save space. Decoding an MP3 and re-encoding it as another MP3 discards a second round from what is already an approximation, and the artifacts compound. If your song only exists as an MP3, upload that MP3 as-is rather than "cleaning it up" through another export. If you have the WAV, use the WAV.
- Upload the version you are actually releasing. If you are re-cutting vocals tomorrow, wait. The tempo, the sections and the plan are all built from the file you send, and a new mix means a new plan.
- Bounce a stereo mixdown, not stems. One file, both channels.
If your song came out of a generative tool rather than a DAW, the export rules are the same: download the finished file from the tool and upload it. Melodious does not fetch songs from links.
Contrast: give the listen a shape it can read
This is the least obvious lever and the one worth understanding, because it explains a whole class of "why did it plan it there" reactions.
Your song's sections are not read from tags or metadata. The director finds them from the audio itself and shows them on the song map. A song with clear contrast between its parts gives it more to work with than one squashed to a near-constant level.
What that means in practice:
| Your master | What the section read tends to look like |
|---|---|
| Clear arrangement contrast, such as a sparse verse and a big chorus | Boundaries land close to where you hear them |
| Heavily limited, near-constant level throughout | Sections still appear, but boundaries may drift from the musical ones |
| A long quiet intro or ambient outro | Tends to read as its own low-energy section, which is usually right |
Be honest with yourself about two things here. The sections are inferred from the audio rather than from music theory, so if your actual chorus is the quiet one, expect the plan's peak to land somewhere else. And you do not need to un-master your song or hand over a dynamic "pre-limiter" version to fix this. Modern masters are loud; that is fine. The fix is to check the song map against your song's real structure and pick the part yourself: tap a section, or drag across the waveform, and press Use my selection in the chat.
If you already know your song's map, where the pre-chorus lifts and where the bridge drops out, writing it down before you upload is worth more than any export setting. Storyboarding against the song's sections walks through how to do that.
Vocals: can the mix be heard clearly enough to read?
The director reads the vocal and the lyric lines from the mix you upload, so how present the vocal is in your master directly affects how much of it comes back.
A vocal sitting where a finished pop or hip-hop master would put it comes back cleanly. A vocal buried under a wall of guitars, drenched in reverb, heavily pitch-processed, or double-tracked in a dense arrangement is harder to read, and you will see that in the result: words missing, words wrong.
Which leads somewhere useful rather than somewhere depressing:
- Check the lyric lines before you rely on them. The song map shows them. If the director misheard a word, mention it in the chat, so the error never reaches the plan.
- If you want words on screen, this is a different tool. Melodious does not put lyrics or captions on screen. A present vocal still matters, because the director plans the shots against the lines it hears. See the lyric video guide for tools that do show words.
- Do not remix your song for the AI. Pushing the vocal 3 dB above where it belongs to help the listen is a bad trade: you would be publishing a video whose audio is not your record. Upload the master, and mention what comes back wrong.
Which part of the song will you get?
The director's first question is which part of the song to make, from a short part up to the full song. It offers the parts that fit your song, for example the hook, the chorus, a verse and the chorus, or the whole song. Tap one, type the part you want in your own words, or tap a section on the song map (or drag across the waveform) and press Use my selection. A video can be up to 5 minutes long.
Everything after that question, meaning the ideas, the plan, the storyboard and the video, covers only the part you pick, so you never pay to draw scenes you won't use. A short part is the cheapest thing worth looking at: make 30 seconds first, see the look and the pacing, and make the longer one once you love it.
So before you press Start, ask yourself which passage of the track you want a video of. If it is a quiet second verse, or a bridge, or the a cappella intro, say so in your first sentence, or pick it yourself on the song map. "Build the whole thing around the stripped-back bridge, not the big chorus" is a steer worth spending a sentence on.
What not to do is trim the file down to the part you want. A trimmed upload is read as if that excerpt were the entire song, which throws off the sections and the pacing that come out of it.
What to do if your track is an instrumental
Instrumentals work. A song with no vocals is planned as an instrumental, with no invented lyrics and no singing shots; tempo and sections still drive the plan.
Two things change, and both are on you rather than the file:
- What you write carries more of the meaning. With a vocal, there is a lyric giving the director something concrete to picture. Without one, whatever you type is the only semantic input. "Instrumental post-rock, a lighthouse keeper through one winter, no people in the wide shots" is a brief. "Instrumental track, make it cinematic" is not. How to write a brief for your AI music video director goes deeper on this.
- There are no singing shots. There are no words to sing on camera, so the video is carried entirely by the images and the edit.
The pre-flight checklist
Two minutes, once, before you press Start:
| Check | Why it matters |
|---|---|
| Finished master, not a rough mix | The plan is built from this file; a new mix means a new plan |
| Full length, not a trimmed excerpt | A trim is read as if it were the whole song |
| MP3 / WAV / FLAC / M4A / AAC / OGG, under 50 MB | These are the accepted formats and the size limit |
| Exported, not recorded off a speaker | Room noise and clipping blur what the director hears |
| Not re-encoded from another lossy file | Lossy-on-lossy artifacts compound |
| You know which part of the song you want a video of | The director's first question is which part to make |
| You know whether the vocal is clearly audible | It determines how much of the lyric comes back |
| One or two sentences written about what you want to see | It is what the director builds the three ideas from |
Mistakes that cost you a round
| Mistake | What happens | Fix |
|---|---|---|
| Uploading a rough mix "just to try it" | You plan against a song that will not ship | Wait for the master, or accept you will redo it |
| Trimming to the section you want | The excerpt is read as the whole song | Upload full length, pick the part in the chat |
| A phone recording of a playback | Noise and clipping blur what the director hears | Export the file properly |
| Re-exporting an MP3 as another MP3 | Compounded lossy artifacts | Upload the original file untouched |
| Pressing Start with no idea at all | The director has tempo and structure and nothing about intent | Write a sentence or two first |
| Assuming the lyric lines are correct | A misheard word shapes the plan | Read the lyric lines on the song map before you approve anything |
| Assuming it will use the start of the song | It makes the part you pick | Choose the part on the first question |
What you do not need to bother with
Just as useful as the list above:
- Perfect loudness targets. There is no number to hit.
- Choosing WAV over MP3 on principle. Within the accepted formats the difference is not what stands between you and a plan you like.
- Providing a BPM or a key. The director reads the tempo from the audio.
- Supplying a lyric sheet. The lyric lines are read from the mix and shown on the song map.
- Separate stems. One stereo mixdown is what is wanted.
Two free checks help here: the BPM and key finder reads your track's tempo and key, and the audio cutter trims it to the part you want.
Two minutes now, or a redo later
The compressed version of all of this: upload the release master, at full length, exported rather than recorded, and know two things about it before you press Start: which part of the song you want a video of, and how clearly the vocal sits in the mix. Those two facts predict most of what comes back in the ideas and the plan.
Everything after that is changed by asking in the chat, because the panels are view-only, and nothing is made until you approve its price. But the director only ever gets one thing to listen to, and it is the file you hand it. Make that the good one.
Got a finished master sitting on your drive? Start with your song and see what the director makes of it.
Upload the master, not the rough
The video is planned against the file you send: tempo, sections, vocal and all. Send the version you'd release.
Start with your songFrequently asked questions
What audio format should I upload for an AI music video?
MP3 or WAV are both fine, and Melodious also accepts FLAC, M4A, AAC and OGG, up to 50 MB. Export straight from your DAW, distributor or the tool that made the song. What matters far more than the format is that it is the finished full-length mix rather than a rough bounce, a trimmed preview, or a lossy file that has been re-encoded several times over.
Does the audio quality of my song affect the music video?
It affects what the director can hear, which is what the plan is built from. A clean export gives a clear picture of tempo, sections and vocal. A phone recording of a speaker adds room noise, clipping and a wrecked low end, which blurs all three. A good MP3 and a WAV are close enough that the difference is not worth agonizing over; a recording of a playback is not.
Should I upload a mastered version or a rough mix?
The mastered version, the one you are releasing. A finished master has the intended shape. A rough mix with an unbalanced vocal or a missing arrangement section produces a plan for a song that will not exist by release day. If your master is not ready, it is worth waiting rather than starting again against a second file later.
Do I need to trim my song to 30 seconds first?
No, and it is usually the wrong move. Upload the full track. A trimmed file is read as if it were the whole song, so the sections it reports are those of the excerpt, not of your record. The director's first question is which part of the song to make, from a short part up to the full song, so you can ask for just the part you want, or tap a section on the song map and press Use my selection.
What if my track is an instrumental?
It works. A song with no vocals is planned as an instrumental, with no invented lyrics and no singing shots, and tempo and sections still drive the plan. Two things change: there are no lyric lines to lean on, so what you write about your idea carries more of the meaning, and there are no singing shots. Say what the piece is about in your first sentence rather than assuming the music will imply it.
Does a heavily compressed, loud master hurt the plan?
It can flatten the contrast the director hears the sections from, so boundaries may land less where you would put them. You do not need to un-master anything. Check the sections on the song map against your song's real structure, and pick the part yourself by tapping a section or dragging across the waveform, then pressing Use my selection.