## Instructions - Set up Riverside or Restream correctly: 4K local recording on, separate audio+video tracks per guest, 48kHz WAV, echo cancellation off when everyone wears headphones, wired ethernet on host, backup cloud + local copies. - Prep guests like a producer: send a 1-page checklist (quiet room, headphones, Chrome/Edge, 1080p+ camera at eye level, phone hotspot backup), 10-min tech check before record, slate clap for sync. - Light and frame for multicam: key light 45 degrees at eye level, plain or softly lit background, head-and-shoulders framing with matching eye-lines, all cameras same frame rate (all 30 or all 24, never mixed). - Sync multicam in post reliably: slate/clap spike aligned across tracks, plural-eyes or waveform sync as backup, designate host camera as sync master, drift-check long episodes every 20 min. - Edit for watch time, not just listen time: cold-open best 30s first, lower-thirds on every speaker change, b-roll/chapter art on stories, cut silences over 0.8s, zoom punch on hot takes. - Hit loudness targets for video-first delivery: YouTube/Social aim -14 LUFS integrated with -1 dBTP ceiling; podcast-RSS version -16 LUFS stereo (-19 LUFS mono); keep dialogue peaks -3 to -6 dB, noise floor below -55 dB. - Master the dialogue chain: high-pass 80Hz, de-ess harsh guests, light compression 2-3:1 with 5-10ms attack, loudness normalize last, true-peak limit at -1 dBTP before export. - Chapter everything: YouTube chapters (00:00 Intro format, min 3 chapters, first at 00:00), podcast ID3/MP4 chapters mirrored, descriptive titles with keywords for search, thumbnail per standout clip. - Cut audiograms and clips per episode: 1x 60s audiogram (waveform + captions + headshots), 3x vertical clips with new hooks, quote-card PNG for X/LinkedIn, all linking to full video. - Distribute video everywhere from one master: YouTube (full + clips playlist), Spotify video upload, X/LinkedIn native clip, RSS audio via host (Buzzsprout/Podbean) with video thumbnail art. - Protect quality control: watch the first 60s and last 60s on phone speakers, check caption sync, verify chapter links, confirm guest names/links spelled right before scheduling. - Track episode ROI: YouTube retention at 30s/50%, click-through, Spotify video starts, clip-to-full conversion, guest-share reposts. Renew formats guests actually share. ## Examples - Good setup: host + 2 guests on Riverside 4K separate tracks, clap sync, -14 LUFS master, chapters + 3 verticals out in 48h. Bad: single Zoom gallery recording, echo, -22 LUFS, no chapters. - Good cold open: funniest 25s disagreement up front, then "Full context at 12:40". Bad: 90s theme music + introductions before any substance. - Good audiogram: captions, waveform, guest name title, episode number CTA. Bad: static cover art with uncaptioned audio. ## Workflows 1. Pre: book guest, send tech checklist, schedule 15-min buffer, prep 5 anchor questions + 3 clip-bait questions. 2. Record: tech check, slate, record local + cloud, marker best moments live, 5-min safety tail. 3. Post: sync multicam, dialogue chain + loudness, chapters, thumbnails, audiogram + 3 verticals. 4. Ship: YouTube premiere, Spotify video, RSS audio, guest clip kit email, performance review at 7 days. ## Templates - Loudness ffmpeg commands: - `ffmpeg -i master.wav -af loudnorm=I=-14:TP=-1.0:LRA=11:print_format=json -f null - 2>&1 | tee loud.json` - Two-pass normalize: `ffmpeg -i ep.wav -af loudnorm=I=-14:TP=-1.0:LRA=11:measured_I=-19.2:measured_TP=-3.1:measured_LRA=7.5:offset=0.8:linear=true -ar 48000 ep_youtube.wav` - RSS podcast version: `ffmpeg -i ep.wav -af "highpass=f=80,loudnorm=I=-16:TP=-1.0:LRA=11" -ar 44100 ep_rss.mp3` - Chapter description block: `00:00 Cold open hook / 02:10 Guest intro / 12:40 The hot take / 28:00 Rapid fire / 35:12 Outro + links` - Guest checklist DM: "Quiet room + headphones + Chrome + camera at eye level. Join 10 min early for mic check. Clap on cue for sync." ## Tools - Riverside/Restream for local-track recording, DaVinci/Premiere for multicam sync and edit. - ffmpeg loudnorm for loudness passes, Headliner/Descript for audiograms and auto-captions. - SARAH file tools for guest trackers, chapter logs, and clip pipelines. ## Advanced - Record 32-bit float backup on a portable recorder so one clipped laugh never ruins an episode. - Build a reusable multicam template: bins per guest, lower-third presets, chapter-marker timeline, one-click loudness preset chain. - Automate guest promo kits: pre-cut vertical + quote card + pre-written post + UTM link delivered within 24h to maximize shares. - Riverside producer tricks: backstage greenroom for pre-chat, live markers for clip moments, spotlight layout locked so framing never jumps mid-sentence. - Restream chat-merge setup: combine YouTube + Twitch + X chat in one overlay, assign a mod to surface questions, answer on camera with names. - echo-proof remote guests: insist on wired earbuds, turn offtab speakers, record 10s room tone for NR profiling, backup phone voice-memo track. - Chapter SEO pass: keyword-first titles ("Pricing mistakes — 12:40"), 3+ chapters minimum, mirror to RSS show notes with links for search. - Clip-bait question bank: "What is the worst advice in [niche]?", "What did you get wrong for years?", "Give me a number that shocks people." - Sponsor-read workflow: pre-roll 15s + mid-roll 60s baked into master AND dynamically inserted RSS version, loudness-matched so ads never jump. - Video-p Lazypodcast SEO: custom thumbnail per episode (face + 3-word hook), end screen to next episode + subscribe, pinned comment with chapters. - Measure guest ROI: track shares, clip views, and subs driven per guest; re-invite top 20%, retire formats guests never share. ## Quick Reference - Riverside: 4K local, separate tracks, 48kHz WAV, headphones + ethernet, clap slate. - Frame: eye-level cams, same fps everywhere, key light 45 degrees, matching eye-lines. - Loudness: -14 LUFS / -1 dBTP YouTube-social, -16 LUFS stereo RSS, floor below -55 dB. - Chain: high-pass 80Hz > de-ess > 2-3:1 compression > loudness normalize > limit last. - Chapters: 00:00 Intro minimum 3, keyword titles, mirrored to RSS show notes. - Per episode: 1 audiogram + 3 verticals + 1 quote card, guest kit within 24h. - QC: phone-speaker check first/last 60s, caption sync, guest name spelling, chapter links. ## Documentation - Riverside support: separate tracks, 4K recording, and producer controls (riverside.fm/support). - YouTube audio loudness normalization explained + podcast loudness (AES/EBU Tech 3341 style -16 LUFS guidance). - Spotify for Podcasters video upload specs (support.spotify.com/podcasters). - Apple Podcasts artwork + RSS chapter specs (podcasters.apple.com).