How We Made a Minimal Lyric Video
We built “Together We Rise” as a black‑on‑black lyric video with stick‑figure loops, glitch passes, and clean captions. Here’s the exact, low‑budget path so you can make one tonight.

This Week at The AI Muse - For the AI‑curious
TL;DR: We built “Together We Rise” as a black‑on‑black lyric video with stick‑figure loops, glitch passes, and clean captions. Here’s the exact, low‑budget path so you can make one tonight.
What we made (and why)
Vibe: conscious hip‑hop, protest energy, chant‑ready.
Look: black canvas, thin white lines, tiny cyan glow.
Hook: TOGETHER WE RISE / IGNITE THE LIGHT.
Why minimal? It’s cheap, punchy, and puts lyrics first. Perfect for YouTube where we keep things b/w and let Reels go color‑crazy later.
What you need (free or cheap)
- Music draft: Udio or Suno (pick one).
- Editor: CapCut (desktop or mobile).
- Visuals: Sora (if you have it) or NightCafe/SDXL stills you animate inside CapCut.
- Captions: any text editor to make an .srt.
Handle everywhere: @theaimuze
Stream/Save:
Step 1 — Lock the vibe (use this prompt style)
Vibe prompt (copy):
Minimal protest energy on a black canvas; militant but warm. Dusty trap swing, chant‑ready chorus. Keep it soulful and human, non‑aggressive. Hook concept: TOGETHER WE RISE / IGNITE THE LIGHT.
Use that to guide your writing/performance. Keep lines short so they caption clean.
Step 2 — Draft the song (Udio/Suno)
- Build 2× versions at different tempos.
- Pick the one with the easiest sing‑along chorus.
- Export WAV/MP3 (peaks under ‑1 dB if you can).
(No need for tech wizardry—we value feeling over settings here.)
Step 3 — Make captions (.srt)
Mini SRT skeleton (paste & tweak times):
1
00:00:08,000 --> 00:00:11,500
No need to turn a blind eye
2
00:00:11,500 --> 00:00:15,000
Time to recognize
Tips:
- Keep each caption 1–2 lines, ≤32 chars/line.
- Minimum duration 1.6s so people can read.
- You can refine timings inside CapCut later.
Step 4 — Generate simple visuals
Pick one path:
A) Sora (text‑to‑video, seamless loops)
March loop (3–4s):
Pure black background. Minimal stick‑figure protesters in profile march left→right in one metronomic cycle. White 2–4 px lines, no faces, no shading. One blank round sign (no text). Subtle cyan glow pulse on the beat. Begins and ends on the same pose for a perfect loop. 9:16 or 16:9, 24fps.
Torch pulse (3–4s):
Single figure lifts a torch; cyan halo blooms once then fades; returns to start pose for a seamless loop. White line art, 2–4 px, flat vector.
Neon shapes (2–3s):
Minimal cyan corner brackets and thin rings drifting on black; very faint glow; seamless loop.
B) NightCafe/SDXL (stills → animate in CapCut)
Image prompt:
Minimal white stick‑figure crowd on a pure black background; thin 2–4 px lines; one round sign outline; soft cyan glow; high contrast; no text; crisp edges.
Animate the stills in CapCut with a 2–3% slow zoom or sideways crawl for motion.
Step 5 — Assemble in CapCut (fast path)
- New 1080p project (24 or 30 fps).
- Drop the song at 0:00.
- Import SRT → Captions. Style: off‑white fill, thin black stroke.
- Add your loops on tracks above: march (verses), fist/torch (chorus), neon shapes (bridge).
- Adjustment layer on top: Contrast +12, Highlights ‑10, Saturation +10, Warmth +6, Grain 8–12%.
FX pops (sparingly):
- RGB split (2–4 px) on snares.
- 2‑frame shake on the torch bloom.
- Tiny glow: duplicate layer → Gaussian Blur 12–18 → Screen 20–30%.
Step 6 — Export & upload
- Export: H.264, 1080p, 20–30 Mbps; audio AAC 320 kbps; peaks ≤ ‑1 dB.
- YouTube: title front‑load keywords → Together We Rise (Lyric Video) — Conscious Hip‑Hop.
- Put Stream/Save link + a one‑line CTA in the first two lines.
- Add end screen to your Conscious Hip‑Hop playlist.
Step 7 — Reels/Stories (colorful alt cut)
Keep YouTube minimal. For socials, go colorful & absurd:
- Dancing megaphone army (cyan/pink rings).
- Rubber‑hose torch relay (cyan torch bloom).
- Unity globe bounce (cartoon Earth + confetti flags).
Cut a 15s militant version and a 15s absurd version. Same chorus, different energy.
Steal these (copy‑paste blocks)
Sora negative prompt: no text, letters, logos, words, gradients, 3D, faces, realistic lighting, heavy particles
NightCafe negative: text, logos, gradients, clutter, messy edges
Chorus card (CapCut text): TOGETHER WE RISE → IGNITE THE LIGHT
Pinned comment: Drop your city + one fix ✊
Hashtags: #AIMusic #HipHop #ConsciousHipHop #LyricVideo #TheAIMuse
Troubleshooting
- Captions drift? Select all captions → shift Offset ±0.25s until verse hits feel right.
- Loops feel dead? Add 2–3% zoom or a slow pan.
- Sora adds text? Strengthen the negative prompt; re‑render.
- Crushed blacks? Keep Lift at 0; glow layer opacity ≤30%.
What’s next
- Afrofuturist protest march reel variant (neon circuits + silhouettes).
- Country‑trap experiment in the Boots universe.
- A dub take on Rise & Unite.
If you build your own version, tag @theaimuze so I can share it. And if you want the exact project file (CapCut), reply and I’ll post it next weekly.