Feature
Instrumental or vocals: a listener follows one set of words
Use a sung track when nobody speaks in the ad, and an instrumental under any voice-over. A listener cannot follow two sets of words at once, and both options cost 50 credits. This page covers when each one wins, the one exception that works, and what the two-track pattern costs.
- Either option, any of five lengths
- 50 CR
- Sets of words a listener can follow at once
- 1
- The two-track pattern for one ad
- 100 CR
The rule, and why it is not a matter of taste
Language processing is single-channel. A listener can follow one stream of words and not two.
A sung vocal and a spoken voice-over also occupy the same frequency band, so they mask each other on top of competing for attention.
Run both and the viewer hears a wash. They do not consciously identify the problem, they simply stop following.
That is why an ad with a great voice-over and a great song can perform worse than either alone.
So the decision is structural rather than aesthetic, and it gets made before you type anything into either box.
Empty lyrics box means instrumental. Anything typed in means somebody sings. That is the entire control surface.
Sung track against instrumental bed
Neither one wins outright. The ad you are making decides, and the first two rows go to the sung option.
| Sung track | Cutroom | |
|---|---|---|
| Ad with no voice-over | The lyrics become the ad's words | Works, but says nothing |
| Carrying a memorable phrase | A tune is the best recall device available | A bed cannot carry a message |
| Under a voice-over | Two sets of words, listener follows neither | Stays out of the speech band if briefed right |
| Reusable across many ads | The words tie it to one message | One description covers a catalogue |
| Attempts before it lands | Lyrics and delivery both have to work | Usually right in one or two |
| Price | 50 credits | 50 credits |
Vocals win when the track is the ad
No voice-over, no talking head, no narration. The words being sung are the only words in the piece.
A tune is the strongest recall device available in fifteen seconds, and nothing an instrumental does competes with it.
The constraint is the word budget. A sung line carries about sixty words a minute, so fifteen seconds is about fifteen words.
Put the brand name in the first three, because that is roughly what a scroller hears before deciding.
Repeat the key line. Repetition inside a short track is the only acceleration mechanism the format has.
And say it out loud at singing speed before generating, because there is no fixing a line after the track returns.
This is the whole editor
Highlight a phrase and a clip lands on those exact words. No timeline, no keyframes, no layers.
Instrumentals win everywhere else, and that is most ads
Any ad with a voice-over, a talking head, a demonstration with narration, or a founder speaking to camera.
For all of those, leave the lyrics box empty and describe the absences: sparse, nothing in the midrange, low end and high percussion only.
An instrumental is also reusable in a way a sung track is not, because it carries no message tying it to one ad.
One saved description can front a whole catalogue, with each ad getting its own sibling track for 50 credits.
And instrumentals tend to land in one or two attempts, where a sung track needs the words and the delivery both to work.
That makes the instrumental cheaper in practice even though the sticker price is identical.
It is also the safer default when you are not sure yet what the ad will sound like, because an instrumental never forecloses a decision about narration.
A sung track does. Once the words are in the music, adding a voice-over later means throwing one of them away.
- Voice-over present: instrumental, no exceptions
- No voice at all: sung is on the table and often stronger
- Reusing across many ads: instrumental
- One memorable phrase to plant: sung
The exception that works: sung opening, instrumental body
The first two seconds have no voice-over yet, so a sung line can live there without competing.
Then the talking starts and an instrumental bed takes over for the rest of the ad.
That is two tracks at 50 credits each, which is 100 credits and the cheapest fix available for this problem.
In the editor the music controls are turn it on, pick the track and set how loud it sits, so the separation has to be structural rather than automated.
There is no ducking and no stem separation, which means a single track cannot be made to step aside when speech starts.
Check the join on a phone at half volume. That is where a sung line and a spoken one collide hardest.
The 40 credit pattern
Two tracks, one ad. The only reliable way to have both.
Sung line
First two seconds, 50 CR
Talking starts
The sung track ends here
Instrumental bed
Sparse, no midrange, 50 CR
Check on a phone
Half volume, arm's length
Make the decision before either track exists
Here is how this usually goes wrong. Teams pick sung because a sung track is more exciting to make, then bolt a voice-over on afterwards.
By then the sung track feels like an asset worth protecting. The ad ships with two sets of words in it.
Decide before either one exists and the problem never arrives. That is the cheapest fix on this page and it costs nothing.
Two ads want neither option. A dense voice-over with no pauses has nowhere for either to sit, and both cost clarity.
A demonstration where the product makes a real sound is stronger with that sound than with anything over it.
Play the cut muted, then voice only. If voice only holds, music is a decision rather than a requirement.
Four plain limits sit around both routes. No editing after generation, no stems, no tempo slider, and 120 seconds is the ceiling.
A specific recognisable song is not available, sung or otherwise. That is a rights negotiation.
The track is generated for you rather than licensed from a catalogue. Confirm the usage terms that apply to your account before running one behind paid spend.
Questions people ask
- Can I have vocals quietly under a voice-over?
- No, and volume is not the fix. The conflict is attention rather than level, because a listener cannot follow two sets of words at once. A quiet sung line still splits attention and still masks the same frequencies.
- How do I make an instrumental?
- Leave the lyrics box empty. That is the only thing separating an instrumental from a song. Adding no vocals to the description as well helps when the genre you asked for usually carries a vocal line.
- Is a sung track worth more attempts?
- It needs them. The words and the delivery both have to work, so budget five to ten attempts at 50 credits each. An instrumental usually lands in one or two, which makes it cheaper in practice despite the same price.
- Who should never use a sung track?
- Anybody whose ads carry narration, which is most direct response advertisers. Leave the lyrics box empty and brief a sparse bed. Save sung tracks for the ads where nobody speaks and the music is doing the talking.
One set of words at a time. The lyrics box is where that decision gets made, and it is made before you write anything.