Rewriting O Fortuna for a Custom Grandchild Gift

When Sandra came to the ChangeLyric done-for-you service and asked me to rewrite Carl Orff's "O Fortuna" into a silly packing song for her grandchild's trip to visit Grandma, I laughed. Then I looked at the original and stopped laughing. O Fortuna is the ominous Latin choral piece from Carmina Burana that every movie trailer since 1982 has ripped off.

It is also one of the hardest kinds of songs anyone has ever asked me to swap lyrics on, and the reason is simple: there is no instrumental in it. The song IS the vocals.

Her note to me was exactly 12 words: "Song must sound identical to original. Only the lyrics change." The original is dense Latin; the new version turned it into English about toothbrush packing, underwear, garlic sausage, and a long-suffering grandkid enduring a ride with Grandma.

The creative brief was perfect. The technical brief was brutal.

Why O Fortuna Is The Hardest Kind Of Swap

Most songs have a clean split between vocals and instrumental. You isolate the vocal stem, you replace what needs replacing, you sit it back on top of the original backing track. For something like a Neil Diamond pronoun swap or a Taylor Swift graduation rewrite, that workflow handles most of the work.

O Fortuna has none of that. The "instrumental" is a full choir stacked with timpani and orchestra. There is no vocal layer you can peel off and rebuild.

A stem separator trying to pull vocals out of O Fortuna gives you the entire song back, slightly phased. That is why the usual bulk V3 flow or targeted Horizon editing both break down here. They cannot find a vocal to hold onto.

Producer's Note: How It Actually Got Made

Here is the exact recipe I used. You can do the same thing inside ChangeLyric's DIY tools if you are willing to put in a little mixing time.

Step 1: Record your own scratch vocal, two octaves. I sang the new English lyrics over the original at whatever pitch was closest, once in a high octave and once in a low one. The scratch take is not meant to sound good. It is meant to give the AI a vocal anchor to steer from later. Stacking two octaves gives the follow-up model something to blend and widen.

Step 2: Duck the original with EQ and volume. I used a surgical EQ cut in the vocal band and a volume rider following my scratch take to push the original choral mass down wherever my new words needed to live. This was all standard DAW work. Nothing fancy. Just making room.

Step 3: Smooth the scratch with an AI cover pass. I ran the ducked mix through Suno as an AI cover to turn my rough scratch into a convincing choral texture. The same idea works with ChangeLyric's built-in Smooth feature if you would rather stay inside one tool. The trick is giving the model a real vocal to clone, not a silent mask.

What It Teaches

If a song is basically 100% vocals, forget clean stem swaps. Record your own scratch take, duck the original underneath it, then let an AI cover model clean up your singing. That three-step pattern turns almost any all-vocal song into a lyric-swap candidate, and it is the kind of workflow you would never guess from the UI alone.

Sandra got her rush WAV back two business days after ordering and left a 5-star review. For a song everyone said was impossible to rewrite, "I love it" is a good outcome.

Browse more projects like this in our lyric swap showcase, or start your own custom lyric swap. If you would rather do it yourself, the ChangeLyric DIY tools come with a 7-day free trial. Try the three-step scratch vocal trick on your own hardest song.