For Scott Zona, from Peter Repetti · updated October 21, 2026
Scott, the current favorite is at the top: your shorter script, read by an ElevenLabs voice. Below it are the things still to decide, then the other versions, and last, everything we tried on the way to choosing the voice.
57 s. Your edit of the shorter script, word for word, with the same cut and music as the voice C version below. The voice is the one you picked, "Brian (younger)", on ElevenLabs' newest model, asked to sound excited. ElevenLabs gives clean studio audio, so the faint background hiss of the free voices is gone. "Love palms? You'll fit right in." was re-recorded because the first take came out flat.
All plants grow. But, palms put on a show! From steamy jungles to quiet desert springs, they shape landscapes all over our planet.
Back in 1956, a few passionate palm lovers founded the society in Florida. Today, the International Palm Society is a home for palm people all over the world.
We publish Palms, our acclaimed journal, and we support research that's uncovering their secrets.
And when wild palms are in trouble, we step in! We find threatened species, and back projects that protect them and the places they grow.
Our members meet up in palm-rich places around the globe, and keep the conversation going on PalmTalk, our online forum.
Love palms? You'll fit right in. Join us at palms.org!
Still to decide
Two choices from you, and I'll rebuild with them.
Music: four options
The current favorite and the other landscape videos use the upbeat track from the TikTok version. You might prefer something calmer, so here are three more, each with your first two paragraphs over it, mixed as they would be in the video. All four are original tracks I composed for this, so there's nothing to license. Tell me which you like and I'll rebuild the video with it.
Which script: shorter or your 45-second one?
The favorite uses the shorter script (57 s). Your full 45-second script is under "Other versions", still in the free voice C (1 min 53 s). If you'd rather use the longer one, I can record it with Brian too.
Other versions
Earlier cuts, kept for comparison and in case you want to use any of them.
58 s. The same script and cut as the favorite, read by the livelier voice C from the free voice software. Your edit of October 21: International Palm Day and Martius are out, and the founding in Florida is in. The sealing-wax palms now go under "Love palms? You'll fit right in." I've also fixed the glitch you spotted: the shot of the sealing-wax palms started during the dissolve from the double-coconut forest, and slowing the footage stretched that into a flash. It now starts after the dissolve, in every video.
1 min 53 s. Your words exactly; a few full stops became exclamation marks, which changes how the voice reads them. The voice is a livelier take of voice C, rendered sentence by sentence so each sentence keeps its lift. The opening holds on the title card and the palm grove, the palm-flower shot covers "For nearly seven decades…", and the 1956 bulletin comes in exactly on "Founded in nineteen fifty-six."
A vertical, TikTok-style version
Download the videoDownload the audioVoice onlyMusic onlyScript Video: MP4, 1080×1920, 27 s. Audio: voice and music together, MP3, 0.7 MB. The voice-only and music-only tracks are the same length and line up with the video, so you can put the voice over a different sound or reuse the music. The script is a Word file, for editing.
A playful 27-second cut of your footage for TikTok, Reels and Shorts, to show what a vertical version could feel like. The script is new and casual, but every claim comes from your scripts (1956, PALMS, research grants, conservation, meetings, PalmTalk, International Palm Day, palms.org). Please change any wording you don't like. Voice C again, faster. The music is an original track I composed for it, so there's nothing to license. When posting to TikTok you could also swap in a trending sound from TikTok's own library. The covers come from your video, and the end card keeps your photo credits.
First draft (October 20): voice C reading your 45-second script
1 min 57 s. Voice C, made a little faster and livelier than in the original sample. To show how a narrated version could feel, I re-cut your footage to the narration rather than squeezing the narration into 50 seconds. Each paragraph plays over the footage that matches it: the opening over the palm landscapes, the history over Bulletin No. 1, the journal and books over the covers, conservation over the forest, the meetings over the garden tour, International Palm Day over the sealing-wax palms, and the web address over the end card. Clips are slowed to fit, which is why some motion looks soft; the shorter script lets the footage run nearer normal speed.
How we chose the voice
Every sample from the search, newest first. Click a heading to open it.
Your two ElevenLabs picks reading the opening, three ways eachBrian (younger) and Logan. You chose Brian on the newest model, asked to sound excited.
Each reads the first lines of the shorter script: "All plants grow. But, palms put on a show! From steamy jungles to quiet desert springs, they shape landscapes all over our planet." These clips are as they came from ElevenLabs, with no editing.
Brian (younger)
Logan
Eleven ElevenLabs voices, from their own sample clipsThe shortlist you picked Brian (younger) and Logan from.
The free voice tools had taken voice C about as far as they could, and they leave a faint background hiss. ElevenLabs, a paid voice service, gives clean, more natural and more expressive speech. These are eleven of its American male voices, closest in pitch to voice C first. Each voice reads its own sample line, not your script.
"Hz" is the voice's typical speaking pitch: lower numbers are deeper. Each clip is short and reads different words, so treat the numbers as a rough guide.
How lively? Your first two paragraphs, five waysMaking voice C more animated with the free voice tools.
The free voice tools trade liveliness against keeping exactly the same voice. Copying voice C's original recording keeps the same man but also his calm delivery. Copying a livelier take gives more energy, from a close relative of voice C rather than the identical voice. The voice C videos use the fourth clip.
The first eight voices (October 20), reading both of your scriptsFree voice software on my own computer, deepest first. We settled on voice C.
Timing: read at a natural pace, the 30-second script takes 65–81 seconds and the 45-second script 92–118 seconds. Both have roughly twice as many words as fit (158 and 237 words; 30 seconds holds about 75 words, 45 seconds about 110). That's what led to the shorter script.
A · deep
Supertonic 3, voice M5 · speaking pitch about 90 Hz
B · deep
Kokoro, voice Onyx · speaking pitch about 87 Hz
C · deep, younger
OmniVoice, made from the description "male, young adult, low pitch, American accent" · about 88 Hz
D · deep
Supertonic 3, voice M2 · about 92 Hz
E · baritone, quicker
Supertonic 3, voice M3 · about 100 Hz
F · baritone, younger
OmniVoice, made from the description "male, young adult, very low pitch, American accent" · about 105 Hz
G · lighter baritone
Kokoro, voice Puck · about 110 Hz
H · lighter baritone
Kokoro, voice Echo · about 112 Hz
Read as written, except that "1956", "April 17th" and the web address are spelled out so each voice says them correctly. Speaking pitch is the measured median; a baritone speaking voice usually sits around 85–120 Hz.