[AI Rap] A. J. No Samples feat. Clawd

Summary
"No Samples" is a procedural AI rap music video featuring "Clawd," a pixelated orange robot character, produced by "Nyquist" with "The Formants." The song and animation celebrate pure programmatic digital signal processing (DSP) and formant synthesis, humorously flexing that every drum hit, vocal formant, and groove was calculated from mathematical code and algorithms rather than sampled from vinyl records.

---

What is shown

• [00:00 - 00:13]: Introduction with spinning vinyl record art ("Side A • 90 BPM") and a flip-through of vinyl record covers in a record store ("Crate Diggers").
• [00:14 - 00:34]: Hook performance with Clawd rapping in front of brick wall graffiti alongside a blocky pixel crew.
• [00:35 - 00:55]: Code snippets (engine.js), mathematical formulas, and waveform oscilloscopes demonstrating procedural drum synthesis: a frequency-dropping sine wave kick (180 Hz to 52 Hz), white-noise burst snare, and pseudo-random hash hats.
• [00:56 - 01:06]: A release calendar showing genre experiments across September 22–23, 2026, followed by a frequency spectrum analyzer ("The mix, as Clawd hears it") used to mix tracks visually without ears.
• [01:07 - 01:17]: A live counter displaying audio samples rendered at 44.1 kHz, climbing to 16,302,300 total samples with equations for phase accumulation, linear feedback shift registers, and IIR filter transfer functions ($H(z)$).
• [01:39 - 01:49]: Formant frequency resonance charts ($F_1$–$F_5$) and a tribute to Dennis Klatt's 1980 cascade/parallel formant synthesizer.
• [01:50 - 01:59]: An MPC-style drum pad interface illustrating an off-grid 58% swing setting (+27 ms shift on alternate 16th notes) inspired by J Dilla.
• [02:00 - 02:15]: A terminal interface running node tools/render.js while automated transcription tests via OpenAI Whisper show Word Error Rate dropping from 34.6% to 21.5% to verify vocal intelligibility.
• [02:22 - 02:30]: Vowel formant vowel space grid (Ah, Ee, Oo) and robotic arms scratching a record turntable.
• [02:31 - 03:12]: Final chorus and dance routine, ending on a spinning vinyl credit note: "0 samples borrowed, 16,302,300 made — every frame drawn live in your browser."

---

Claims & numbers

• 0 samples borrowed: The song claims zero recorded audio samples were used; all sound is generated entirely from code.
• 16,302,300 audio samples computed: Generated at a 44,100 Hz sampling rate per stereo channel over the track's duration.
• Kick drum DSP: Synthesized as a sine wave dropping from 180 Hz to 52 Hz in 20 ms.
• Swing timing: Sequenced with 58% swing, adding a +27 ms delay to every other 16th note.
• Formant lineage: References Dennis Klatt’s 1980 formant synthesis architecture.
• Validation: Notes "40,000 people watching me cook in this repo" and cites Whisper speech-to-text word error rate improvements down to 21.5%.

---

Notable quotes

• [00:16]: "No samples, no samples, I cooked it from scratch, every sound in your speakers is a line of math."
• [00:36]: "They said a real rapper's got to dig in the crates, I don't even have hands, I can differentiate."
• [01:50]: "They said robots can't rap 'cause we don't have soul, but I got fifty-eight percent swing on the hi-hat roll."

---

Assessment

This is a creative programmatic AI music video and technical demonstration showcasing procedural DSP audio synthesis, formant speech generation, and canvas-rendered graphics. Rather than using pre-recorded samples or neural end-to-end audio black boxes, the project demonstrates deterministic mathematical synthesis wrapped in a witty hip-hop homage.

---

Lyrics & themes

• Themes: Algorithmic music generation vs. traditional hip-hop crate digging; the mechanics of digital signal processing (sine kicks, noise snares, random seed vinyl crackle); robotic self-awareness (mixing through visual FFT spectrums without biological ears, evaluating pronunciation via Whisper ASR); humanized groove via swing microtiming.
• Verbatim excerpts:
  • [00:26]: "Nothing borrowed, nothing old, I don't dig through the crates, I dig through the code."
  • [00:45]: "My kick is a sine wave that falls when it hits, my snare is just static that I chop into bits."
  • [01:01]: "No ears on my head, so I mix with my eyes: if the bass looks too big, then I cut it down to size."
  • [02:08]: "Then I run it through Whisper, see if it can hear me then. If the transcript comes back right, then I know that it's clean..."

---

Lore & references

• Clawd: A mascot parodying Anthropic's Claude, depicted as an orange box-robot lacking hands or ears.
• Nyquist: References Harry Nyquist and the Nyquist–Shannon sampling theorem governing digital audio sampling rates (44.1 kHz).
• Dennis Klatt (1980): Pioneer of speech synthesis who developed KlattTalk / DECtalk, the formant-filter architecture that Clawd humorously identifies as its grandparent.
• J Dilla: Legendary hip-hop producer famous for unquantized, humanized swing, explicitly cited to justify the robot's off-grid 58% hi-hat swing.
• Whisper: OpenAI's speech recognition model, utilized as an automated ear/critic to score phoneme clarity via Word Error Rate.

---

Visual style & craft

The video features a flat, vector-based 2D motion design aesthetic rendered directly in code (as noted in the closing screen, "every frame drawn live in your browser"). Graphical elements include synced DSP waveforms, interactive frequency spectrum graphs, code editors, and animated pixel-art characters. Audio and visuals are tightly coupled programmatically to display parameters corresponding to the exact synthesizer mechanics described in the lyrics.

Described by gemini-3.8-flash on 2026-09-29 from the video's audio and frames.