Pro Tools: Post-Production, Podcasts, Live Sound & Game Audio

You finished the master at the end of Chapter 19. [Song]_v8_Master_Streaming.wav sits on your hard drive, ready for distribution. For some projects that is the end—the song goes to streaming and the work is done. For most projects in the modern audio industry, it is the beginning of another phase: the work to picture. Film. Television. Commercials. Podcasts. Video games. Streaming originals. The fastest-growing corners of the audio industry are not music releases—they are post-production, podcasting, and live events. The engineers who can do this work in addition to music are the ones who never run out of jobs.

My first film score was for a short film called Fool Circle. I spent a full twelve-hour day composing, producing, and mixing the entire score from scratch. That version—the one from that single marathon session—was the one we ran with. The film went on to win the Canon/Vimeo “Story Behind the Still” contest. The producer loved the work so much that she connected me with another project: a music video for a song I had produced called “Breakout” by My Hero. That track eventually became the official theme song for both the LA Clippers and the St. Louis Cardinals. One twelve-hour day led to two professional sports franchises playing my music. You never know where a project will take you.

That is the thing about post-production—it is the door most audio students do not even know exists. You learn to record and mix music, and then someone hands you a film and says “make this sound like something.” Suddenly you are not just an engineer. You are a storyteller. Every sound you place, every edit you make, every moment of silence you leave in—it all shapes how the audience feels. And the work is everywhere. Television. Film. Commercials. Podcasts. Video games. Streaming content. The demand for audio professionals who can work to picture has never been higher.

Pro Tools is the industry standard in television, film, and post-production. Whether you are scoring a film, editing dialogue, designing sound effects, or mixing a feature in Dolby Atmos, Pro Tools provides the tools and workflows that professionals rely on every day. This chapter covers all of it—plus podcasting, remote collaboration, live sound, and game audio, because your skills go far beyond the music studio.

After more than twenty years working in this industry, here is what I will tell you: the engineer who can only do music has a smaller career than the engineer who can also work to picture. The technical skills you have built across this entire book—signal flow, microphone technique, EQ, dynamics, time-based effects, mixing, mastering—all transfer directly. What is new is timecode, picture sync, and the discipline of making sound serve a story. Once you have that, the doors that open are doors most musicians never see.

“We had a projection booth with some very, very old simplex projectors. They had an interlock motor which made a wonderful humming sound…”

—Ben Burtt, recalling the creation of the lightsaber sound for Star Wars (filmsound.org interview)

That is post-production in one anecdote: the sound that defined a generation of cinema came from an idling projector motor and a faulty microphone cable picking up the hum of a nearby TV set. The discipline of post-production is the discipline of finding the right real-world sound, recording it, transforming it, and placing it on picture so the audience feels something they cannot name. Before you touch a fader, you need to speak the language. Every role below represents a career path—and on smaller projects, you might fill all of them in a single session.

Film Scoring — is original music written specifically to accompany a film. The score is timed to begin and end at precise points, enhancing the dramatic narrative and emotional impact of each scene. My Fool Circle score had to hit specific moments—a door opening, a character turning, a reveal—down to the frame. Miss the hit point by half a second and the music feels disconnected from the picture. This precision is what separates film scoring from songwriting.

Foley: — Named after Jack Foley, the sound effects pioneer at Universal Studios, Foley is the art of creating everyday sound effects that are added to film in post-production. Footsteps, clothing rustle, doors closing, glass breaking—sounds that were either not captured cleanly during filming or need to be enhanced. Foley artists perform these sounds live in a studio while watching the picture, and the results are recorded into Pro Tools. It is one of the most creative and physical jobs in audio.

The craft divides into three passes. Feet come first: the artist walks every character's steps in sync with picture, on a floor of swappable surfaces—the Foley pits, framed boxes of gravel, sand, concrete, hardwood, and leaves built into the stage floor—wearing shoes chosen to match the character. Moves is the cloth pass: a fistful of fabric worked in rhythm with each actor's body so their clothing rustles when they do; without it, dubbed dialogue plays against an eerily silent body. Specifics are everything handled—props, door latches, cutlery, the coffee cup set on the table. The artistry is rarely literal: the prop that reads as the sound beats the real object (the classic example: a kiss performed on the artist's own forearm, celery snapped for breaking bone), and a good Foley stage is a junk shop curated over decades. Recording is close and dry—usually one good condenser per pass and a tight room—because these tracks must sit under dialogue believably and survive being placed into any space the re-recording mixer needs.

That placement has its own classic technique: Worldizing — re-recording clean audio through real spaces to make it belong to the picture's world: play the element from a speaker in a stairwell, a parking garage, a car interior—and capture it with mics in that space, baked-in reflections and all. Walter Murch named the practice on American Graffiti, re-recording the radio soundtrack through real speakers in real locations as the camera's perspective changed. Convolution reverb (Chapter 17) is its digital descendant, but engineers still worldize when a sound must sit in a space no impulse response quite captures.

ADR (Automated Dialogue Replacement): — When dialogue recorded on set is unusable—too much background noise, a plane flew over, the actor mumbled—the actor comes into a studio and re-records the lines while watching the original footage. The actor must match their lip movements and delivery to the original performance. Production beeps (a series of sine wave tones) cue the actor when to start. VocAlign Ultra is an ARA2-integrated tool that can time-align the new recording to the original, saving hours of manual editing.

Sound Design: — The process of creating the entire sonic world of a film or project. A sound designer might layer dozens of recordings to build a spaceship engine, process animal vocalizations to create a creature's voice, or construct ambient beds that establish location and mood before a single word of dialogue is spoken. In practice, sound design also includes cleaning up production audio—removing noise, fixing levels, making dialogue intelligible. It is part art, part problem-solving, and entirely essential.

Mixing and Mastering for Film: — A typical motion picture will have hundreds of audio tracks mixed for stereo, 5.1, 7.1, or Dolby Atmos surround sound. Unlike music mixing where you balance 40–80 tracks, a feature film mix may involve managing dialogue, ADR, Foley, hard effects, ambient backgrounds, and multiple stems of score simultaneously—often exceeding 500 tracks. The re-recording mixer must ensure dialogue remains intelligible over action sequences, music supports without overwhelming, and dynamic range works in both a theater and a living room.

Post-Production Roles

On a feature film or major streaming series, audio post is its own department with specialized roles. Knowing the structure tells you where you fit and how to talk to the team:

• Supervising Sound Editor (SSE) runs the entire post-production sound team. The SSE attends the spotting session with the director and picture editor, builds the sound budget, hires the editorial crew, and is the creative lead through final delivery. On big features, the SSE also designs key sounds.
• Re-Recording Mixer (RR Mixer) is the senior mixer at the dub stage who creates the final mix. Major features split the role between two or three RR mixers—typically one handling Dialogue/Music, one handling Effects/Foley/Backgrounds, and sometimes a third on the Atmos render. Re-recording mixers are credit-eligible for Oscars and BAFTAs.
• Dialogue Editor cuts production audio, replaces lines with ADR where needed, fills gaps with room tone, and prepares clean dialogue tracks for the mix. The dialogue editor is the unsung MVP of post—most of what audiences perceive as “sound quality” is actually dialogue editing.
• Sound Designer builds non-existent sounds—creature voices, weapon impacts, alien environments, magical effects—through layering, processing, and recording. Often part of the editorial team; sometimes a freelance specialist hired for specific sequences.
• Sound Effects Editor cuts and tracks-out effects (hard FX, backgrounds, walla) from sound libraries and field recordings to picture.
• Foley Artist performs the Foley to picture in the studio (footsteps, clothing rustle, prop handling, fights). Pure performance art—and unionized through IATSE Local 700, the Motion Picture Editors Guild (the MPSE is a professional society, not a union).
• Foley Mixer records and processes the Foley artist's performance.
• ADR Mixer runs the ADR session—records replacement dialogue, slates each take, manages the actor and director through the loop process.
• Music Editor cuts and conforms music to picture, tempo-maps cues for the composer, and prepares music stems for the mix.

On smaller projects you may fill all of these roles yourself. On larger projects you fill one. Either way, knowing every role exists tells you who you are talking to and what they need from you.

The first time you import a video into Pro Tools and the audio drifts out of sync by the end of the scene, you will understand why timecode matters. In music, you work in bars and beats. In post-production, you work in hours, minutes, seconds, and frames—and if your frame rate is wrong, nothing lines up.

The Society of Motion Picture and Television Engineers developed SMPTE timecode—a standard that gives every frame a unique eight-digit address read as hours:minutes: seconds:frames (for example, 01:00:00:00). This allows synchronization whether the media is moving forward, backward, at slow speed, or at fast speed.

The complication is frame rates. Different formats use different rates, and mixing them up will cause sync drift:

Black and white TV in the United States: 30 frames per second.

Color TV (NTSC) in the United States: 29.97 frames per second—either drop frame (certain frame numbers are skipped so the count keeps pace with real clock time—no frames are actually lost) or non-drop (no frames dropped, but the count drifts from real time over long programs).

European TV (PAL): 25 frames per second.

Standard film: 24 frames per second.

When linking multiple devices for video work, designate one as the leader and the rest as followers. All followers lock to the leader's timecode. Without this, every device plays back independently and nothing stays in sync. With DAWs, you can set the program as either leader or follower. Since most DAWs import video directly, making the DAW the leader is typically the simplest approach.

For MIDI to sync with SMPTE timecode, MIDI Timecode (MTC) was developed as a translation layer. MTC sends the standard timecode to all MIDI devices capable of reading it. Most modern MIDI devices include SMPTE-to-MTC conversion built in, but external converters are available if needed.

Here is a mistake I have watched students make more times than I can count: they import a video, start editing audio, get twenty minutes in, and realize the audio is drifting further and further out of sync with the picture. They panic. They try re-importing. They check their buffer settings. They Google “Pro Tools video sync problem” and fall down a two-hour rabbit hole. The fix? They set the wrong frame rate when they created the session. A five-second check at the beginning would have saved them an afternoon.

Before importing video, go to Setup → Session and set the timecode rate to match your video's frame rate (23.976, 24, 25, 29.97, or 30 fps). If you do not know the frame rate, right-click the video file on your computer and check its properties before you open Pro Tools. This is step one. Do not skip it.

Set the session start time to match the timecode of your video. Film and television projects typically start at 01:00:00:00 (one hour) rather than 00:00:00:00, leaving room for pre-roll elements like tone, slate, and countdown. Switch the main time display to Timecode rather than Bars|Beats so you can navigate by SMPTE address. If you are working with film editors who think in feet and frames rather than timecode, Pro Tools supports Feet+Frames display as well.

Pro Tools uses a dedicated Video track to hold imported video, displaying thumbnails in the Edit window and routing output to the built-in video player or an external interface. Pro Tools Studio supports one Video track per session; Pro Tools Ultimate supports up to 64.

Importing video: Go to File → Import → Video. MP4 or MOV files work best. Pro Tools will ask whether to import the audio from the video as well—if so, choose an appropriate folder for the audio files. The video appears in the Video track, and you can resize the player window or move it to another monitor. Toggle the video display with Ctrl+9 (PC) or Cmd+9 (Mac) on the number keypad.

Recommended codecs: This is where new engineers get burned. A client emails you an H.264 file—the same format used for YouTube and streaming—and you import it into Pro Tools. It plays, but when you scrub through the timeline, it stutters. Frames drop. Sync drifts. The problem is that H.264 and H.265 use interframe compression—each frame depends on the frames around it, which means your CPU has to decode multiple frames just to display one. That is fine for watching a movie, but terrible for editing.

Request Avid DNxHD, DNxHR, or Apple ProRes from your video editor whenever possible. These are intraframe codecs—each frame is decoded independently, which means smooth scrubbing and rock-solid sync. If you receive H.264 files (and you will—clients send whatever they have), convert them to an intraframe codec using DaVinci Resolve or FFmpeg before importing. Five minutes of conversion saves hours of frustration.

Spot Mode and Placing Audio to Picture

In music production, being a few milliseconds off is a vibe. In post-production, it means the gunshot happens after the muzzle flash and the audience notices. That is why Spot mode exists, and it is arguably the most important editing mode for working to picture. When you place or move a clip in Spot mode, Pro Tools opens a dialog asking for the exact timecode location where you want the clip to land. A door slam needs to hit the exact frame the door closes on screen. A footstep needs to land when the shoe touches the floor. Spot mode makes this frame-accurate precision possible.

Enter Spot mode by pressing F3 or clicking the Spot mode button in the Edit window toolbar. When you drag a clip from the Clip List or move an existing clip, Pro Tools prompts for the exact SMPTE timecode destination. Type the address, hit Enter, and the clip lands exactly where it needs to be. No dragging, no guessing, no eyeballing.

Spot mode also solves a common ADR problem. When a post house sends a video with burned-in timecode (a TC window burned into the picture), your first job is to line the video up to Pro Tools' own timecode: find a frame where the burned-in number hits a round value, then use Spot mode to place the video clip at that exact SMPTE address. Now the timeline and the on-screen TC agree, and every cue you place lands on the right frame.

Markers and Memory Locations

When I scored Fool Circle, the first thing I did was watch the entire film and drop markers at every moment that needed a musical hit—a door opening, a character turning, a reveal, a cut to black. By the time I started writing music, I had a roadmap. Without those markers, I would have been scrubbing back and forth through the timeline for twelve hours trying to remember where things happened.

Memory Locations allow you to mark hit points—moments where a sound event must occur: a gunshot, a door closing, a music cue entrance, a scene change. Press Enter on the numeric keypad to create a marker at the current playback position. Name each one descriptively (“Car crash,” “Music in,” “Scene 4 start”) so you can navigate quickly. On a feature film, you might have hundreds of markers. On a short, maybe thirty. Either way, they are the difference between organized work and chaos.

For film scoring, markers are invaluable for identifying sync points and tempo changes. Composers use markers in combination with Pro Tools' tempo map and click track to ensure music hits at precisely the right moment. By adjusting tempo at specific markers, you create a variable-tempo click track that guides musicians or virtual instruments to lock with the picture. The music speeds up and slows down with the action—and the audience never knows it is happening.

A handy tempo trick: film runs in seconds, music in bars and beats, and the two line up cleanly at certain tempos. At 60 BPM one beat equals one second; at 120 BPM two beats equal one second (one 4/4 bar = two seconds). Score at 60 or 120 (or a multiple) and your bars and beats map directly onto timecode, so hit points fall on round numbers instead of awkward fractions of a frame.

Session Interchange: AAF and OMF

At some point, a video editor is going to hand you a project and say “here's the session.” It will not be a Pro Tools session. It will be an AAF or OMF file exported from Premiere, Resolve, or Media Composer. If you do not know what to do with it, the project stalls.

AAF — (Advanced Authoring Format) and the older OMF (Open Media Framework) are interchange formats that preserve audio, fades, volume automation, and track layout between platforms. AAF is the modern standard—more reliable, preserves more metadata (plugin assignments where supported, color coding, pan automation, embedded media), and works between Pro Tools, Premiere, Resolve, Media Composer, Nuendo, and Logic. OMF is the older Avid format from the 1990s—simpler, less metadata, but still occasionally requested by broadcast TV facilities and legacy clients. Always request AAF as primary; accept OMF when it is what the picture department can send.

The roundtrip between picture and sound departments is how professional post-production actually works. Here is the canonical six-step AAF roundtrip from Premiere or Resolve into Pro Tools and back:

1. Picture department exports the AAF. The video editor exports an AAF with embedded media (audio files travel with the metadata, not just clip references) and handles—a minimum of 120 frames (about five seconds), often 240, of audio padding around every cut so you have headroom for re-cuts and crossfades.
2. You import into Pro Tools. File arrow Import arrow Session Data, browse to the AAF, accept the prompts. Pro Tools rebuilds the tracks, clip placement, fades, and volume automation from the picture department's timeline.
3. Verify timecode and frame rate. Confirm the session timecode rate matches the source video. If the picture editor exported at 23.976 and you opened a 24 fps session, every clip will drift. This is the single most common roundtrip failure.
4. Do the work. Dialogue editing, ADR, Foley, sound design, score placement, mix—all the chapter walks through above.
5. Conform if picture changes. When picture lock breaks (and it sometimes does), the assistant editor exports a new EDL or AAF with the changes. Conform your existing audio session to the new cut using your handles.
6. Export back. File arrow Export arrow Selected Tracks as AAF/OMF. Choose embedded media. Send the AAF (plus stems and the print master / IMF / ADM BWF as required) back to the picture department for relink.

Learn this workflow. On any post-production gig in the world, this is the first hour of the day.

“If everything is loud, nothing is loud. You really need troughs for the waves to be significant.”

—Randy Thom, Director of Sound Design, Skywalker Sound (Frame.io, July 2021)

A feature film does not get mixed in one pass. The dub stage process is a structured sequence designed to manage hundreds of tracks across hours of picture, and understanding the stages tells you why post takes weeks and where your work fits.

Spotting Session. Before any editorial work begins, the SSE, the picture editor, the director, and sometimes the producer walk through the locked picture together identifying every sound that needs attention—music in/out, effects to design, ADR cues, Foley moments, problem dialogue. Spotting notes drive the entire schedule.

Temp Music and Temp Dub. Long before the final mix, the picture editor or music editor places temporary music (often pulled from existing scores) and a rough sound mix to make the cut work for early screenings. The composer references the temp music for tone and timing; the final score replaces it. Sometimes a director gets so attached to the temp that the composer is asked to write something close to it—“temp love” is a working hazard.

Pre-dubs (Premixes). On the dub stage, the mixers do not jump straight to the final mix. They build pre-dubs first: a Dialogue Pre-Dub (clean dialogue, balanced and de-noised), a Music Pre-Dub (the score balanced and routed to its stem), and an Effects Pre-Dub (Foley, hard FX, and backgrounds bussed and shaped). Each pre-dub may run a full pass of the picture; each generates its own stem. The pre-dub is where most of the heavy lifting happens—attack, release, level, EQ, automation per element.

Final Mix. With pre-dubs printed, the mixers run the picture again, this time balancing the three pre-dub stems against each other in real time, riding dialogue under music, ducking effects under speech, automating the room. Final mix is fast compared to pre-dubs because the shape of every element is already locked. Pre-dubs are sculpture; final mix is balance.

Print Master and Stem Delivery. The final mix produces the print master—the finished mixed soundtrack in its delivery format (stereo, 5.1, 7.1, or Atmos). Alongside it, the dub stage prints stems: dialogue stem, music stem, effects stem, sometimes Foley as a separate stem. Stems are the deliverable for international dubbing (M&E), promotional spots, alternate cuts, and downstream re-mixes. Lose a stem, and the title cannot be released in another language without a remix.

Conform. The reality of post-production is that picture sometimes changes after picture lock—a producer note, a test screening, a legal issue requiring a recut. When this happens, the assistant editors export a new EDL or AAF, and audio post must conform the existing mix to the new cut: re-spotting effects, re-aligning dialogue, re-cutting music to fit the new pacing. Conform is unglamorous, expensive, and often unavoidable. Build your sessions with handles (120 frames minimum, 240 preferred, of audio padding around every cut) so conforming is a one-day job instead of a one-week disaster.

Here is the reality of production audio: it is almost never clean. Air conditioners hum. Traffic bleeds through walls. An actor's lapel mic rubs against their shirt every time they move. The dialogue editor's job is to take all of that and make it sound like it was recorded in a perfect, silent room—without the audience ever noticing the work.

A core skill is working with room tone (also called ambience or fill). Every location has its own ambient sound. When you edit dialogue and remove sections of audio, the gaps between lines sound unnaturally silent. You fill these gaps with matching room tone recorded on set, creating seamless edits. Without proper room tone fills, the audience hears jarring jumps between silence and ambience—and they may not know why something sounds wrong, but they will feel it.

iZotope RX is the industry-standard audio repair suite for post-production, and it is not close. RX combines visual spectral editing with AI-powered tools including Dialogue Isolate (separates speech from background noise using machine learning), Spectral De-noise, De-reverb, De-click, De-clip, De-hum, and a Repair Assistant that intelligently identifies and addresses problems. RX works as both a standalone editor and a plugin within Pro Tools. Its ARA2 integration allows non-destructive spectral editing directly on the Pro Tools timeline—you can visually identify and paint away unwanted sounds without leaving your session. If you work in post-production, you will use RX. Learn it.

Other noise reduction tools: Waves offers several useful options—WNS Noise Suppressor for real-time dialogue cleanup, NS1 for single-fader intelligent noise suppression, and X-Click for restoring vinyl recordings. The principle behind all of these tools is the same: identify what is noise and what is signal, then attenuate the noise without destroying the signal. The better the tool, the more transparently it does this.

Field Recording

“I didn't want to go in there and activate the space myself. I wanted to go in there and listen to what was already there.” —Hildur Guðnadóttir on recording the score for Chernobyl at the Lithuanian power plant, Pop Disciple, August 2019

When Guðnadóttir scored HBO's Chernobyl, she did not write a single note on a piano. She traveled to a real decommissioned power plant in Lithuania, recorded hours of ambient sound from the building itself, and built the entire Emmy-winning score from those recordings. The hum of dead transformers became the strings. The metallic creak of forgotten machinery became the percussion. That is the power of field recording at its most committed—using real-world sound to tell a story no synthesizer could.

Every sound in a film started somewhere real. The thunder in a horror movie might be a recording from a storm in Oklahoma. The footsteps in a period drama might be leather shoes on cobblestone captured in an alley in Prague. Field recording—capturing sounds on location using portable recorders—is where a massive amount of post-production source material comes from.

Field recordists use devices like the Sound Devices MixPre series or Zoom F-series to capture production dialogue, ambience, and sound effects at high quality (typically 24-bit, 48 kHz or higher). These recordings get imported into Pro Tools for editing, cleaning, and mixing into the final soundtrack. Here is a career tip most textbooks will not tell you: build your own sound effects library. Every time you are somewhere interesting—a train station, a rainstorm, a construction site, a quiet forest—record it. Over the years, that library becomes one of your most valuable professional assets. The sounds you own are the sounds you never have to license.

That tip deserves the technique to back it up. Here is the craft of field recording, compressed to what actually matters on location.

Gear, in three tiers. The recorder you have beats the recorder at home: a phone with a clip-on mic captures usable scratch material and trains your ear for what is worth returning to. The working tier is a handheld recorder with a built-in stereo pair (the Zoom H-series; the H3-VR adds the ambisonic full-sphere capture from Chapter 6, which lets you re-aim the recording in post). The professional tier is the MixPre or F-series with external mics—a shotgun for isolated effects, a stereo or ambisonic pair for ambiences. At every tier, the non-negotiable accessory is wind protection: the foam ball that came with the mic is a studio pop filter, not a wind solution; outdoors you need a furry windjammer or a full blimp, because wind does not sound like wind on a recording—it sounds like ruined takes.

Location strategy. Scout with your ears, not your eyes: stand still for thirty seconds and inventory everything making noise before you unpack. Time of day is your biggest noise-floor control—dawn for nature beds before traffic wakes up, late night for room tones and cityscapes. Get close to the subject: the inverse-square law (Chapter 2) is the best noise gate you own, because every halving of distance buys you 6 dB of subject over background. And record longer than feels necessary—sixty to ninety seconds minimum for an effect, three to five minutes for an ambience you may need to loop. The take always ends right before the sound you actually wanted.

Noise management, in order. Kill what you can (HVAC, refrigerators, your own phone—airplane mode, always). Wait out what you cannot (let the plane pass; the recorder is already rolling, which is why you roll early). Aim what remains: point the mic's null—the dead side of the pattern (Chapter 5)—at the noise you cannot remove. What survives all three becomes a decision: re-schedule, relocate, or accept and plan to clean it with RX later.

Field discipline. Slate every take by voice—what, where, when—because three months later, two hundred files named ZOOM0047.WAV are a landfill, not a library. Record a safety take of anything you cannot repeat. Capture thirty to sixty seconds of clean ambience bed at every location before you leave; that is the room tone (covered earlier in this chapter) your future dialogue edit will beg for. Run 32-bit float if your recorder offers it, conservative levels if it does not—transients in the field are wilder than anything in a studio. And mark your keepers in the moment, while you still remember which thunderclap was the good one.

Library hygiene. A library you cannot search is a hard drive, not an asset. Name files by category, subject, and location at import time, and embed metadata while the memory is fresh; if you want the professional standard, the Universal Category System (UCS) is the freely published naming scheme most commercial effects libraries and search tools organize around. The habit costs minutes per session and compounds for the rest of your career—which is the same lesson as the session-naming discipline in Chapter 14, applied to the world outside the studio.

Production Sound for Picture

It is 6 AM on a film set. The first assistant director calls “quiet on set.” Thirty crew members freeze. The production sound mixer lifts a hand: “Speed.” The boom operator raises a Sennheiser MKH 416 above the actor's eyeline, the slate claps, and the director calls action. In the next thirty seconds, every word the actor speaks must land on a recorder with enough clarity that a dialogue editor—possibly months later, in a different city—can cut it into a feature film. Field recording (see above) is about capturing sounds you choose. Production sound is about capturing dialogue under conditions you do not control.

Double-system sound — recording audio and picture on completely separate devices. The camera records a reference scratch track; the production sound mixer's field recorder captures the full-quality audio. In post, the two are synced using timecode and the clapper slate. Every professional film and television production uses double-system sound because a camera's built-in preamps and media cannot match the dynamic range, track count, or reliability of a dedicated audio recorder.

The production sound mixer's bag rig is the on-set recording system: a field mixer-recorder worn in a harness bag, fed by wired and wireless microphone inputs, generating timecode, and recording every take to redundant media. The current standard is the Sound Devices 888 — 8 preamps, 16 channels, 20 tracks, internal 256 GB SSD plus dual SD cards, 32-bit A/D with 32-bit-float recording, Dante networking — compact enough for a bag, powerful enough for a multi-camera episodic set. (The Sound Devices Scorpio handles larger cart-based sound packages with 16 mic preamps and 32 channels.) Wireless lav systems ride in the same bag, their receivers strapped to the outside or mounted inside pockets, with RF antennas on short extenders for line-of-sight to transmitters on the talent.

Wireless Lavalier Systems

A lavalier (“lav”) is a miniature omnidirectional condenser microphone, typically 3–5 mm in diameter, clipped or taped to clothing near the actor's chest. The four wireless systems you will see on professional sets today:

• Lectrosonics SMWB/SMDWB + SRc — the North American touring standard. The SMWB is a single-AA Digital Hybrid Wireless^® beltpack transmitter; the SMDWB adds a second battery for double runtime. Both tune across 76.8 MHz of UHF spectrum in 25 kHz steps (3072 frequencies per band, bands A1–C1 plus a 941 MHz Part-74 band). The SRc dual-channel slot-mount receiver rides camera-side or in the bag. Digital Hybrid Wireless encodes audio digitally before transmission, decodes on receive, with no compandor artifacts.
• Wisycom MTP61 + MCR54 — the European narrative and EBU broadcast standard, increasingly common on US streaming productions. The MTP61 is the world's smallest multiband UHF bodypack transmitter. The MCR54 is a four-channel modular diversity receiver.
• Sennheiser EW-DP — a digital UHF system targeted at smaller crews and run-and-gun documentary work. The EW-DP EK receiver mounts directly to a camera cold shoe via a magnetic cheese plate; the SK bodypack transmitter pairs with ME 2 (omni) or ME 4 (cardioid) lavaliers. 134 dB dynamic range, 56 MHz bandwidth, exchangeable batteries.
• RØDE Wireless PRO — a 2.4 GHz dual-channel system for documentary and content production. The transmitters include 32-bit float on-board recording (backup audio to microSD card regardless of RF) and a built-in SMPTE timecode generator for double-system sync without a separate timecode box. Series IV digital transmission with 128-bit encryption and up to 260 m line-of-sight range. (RØDE also announced the RØDELink II—a professional UHF system built on the UHF engineering of Lectrosonics (which RØDE's parent acquired in 2025), with 32-bit float on-board recording and a dedicated timecode I/O port.)

Hiding Lavs and Managing Clothing Noise

Placing a microphone on an actor's body without the audience ever seeing or hearing it is a craft skill. Topstick double-sided tape holds the mic capsule to skin or fabric and simultaneously tacks the overlying clothing against the mic, preventing the fabric from sliding across it—that sliding is the source of most clothing rustle. Moleskin (the foot-care material, cut to roughly 25×50 mm rectangles) wraps around the capsule as a fabric isolation cradle before it is taped down; the felt surface absorbs friction instead of transmitting it as noise. For scenes with heavy clothing layers—leather jackets, suits, thick knitwear—a plant mic is the alternative: a lav taped to a prop, a piece of set furniture, or a camera rig in the scene rather than the actor's body. Plant mics sacrifice proximity and level for freedom of movement; the dialogue editor blends them with boom at the mix stage.

Boom Operation

The boom operator holds a 3–5 meter carbon-fiber pole with a directional shotgun microphone at the end—most commonly the Sennheiser MKH 416 (supercardioid/lobar, extremely popular outdoors for off-axis rejection) or the Schoeps MK 41 (supercardioid, preferred in controlled interiors for its natural off-axis character). The boom is positioned above the actor's eyeline, microphone angled downward toward the mouth at roughly 45–60 degrees, as close as the camera frame allows without dipping into shot.

Two non-negotiable rules govern every boom shot: (1) never block a light source. A boom pole and mic between a practical and your actor casts a shadow on the wall or face. Boom operators watch the camera monitor constantly and coordinate with the gaffer before the boom position is locked. When overhead creates a shadow problem, the boom comes from below—aimed upward at the actor's chin. (2) match perspective to the lens. A tight close-up tolerates a boom six inches from the frame line; a wide master shot demands the mic step back, which costs presence. The production sound mixer tells the director when a shot is forcing the mic so far back that usable dialogue is unlikely—the director's call is whether to accept the compromise or plan an ADR line.

Timecode Jam-Sync

Because audio and camera record on separate devices (double-system), both must share an identical clock reference so that any frame of picture corresponds to exactly one moment in the audio file. This is timecode jam-sync: one master timecode generator “jams” its SMPTE clock into every other device at the start of the day, and all devices free-run independently from that locked reference.

The Tentacle Sync E Mk2 is the current field standard for jam-sync. It is a Bluetooth-configured unit the size of a large USB stick, with a 3.5 mm LEMO-compatible LTC output, a 35-hour built-in battery, and support for all SMPTE rates. At the top of the day, the sound mixer jams the 888's internal clock from a master Tentacle, then jams a Tentacle into each camera's timecode input. All devices now share a common SMPTE address—when picture and sound are ingested in post, they align automatically by timecode. The jitter between devices after a full 12-hour day of free-running is typically less than one frame at 24 fps. Some wireless systems eliminate the external box entirely: the RØDE Wireless PRO and the incoming RØDELink II have internal timecode generators that jam directly from the receiver.

The Sound Report and Handoff to Post

At the end of each day, the production sound mixer prepares a sound report (also called a sound log)—a document listing every scene, take, track count, sample rate, frame rate, media card serial number, and any notes about problem takes or preferred takes (“circle takes” or “print takes”). The report travels with the media cards to the digital imaging technician (DIT) and, from there, to post-production.

The post house ingests the audio BWF / WAV files, reads their embedded timecode metadata, and auto-syncs them to the corresponding camera rolls in Premiere, Media Composer, or DaVinci Resolve. Once picture is locked, the picture editor exports an AAF—with handles—to the dialogue editor in Pro Tools. That AAF carries the production audio already in sync on its tracks. The dialogue editor rebuilds the timeline, routes individual takes to their own tracks, and begins the repair-and-cleanup pass with iZotope RX (see the Dialogue Editing section, above). The cleaned dialogue session is returned as an AAF for the re-recording mixer on the dub stage.

The chain is: sound report + BWF timecode files arrow sync in Resolve/MC/Premiere arrow picture-lock AAF to Pro Tools dialogue editor arrow RX cleanup arrow AAF roundtrip to the dub stage. Miss any link—wrong frame rate, no handles, missing timecode, no sound report—and the handoff stalls. Spend five minutes filling out the sound report. It saves the post house five hours.

The first time you hear a properly mixed Dolby Atmos track on a calibrated system, it rewires how you think about sound. A helicopter does not just pan from left to right—it lifts off from behind you, sweeps overhead, and disappears into the distance in front. Rain does not come from speakers—it falls from above. Once you hear what three-dimensional audio can do, stereo starts to feel like looking at the world through a keyhole.

Channel-Based Surround: 5.1 and 7.1

Traditional surround sound is channel-based—each speaker receives a dedicated audio feed. A 5.1 system uses six speakers: Front Left, Center, Front Right, Left Surround, Right Surround, and a Subwoofer (the “.1”). Configure your DAW and DA converters to route to all six outputs. If you only have a stereo speaker system, plugins like Waves UM225/UM226 can upconvert a stereo mix to 5.1—not ideal, but functional when a full surround rig is out of reach.

Object-Based Audio: Dolby Atmos

Dolby Atmos changed the game. Instead of assigning audio to fixed speaker channels, Atmos treats individual sounds as “objects” with metadata describing their position and movement in three-dimensional space. The Atmos renderer translates those positions to whatever speaker configuration the listener has—a 7.1.4 theater, a soundbar, or headphones using binaural rendering. The notation describes ear-level channels.subwoofers.height channels: a 7.1.4 system has seven ear-level speakers, one subwoofer, and four overhead speakers.

Pro Tools Studio and Ultimate include an integrated Dolby Atmos Renderer, so you can mix, monitor, and render Atmos content directly within a single session—no external rendering application required. The workflow: create a 7.1.2 main output bus (7.1.2 is the maximum bed format—route overhead-rich material through objects), enable the Atmos Renderer window, and assign audio objects with position metadata using the Object Panner. You do not need a 7.1.4 speaker array to start learning—Atmos can be monitored on headphones using binaural rendering, making it accessible even in home studios.

Apple Music launched Spatial Audio with Dolby Atmos support in 2021, and it has rapidly become a priority for major labels and independents alike. Mixing music in Atmos is fundamentally different from stereo—you are thinking in three dimensions, not two. Where a stereo mix pans left and right, an Atmos mix places instruments above, below, behind, and around the listener. It is a different skill, and the engineers who develop it now will have a significant career advantage.

For music delivery specifically, the full Atmos workflow—beds vs. objects, the integrated Pro Tools renderer, ADM BWF master export—was walked through in Chapter 18 (Pro Tools: Mixing). Treat film/television Atmos and music Atmos as related but distinct disciplines: the LUFS targets, dynamic-range expectations, monitoring formats, and delivery formats differ by platform and by use case. A film Atmos mix at −27 LKFS dialog-gated is built differently than a music Atmos mix at −18 LUFS integrated.

Pro Tools also supports Sony 360 Reality Audio (object-based spatial audio now carried mainly by Amazon Music, after Tidal and Deezer dropped the format) and Audio Vivid for immersive workflows.

Eclipsa Audio and Open-Source Spatial Formats

In January 2025, Samsung and Google unveiled Eclipsa Audio—a royalty-free, open-source 3D audio format designed to compete with Dolby Atmos. Built on the IAMF (Immersive Audio Model and Formats) specification developed by the Alliance for Open Media (whose members include Netflix, Apple, Meta, Amazon, and Microsoft), Eclipsa Audio supports up to 28 input channels, binaural rendering for headphones, and works with standard codecs including LPCM, AAC, FLAC, and Opus. YouTube began accepting Eclipsa Audio uploads in 2025, and Android 16 includes OS-level support. In Pro Tools, Eclipsa is handled through a free open-source plugin Google released in 2025—not a built-in (native) format.

The key difference is economic: Dolby Atmos requires licensing fees for both content creators and hardware manufacturers. Eclipsa Audio is free. Whether Eclipsa will achieve the adoption necessary to challenge Atmos remains to be seen, but as an audio professional, you should understand both formats and be ready to work in either.

You can mix the most beautiful soundtrack ever recorded, and if you deliver it in the wrong format, at the wrong loudness, with the wrong codec, the client sends it back. Delivery is not glamorous. It is non-negotiable.

Broadcast Loudness Standards

When mixing for broadcast television and film, you must comply with loudness standards—not as a suggestion, but as a legal requirement. In the United States, the ATSC A/85 standard (enforced by the CALM Act of 2010) requires dialogue level to average approximately −24 LKFS (Loudness K-weighted relative to Full Scale, equivalent to LUFS). In Europe, the EBU R128 standard targets −23 LUFS. These standards exist because audiences were tired of commercials blasting 15 dB louder than the show they were watching. Congress literally passed a law about it.

Use a loudness meter that supports these standards—Avid Pro Limiter, FabFilter Pro-L, or Waves WLM Plus—and check your mix before delivery. Getting rejected for a loudness violation after you thought the project was done is a mistake you only make once.

Picture Lock

Picture lock — means the video editor has finalized all cuts, transitions, and timing—no more changes to the picture. Audio work should not begin in earnest until picture lock is confirmed. If you start editing dialogue, spotting Foley, and composing score to a cut that is still changing, you will redo work every time the editor moves a scene, shortens a shot, or reorders a sequence. I have seen students burn entire weekends on audio that became useless because the picture changed Monday morning. Always confirm: “Is the picture locked?” If the answer is anything other than “yes,” wait.

Exporting and Deliverables

When your mix is complete, highlight the section to export and go to File → Bounce to → Movie. Select sample rate, source, video codec, and conversion quality, then click Bounce. Pro Tools exports MOV files using DNxHD, DNxHR, Apple ProRes, and H.264 codecs. For audio-only delivery, bounce to disk as discussed in previous chapters.

For professional film deliverables, the re-recording mixer creates a print master—the final mixed audio accompanying the picture. A standard delivery package includes:

• The stereo or surround print master
• Dialogue, music, and effects (DME) stems
• An M&E (Music and Effects) mix—everything except dialogue, used for international dubbing
• Immersive audio files if applicable (Dolby Atmos ADM BWF)
• Lo/Ro and Lt/Rt downmixes—“Left only/Right only” is a simple stereo downmix from the surround master; “Left total/Right total” is matrix-encoded for legacy Pro Logic decoding. Both are required deliverables on most broadcast and theatrical specs.

Modern Master Formats: IMF and DCP

Two delivery formats dominate professional post-production today, and a Paramount, Netflix, or Disney spec sheet will name them by acronym.

IMF (Interoperable Master Format) is the streaming-era replacement for the print master. An IMF package contains the picture, all audio masters (stereo, 5.1, 7.1, Atmos), all language tracks, all subtitle tracks, and metadata—in a single component-based deliverable that streamers can repackage for any output. Netflix, Disney+, and Apple TV+ all require IMF for original content delivery. Built on the SMPTE ST 2067 standard.

DCP (Digital Cinema Package) is the theatrical deliverable—a JPEG 2000 picture stream with PCM 24-bit/48 kHz audio (typically 5.1 or 7.1) in a sealed package conforming to SMPTE ST 429 (the Digital Cinema Package standard). Atmos delivery to theaters uses a DCP plus a separate Atmos sidecar file. If your project is going to theaters, the deliverable is a DCP, not a ProRes file.

Streamer QC Frameworks

Every major streaming platform publishes a technical delivery specification, and they have automated QC systems that reject masters that fail the spec. The most-cited framework is the Netflix Originals Delivery Specification (NOSD), used for Netflix originals and widely referenced by other platforms. Key requirements you will see across most spec sheets:

• Integrated Loudness: −27 LKFS dialog-gated for all near-field content (Netflix), with a ±2 LU tolerance; −24 LKFS program-gated is the fallback only for low-dialogue programs (dialogue under roughly 15% of the runtime). Other streamers vary by ±2 LKFS.
• Loudness Range (LRA): typically capped at 18 LU for film, lower for episodic. LRA measures the spread between quiet and loud passages per ITU-R BS.1770 (ITU-R BS.1770-5).
• Maximum True Peak: −2.0 dBTP standard for streaming, sometimes −1.0 dBTP. Anything above gets rejected.
• Dialogue centering: dialogue must be anchored in the center channel of the surround mix—no panning the lead character's voice across speakers.
• File specs: 24-bit PCM, 48 kHz, no clipping, BWF metadata correctly populated.

Cinema Monitoring Standards

If you ever step onto a theatrical dub stage, you are mixing on a system calibrated to a published standard. Each screen channel (L, C, R) is calibrated to 85 dB SPL at the reference position with pink noise at −20 dBFS RMS, and each surround array (Ls, Rs) sits 3 dB lower—the standard cinema convention (SMPTE ST 2095-1 later refined the reference noise to −18 dBFS RMS to land precisely on 85 dB). The LFE (subwoofer) is calibrated 10 dB hotter for its band-limited content. Above 2 kHz, theatrical playback uses the X-Curve (a high-frequency rolloff curve) so that a movie mixed on the dub stage translates correctly through the theater's house speakers. The signal path from the cinema processor's main-fader input through EQ, amplifiers, crossovers, and speakers (plus the room itself) is the B-chain; the playback server feeding it is part of the A-chain. The dub stage's monitor system is calibrated to match B-chain playback. Mixing for theaters without understanding B-chain calibration is mixing blind.

The chart below is the snapshot I keep at hand—each major streamer with its loudness target, true-peak ceiling, and master format requirement. Spec sheets evolve; always verify the current numbers at the platform's portal before final delivery. But these are the working defaults:

Always confirm delivery specifications with the distributor or post-production supervisor before bouncing. Codec, sample rate, bit depth, loudness target, LRA cap, true-peak ceiling, downmix flavor, and file naming conventions vary by platform and by year. The numbers above are working defaults—the spec sheet is the law. Get the spec sheet first. Bounce second.

Everything up to this point has been about working to picture—film, television, and broadcast. But the skills you have just learned do not stop at the mix stage. Dialogue editing, noise reduction, loudness standards, remote collaboration—these are the same skills driving the fastest-growing corners of the audio industry. Podcasting alone is a billion-dollar market, and it runs on the same signal chain and the same ear you have been developing since Chapter 1.

“All of us who do creative work, we get into it because we have good taste. But there is this gap. For the first couple years you make stuff, and it's just not that good… The most important thing you can do is do a lot of work.”

—Ira Glass, This American Life, 2009 video interview

Glass's “taste gap” is the truest thing said about audio storytelling, and it applies more brutally to podcasting than to any other medium. Anyone with a USB microphone can start a podcast. Almost no one makes a great one. The difference is volume and iteration—and the technical foundation that lets you focus on the story instead of the gear.

It is 7 AM. A host sits in a converted closet lined with moving blankets, a Shure SM7B six inches from their face, a cup of coffee just out of arm's reach. On the other side of a Zoom call, a guest is in their home office with AirPods and a MacBook. In ninety minutes, this conversation will be downloaded by 50,000 people. This is podcasting today—and it is the fastest-growing segment of the audio industry.

More than 150 million Americans listen to podcasts every month. Schools are adding dedicated podcast production courses. And yet it rarely gets the serious, end-to-end treatment it deserves in an audio-engineering textbook.

Room, Mic, and Signal Chain

There is a reason Joe Rogan records on a Shure SM7B in a treated room and not on an AKG C414 in his living room. Podcast recording is about voice clarity in imperfect spaces, and dynamic microphones win that fight every time.

The SM7B, the Electro-Voice RE20, and the Rode PodMic are the industry standards for podcasting because they reject room noise and off-axis sound far better than condensers. When your “studio” is a spare bedroom with drywall and a ceiling fan, that rejection is everything. A condenser in the same room picks up every reflection, every HVAC hum, every car driving by outside. A dynamic mic six inches from your mouth picks up your voice and almost nothing else.

The ideal podcast recording space is small, dry, and well-treated. A closet lined with moving blankets will sound better than an untreated living room with hardwood floors—that is not an exaggeration. Position the mic 4–8 inches from the speaker's mouth, slightly off-axis to reduce plosives. If you are building a podcast setup for a client, start with the room treatment before you spend a dollar on gear. No microphone fixes a bad room.

Editing Workflow

Podcast editing is not glamorous, but it is where good shows become great ones. The host says “um” forty times. The guest's dog barks in the background. One speaker is 6 dB louder than the other. Your job is to make it sound like two people had a smooth, professional conversation—even when the raw audio tells a different story.

A typical workflow: import all tracks, sync if recorded separately, use Strip Silence to remove dead air, apply clip gain to level-match speakers so nobody is louder or quieter than anyone else. Crossfade edits so cuts are invisible. Remove filler words where they distract, but leave enough so the speaker sounds human—over-editing makes people sound like robots.

For processing: high-pass EQ at 80–100 Hz to kill rumble, a gentle presence boost around 3–5 kHz to add clarity, compression at 2:1 or 3:1 to even out dynamics, and de-essing to tame sibilance. A noise gate or expander between phrases keeps the silence clean. On the master bus, a limiter targeting −16 LUFS brings everything to the right loudness. This chain is not fancy. It works on every podcast, every time.

Loudness Standards for Podcasts

Listen to three random podcasts on Spotify back to back. One is quiet. One is loud. One clips. Every time the listener switches episodes, they reach for the volume knob. That is a failure of loudness consistency, and it is the fastest way to lose an audience.

Apple Podcasts targets −16 LUFS. Spotify normalizes podcasts to −14 LUFS. The general standard: −16 LUFS with a true peak ceiling of −1 dBTP. Hit these numbers, and hit them the same way every episode. A listener who has to adjust their volume every week will find a different show. Consistency is not just professional—it is survival.

Multi-Guest Recording

I will say this once: never rely solely on Zoom audio. Zoom compresses the audio, mixes all participants to a single file, and if one person's connection drops for even a second, everyone's audio glitches. I have seen engineers lose entire interviews this way.

Always record each participant on a separate track. Use software that captures isolated audio—Riverside.fm, SquadCast, and Zencastr all record locally on each participant's machine and upload lossless files after the session. Alternatively, have each guest record on their own device and send the file afterward. This is called the “double-ender” method, and it has been the standard in broadcast radio for decades.

The rule: record locally, always. The internet is a convenience for monitoring, not a substitute for proper recording. And always—on every session, without exception—record a local backup on your end too.

The episode is mastered—−16 LUFS integrated, −1 dBTP ceiling, every speaker level-matched. Now it has to reach an audience, and this is where podcasting inverts everything Chapter 21 says about music distribution. A song is pushed to every platform as a copy. A podcast is published to exactly one place—and the entire ecosystem comes to it.

RSS: You Own the Feed

RSS — an XML document, generated and updated by your hosting provider, listing every episode with its title, description, artwork, and the address of the audio file. Every podcast app works the same way: it subscribes to the feed, checks it for changes, and pulls new episodes when they appear. Apple Podcasts, Spotify, and YouTube Music do not host your show—they read your feed and display what it declares. That is the most important fact in this section: the feed is the product, and you own it. One feed serves every app. Change hosting companies and a properly executed redirect carries every subscriber to the new feed without any of them noticing. No music platform offers an artist anything close to that portability.

Hosting and Directories

A podcast host stores the audio, generates the feed, serves every download, and reports listener statistics. Buzzsprout, Libsyn (the oldest name in the space), Transistor, and Spotify for Creators (free—the platform formerly called Anchor) are the names you will meet first; plans and features shift constantly, so compare current offerings rather than trusting any book's snapshot. What you are buying is delivery that survives a viral spike plus download numbers certified to the IAB measurement guidelines—the statistics sponsors actually accept. Hosting episodes on your own website provides neither.

A directory is where listeners find the show. Apple Podcasts Connect (podcastsconnect.apple.com), Spotify, and YouTube Music are the three that matter most today, and each works the same way: you submit the feed URL exactly once, and every future episode appears automatically the moment your host updates the feed. Submit once; RSS updates forever.

Episode Craft

The per-episode discipline is the same one Chapter 19 applied to masters. The MP3 carries its ID3 tags—the same TIT2/TPE1/TDRC frames from the mastering chapter, plus episode artwork. The feed carries the show notes: the description the directories display, with guest names, links, and timestamps, written for a listener deciding whether to press play. Chapter markers embed named skip-points—with per-chapter art and links in apps that support them—so a forty-minute interview becomes navigable. Artwork is its own spec: 3000×3000 pixels, square, Apple's ceiling and the working standard everywhere, designed to read at thumbnail size because a search result is where every future subscriber meets the show first. And title episodes by their content, not their number—“Episode 47” tells a search algorithm nothing.

Monetization, Honestly

Sponsorship money arrives two ways. Host-read sponsorship — the host reads the ad in their own voice, and it is baked into the episode file forever. Rates are quoted as CPM—cost per thousand downloads—and advertisers pay a premium for the host's credibility with the audience. Dynamic ad insertion (DAI) — flips the model: the episode on the host's server carries timestamp markers, and ads are stitched in server-side at download time. Different listeners hear different ads, campaigns are sold and swapped without touching the episode, and—the real shift—the back catalog becomes inventory: a two-year-old episode downloaded today serves today's campaign. The honest trade: DAI scales, host-read converts. The third lane skips advertisers entirely—listener support through Apple Podcasts subscriptions and Patreon-class membership tiers, typically sold as bonus episodes and ad-free versions delivered through private RSS feeds, the same mechanism again, scoped to one paying subscriber. And the honesty the gear ads will not give you: none of this pays at episode three. Sponsors do not buy potential; they buy a download number that shows up every week. Build the number first.

As of this writing, the biggest distribution shift is video: YouTube and Spotify both carry full video podcasts, YouTube has become a primary discovery engine for shows at every scale (the next section quotes the numbers), and Spotify ingests video through Spotify for Creators. The RSS feed, though, does not carry your video—video versions are uploaded natively to each platform while the audio feed remains the canonical product. The production side—multicam sync, live streaming, short-form clips—is exactly what the next section covers.

Here is the part to internalize: distribution is a Tuesday-afternoon task. Pick a host, upload the master, submit the feed to the directories once, and the pipeline runs itself from then on. What cannot be automated is what the loudness section already told you—the same numbers, the same day, the same quality, every week. A listener does not subscribe to an episode; they subscribe to a feed, and a feed is a promise. Consistency is the product.

Podcasting is no longer audio-only. Over 51% of Americans have watched a podcast, and YouTube reports over one billion monthly podcast viewers. The biggest shows—Rogan, Lex Fridman, Call Her Daddy—are as much video as audio now. If you are setting up a podcast studio for a client today and you are not planning for cameras, you are already behind.

Multicam sync is the first technical challenge. With multiple cameras rolling, you need every angle locked to the same audio timeline. The simplest method is a clap at the beginning of recording—same principle as a film slate, and it has worked for a hundred years. For more complex setups with three or four cameras, timecode generators like Tentacle Sync jam the same clock to every device. In post, tools like PluralEyes or DaVinci Resolve can auto-sync all your angles based on waveform analysis. The point is: get sync right at the source. Fixing it in post is possible but painful.

Live streaming adds another layer. OBS Studio (free, open-source) is the standard for routing audio and video to YouTube, Twitch, or other platforms in real time. You route your Pro Tools output into OBS as an audio source alongside your camera feeds. The audio quality you deliver live is the quality the audience hears—there is no fixing it later. Get your levels, your processing, and your monitoring right before you go live.

Short-form content is where the money is shifting. Nobody discovers a podcast by searching for it. They find a short clip on TikTok, Instagram Reels, or YouTube Shorts—formats that now run anywhere from seconds to several minutes—and if that clip hooks them, they subscribe. Many producers now create 5–10 clips from each episode—the best moments, captioned and formatted for vertical video. An engineer who can handle the audio and cut compelling short-form clips is worth their weight in gold. If you can do both, you will never be short on work.

It is 2 PM in Los Angeles. A singer is in a vocal booth. The producer is in Nashville. The mixing engineer is in London. The label executive is listening from a hotel room in Tokyo. All four are hearing the same audio, in real time, with professional fidelity. This is not a special occasion. This is a Tuesday.

The pandemic did not create remote collaboration—studios have been linking up over ISDN lines since the 1990s—but it made remote sessions normal instead of special. Before 2020, a client in Tokyo listening to a mix happening in Los Angeles was a big-budget luxury. Now it is a Tuesday afternoon. If you cannot run a remote session smoothly, you are leaving work on the table.

Source-Connect — has been the standard for remote recording since 2005, and for good reason. It transmits real-time, broadcast-quality audio between two or more studios over the internet, with built-in talkback, MIDI streaming, and a Q Manager that repairs lost packets after the fact so nothing is permanently lost. It supports multi-channel audio including surround formats. When a voiceover artist in New York needs to record a session directed by a producer in LA, Source-Connect is how it happens. When an actor does ADR from their home studio because they cannot fly to the mix stage, Source-Connect is how it happens. Learn it.

Audiomovers LISTENTO — solved a different problem: how do you let a client hear your mix in real time without making them install anything? You insert the LISTENTO plugin on your master bus, it generates a shareable link, and the client clicks it in their web browser. That is it. No downloads, no accounts, no tech support calls. It supports up to 32-bit PCM and multichannel formats including 7.1.4 and Dolby Atmos. Abbey Road Studios acquired Audiomovers in 2021—that should tell you something about how seriously the industry takes this tool.

Pro Tools Cloud Collaboration — takes a different approach entirely. Instead of streaming audio, it lets multiple Pro Tools users work on the same project through Avid's cloud infrastructure—sharing tracks, commits, and session data without being in the same room. Think of it like GitHub for audio sessions. A producer in Nashville adds a guitar overdub, commits it, and the mixing engineer in London sees the new track appear in their session.

Here is the non-negotiable advice for any remote session: use wired Ethernet. Not Wi-Fi. Not your phone's hotspot. A wired connection. Test it before the session starts, not when the client is sitting there waiting. And always—on every remote session, without exception—record a local backup on both ends. Internet connections fail at the worst possible moment. Your local recording is the only insurance policy that matters.

Now imagine doing everything you have learned in this book—EQ, compression, gain structure, signal flow, monitoring—live, in front of two thousand people, with no undo button. That is live sound. And it is one of the largest employers of audio professionals in the world.

Concerts, festivals, theaters, houses of worship, corporate events, sports arenas—all of them need skilled engineers. The core principles are identical to studio work. The difference is that every decision is permanent. You cannot punch in. You cannot comp takes. If the vocal is too quiet during the chorus, two thousand people heard it too quiet during the chorus. The pressure is real, and the engineers who thrive on it build incredible careers.

FOH (Front of House) — is the engineer who sits in the audience—usually on a platform at the back of the venue—and mixes what the crowd hears through the PA system. This is the person shaping the entire sonic experience of the show. The Monitor engineer sits at the side of the stage and mixes what the performers hear through their stage monitors or in-ear monitors. What the band needs to hear to play well is completely different from what the audience needs to hear to enjoy the show. On smaller gigs, one engineer handles both. On arena tours, these are two entirely separate consoles with two entirely separate engineers.

The signal flow follows the same logic you learned in the studio, just stretched across a much larger space: microphone → stage box or snake → console → processing (EQ, dynamics, effects) → amplifiers → speakers. Modern systems increasingly use Dante or other Audio over IP protocols to replace heavy analog snakes with a single Ethernet cable carrying dozens of channels. The gear changes. The signal flow does not.

Modern live consoles are the same DAW-style processing you learned in the studio, packaged into a control surface built for the road. Avid VENUE | S6L is the U.S. touring and festival standard—runs Pro Tools plugins natively. DiGiCo SD/Quantum series is the European and high-end international touring standard. Yamaha Rivage PM and the smaller CL/QL consoles dominate broadcast and Asian markets. Allen & Heath dLive is the mid-tier touring and houses-of-worship workhorse. All of them offer hundreds of input channels, on-board DSP, full snapshot recall, and—increasingly—direct multitrack recording to a USB drive for live mixes you can re-mix later in the studio.

System tuning is its own craft. Tools like Rational Acoustics Smaart and Meyer Sound SIM measure the PA's response in the actual room and reveal where speakers are over- or under-coupled, where reflections create null points, where EQ has to compensate before the doors open. Every major tour carries a system engineer whose entire job is making the PA respond consistently from the front row to the back wall. The mix engineer mixes the show; the system engineer makes the room ready for the mix to be heard.

Feedback — is the single biggest enemy in live sound—that piercing squeal that makes an entire room flinch. It happens when a microphone picks up its own amplified signal from a speaker, creating a loop that builds until it screams. Every live engineer learns to manage feedback through proper monitor placement, ringing out the system before the show (slowly raising gain on each mic until feedback frequencies reveal themselves, then notching them with narrow EQ cuts), and maintaining careful gain structure throughout the performance. A great FOH engineer can feel feedback building before the audience ever hears it.

Before a show, the band's management sends two critical documents: a stage plot—a diagram showing where every musician and piece of equipment sits on stage—and an input list detailing every microphone and DI channel, numbered and labeled. These go to the venue in advance so the house engineer can prepare patches, set up monitors, and have the system ready before the band arrives for soundcheck. If you ever work a venue gig, these documents are your blueprint. Without them, setup is chaos.

One of the most valuable skills that bridges studio and live work is live recording from FOH. Many engineers multitrack-record shows using a splitter that sends mic signals to both the live console and a separate recording rig. The resulting multitrack can be mixed later in the studio with all the tools, time, and control you are accustomed to—giving you the raw energy of a live performance with the precision of a studio mix. Some of the greatest live albums ever made started as a multitrack recorded off a splitter at FOH.

Live Multitrack Recording

I had 48 inputs on stage, two consoles, a recording rig parked in a side corridor, and about forty minutes of line check before doors. That is a normal Wednesday night in live sound—and the multitrack I walked out with that evening became a record that never would have happened without the split.

The Stage Splitter

The fundamental problem in capturing a live show is that you need the same microphone signals in at least three places simultaneously: the FOH console, the monitor console, and your recording rig. A stage splitter — a hardware device that distributes each mic input to multiple destinations without letting those destinations interact electrically with each other.

Passive transformer splitters use a Jensen-type isolation transformer on each secondary output to provide galvanic isolation—the two coils of wire transfer the audio signal without a direct electrical connection, which eliminates the ground loops that turn a live rig into a 60 Hz buzz machine. The Radial JS3 is the touring standard for single-channel passive splits: one XLR input, one direct (“thru”) output with phantom pass-through, and two Jensen-isolated secondary outputs. Rack the JS3 modules into a Radial J-Rak 8 and you have an 8-channel splitter in 2U. Passive splits add roughly 6 dB of insertion loss on the isolated outputs, which is trivial—but you must remember to feed phantom power from only one console, typically FOH, since the direct output passes phantom through while the isolated outputs block it.

Active digital splitters solve the problem differently. A digital console with a built-in stagebox (DiGiCo SD-Rack, Yamaha RIO, Allen & Heath DX168) outputs a full MADI or Dante stream that can be simultaneously received by FOH, monitors, and the recording rig with no transformer at all. The signal stays digital from the stagebox to every destination. Gain is set once, at the head-amp in the stagebox, and all three destinations see the same level. This matters more than it sounds.

Why the Recording Rig Needs Its Own Gain Structure

Here is the mistake I see from engineers new to live recording: they assume they can ride FOH channel faders to shape the recording. They cannot. The split happens at the microphone preamp—before any fader, any EQ, any dynamics. The recording rig captures whatever the mic sees at the preamp output, unprocessed. If the FOH engineer pulls back the kick because the room is loading up, the recording is not affected. If they push a vocal for a big chorus, the recording hears what the mic heard, not what the PA heard.

This is the correct architecture. You want the multitrack to be pre-fader, pre-processing captures. That is what gives you a full mix-from-scratch session back in the studio. But it also means your recording gains are completely independent: they have to be set conservatively enough that nothing clips on the unexpected moments—the drummer who decides to whale on the snare during line check, the guitarist who doubles his amp volume for the show because he changed the setlist. Live sound is full of surprises. 32-bit float recording on portable rigs like the Sound Devices 8-series or Zoom F8n Pro eliminates the clipping risk entirely by storing the full dynamic range of the A/D converter without a fixed clip ceiling—you can recover a 20 dB hot signal in post. On MADI or Dante feeds from a console, you are typically recording at 24-bit, so target a conservative −18 dBFS nominal and leave headroom for the unexpected.

Capture Paths: MADI, Dante, and Standalone Interfaces

MADI from the console is the traditional touring-truck solution. Every major live console offers MADI output: plug a single BNC coaxial or fiber cable into a MADI-equipped interface in your recording laptop, and you have up to 64 discrete channels at 48 kHz landing directly in Pro Tools, Reaper, or any MADI-capable DAW. A dedicated MADI interface (RME HDSPe MADI, TASCAM IF-MA64) bridges the console to your computer. MADI on fiber runs up to 2,000 meters—critical when the recording truck is in a loading dock three hallways from the stage. As covered in Chapter 8, MADI drops to 32 channels at 96 kHz; most live recording is done at 48 kHz where the full 64-channel bandwidth is available.

Dante from the console is increasingly how mid-tier and touring rigs capture. If the console already runs Dante (Yamaha CL/QL, Allen & Heath dLive, many Avid VENUE configurations), you subscribe to a Dante Virtual Soundcard license on your recording laptop, connect to the same managed Gigabit switch, and the Dante Controller routes whichever channels you need directly into your DAW as a standard audio interface. Dante Virtual Soundcard supports up to 64 bidirectional channels at 48 kHz on a standard license; the Pro subscription adds 128×128 channels and clock-leader support. Latency is as low as 4 ms on a properly configured switch—low enough that you can monitor in real time without audible delay. As covered in Chapter 8, Dante requires a managed switch with Quality of Service (QoS) configured; a consumer unmanaged switch at a venue will cause dropouts under load.

Portable standalone recorders are the right answer when you cannot tap a console output—or when you want a self-contained backup that runs regardless of whether the FOH engineer cooperates. The Sound Devices 888 records 16 input channels (eight mic/line preamps plus eight returns) simultaneously to an internal 256 GB SSD and dual SD cards, with 32-bit-float recording and sample rates up to 192 kHz. The smaller Sound Devices 833 handles 12-track recording with six high-gain preamps. For budget-conscious rigs, the Zoom F8n Pro captures 8 inputs and 10 tracks at 32-bit float, with dual SD card slots up to 1 TB each for redundant recording—at around $1,000 street, it is the workhorse of independent live recording.

The Virtual Soundcheck Workflow

Virtual soundcheck is the highest-leverage live-recording technique most engineers never use until someone explains it to them. The idea: record every channel during line check, before the audience arrives. Then, while the band is off stage eating dinner, play the multitrack back into the console inputs—replacing the live mics with the recorded line-check audio—and use that audio to tune the PA, build monitor mixes, and dial in the FOH mix without any musicians present. You can loop a thirty-second section of the drummer's check pattern for an hour while you ring out each monitor mix, solve feedback, and save full console snapshots. When the band returns for soundcheck, you have a working mix already loaded.

Harrison LiveTrax (now at version 3) is purpose-built for this workflow and integrates directly with DiGiCo, Allen & Heath dLive, and SSL Live consoles, including transport sync and scene-marker alignment that jumps playback to the correct moment when a console snapshot is recalled. Pro Tools and Reaper both handle the playback role if you route returns from your DAW back into the console's stagebox return channels—the workflow is DAW-agnostic; the console needs only to have its input source switchable between the stagebox and the return path.

Managing Stage Bleed and Unknown Venue Acoustics

Every live recording inherits the room—and unlike the studio, you cannot design it. The audience absorbs high frequencies unevenly depending on how full the room is. Low-end nodes shift depending on where people are standing. The overheads and room mics you might pin to the rafters capture not just the kit but every guitar amp, stage monitor, and crowd cough in the building. There is no fix for this in the moment: your job is to ensure every close mic is gated or managed tightly enough that bleed does not destroy your mix options in post. Drum overheads in a live context are often omitted from the multitrack entirely in favor of room mics; close-mic the kit aggressively and add ambience from a dedicated stereo room pair placed at FOH height, where you can hear the venue properly.

Gain staging for unknown rooms: set your recording levels to match the line-check peaks, add 6 dB of headroom beyond that for show dynamics, and do not touch the recording gains during the show. If you are on a 32-bit float recorder, this is academic—but if you are recording from a MADI feed at 24-bit, conservative gains protect the performance.

Wireless and RF Coordination

A multitrack live recording lives or dies by the IEM and wireless mic system. Every wireless channel—microphone, IEM, talkback—occupies spectrum in the UHF or 2.4 GHz band, and the venue adds its own RF environment on top: house LED dimmers, building Wi-Fi, neighboring events. RF coordination is the process of scanning the local spectrum before the show, identifying occupied frequencies, and assigning each wireless system to a clear, intermodulation-free slot. Shure Wireless Workbench (free download) performs this coordination for multi-manufacturer systems and exports frequency lists you can load directly into Shure, Sennheiser, and other wireless gear. Best practice: run a spectrum scan at the venue with all other production RF powered on—lighting dimmers are the worst offenders and must be dimmed at show levels during the scan, not at dark.

Live multitrack recording bridges the third engineering path—live sound—directly back into the studio, where the mix happens under controlled conditions with every tool you have learned in this book. It is one of the most direct ways to earn both a live-sound credit and a studio-mix credit on the same project, and on a good night with a great band, it is the most exciting session you will ever run.

A sword swings. A dragon roars. Footsteps crunch through snow. A distant bell tolls. Every sound you hear in a video game was designed, recorded, processed, and implemented by audio professionals—and the gaming industry is one of the largest employers of audio talent in the world.

Game audio is fundamentally different from music or film audio because it is interactive. In a film, the soundtrack is fixed—every audience member hears the same thing. In a game, audio must respond to the player's actions in real time. Walk into a cave and the reverb changes. Get injured and the music shifts to a minor key. Sprint and your footsteps speed up. This interactivity is managed through middleware—software that sits between the game engine and the audio assets.

The two dominant middleware platforms are Wwise (Audiokinetic) and FMOD (Firelight Technologies). Both allow sound designers to create complex audio behaviors—randomized variations so the same footstep never sounds identical twice, distance-based attenuation, interactive music systems, and real-time mixing—without writing code. Learning one of these tools is the gateway to a career in game audio.

A typical Wwise workflow: create a new project, import audio assets into SoundBanks (compressed audio containers loaded by the game engine), define Events (the triggers the game code calls—PlayerJump, EnemyDeath, DoorOpen), and assign audio to those events with optional randomization, distance attenuation, and effect chains. The game engine references events by name; when the player jumps, the code calls PlayerJump and Wwise plays a randomized footstep variation with reverb matching the current room. FMOD Studio works similarly with slightly different terminology and a more visual interface. Both tools offer free indie tiers (Wwise's free Indie license has unlimited sounds and is gated by total production budget—the 200-asset cap belongs to the non-commercial Free Trial, not the Indie license; FMOD's free Indie license is gated by both annual revenue (under $200K) and project budget (under $500K)). Wwise dominates AAA console development; FMOD is more common in indie and mobile games. Pick one, build a small game-audio demo, and you have a portfolio piece for the industry.

Adaptive music uses horizontal re-sequencing (rearranging musical sections based on gameplay) and vertical layering (adding or removing instrument layers based on intensity). A battle sequence might start with percussion only, add strings when enemies appear, add brass when the boss arrives, and fade to solo piano when the player is near death—all in real time, triggered by the game engine. The composer does not write one piece of music. They write a system.

Game audio also ships to a loudness spec, and the spec reads like broadcast: Sony's ASWG-R001 guideline targets console titles at approximately −24 LKFS—the same ITU-R BS.1770 arithmetic, and the same number, as US broadcast's ATSC A/85—Xbox guidance sits in the same territory, and mobile titles typically run hotter because they live on phone speakers and earbuds. Wwise includes loudness metering built in, so the target is checkable in-engine during implementation rather than discovered at platform certification. The number matters for the same reasons it does on television: players hold sessions for hours, where an over-crushed mix becomes fatigue, and a game that ignores the target lands jarringly loud—or quiet—next to every broadcast-calibrated source the living-room TV plays.

If game audio interests you, start here: take existing game footage and replace all the sounds from scratch. Learn the basics of Wwise or FMOD—both offer free versions. Build a game audio demo reel. The skills you have learned throughout this book—recording, editing, mixing, signal processing, spatial audio—all apply directly. The only new skill is interactivity, and middleware handles the heavy lifting.

AI in Game Audio

Imagine you are playtesting an open-world RPG with a hundred named NPCs, each scripted to deliver dozens of lines. Recording all of them with union voice actors would take months and cost a production budget most studios do not have—so the dialogue team drops a text file into an AI text-to-speech pipeline, generates scratch lines in hours, and ships the game to QA the same week. That is the job today.

AI-assisted dialogue and localization. Tools like Replica Studios and ElevenLabs have become standard in game development pipelines for generating temp VO (temporary voiceover used in early builds before final casting) and placeholder NPC dialogue. Both platforms offer SDKs that integrate directly with Unity and Unreal Engine, allowing lines to be generated at runtime from text—meaning an NPC can speak new dialogue the designers wrote that morning without a recording session. At scale, the same pipeline handles AI-driven localization: rather than re-recording a game in six languages, publishers synthesize localized lines from the original script and a target-language voice model, dramatically cutting turnaround time.

Procedural and generative sound design. Middleware vendors are integrating AI tooling directly into their workflows. Audiokinetic's Wwise introduced Similar Sound Search, a feature developed jointly with Sony AI that lets a sound designer describe a sound in plain text—or drop in a reference clip—and retrieve semantically matched assets from their sound library without relying on filenames or hand-tagged metadata. That is retrieval, not generation, but it compresses the asset-hunting phase that consumes hours of every SFX pass. On the synthesis side, Wwise's SoundSeed Grain plugin provides granular, pulsar, and concatenative synthesis inside the middleware session itself, enabling procedural sounds—engine drones, wind, water—that never repeat and draw almost no memory compared to a pre-rendered loop library.

Adaptive music generation. The horizontal re-sequencing and vertical layering systems covered above (see p. ) require composers to pre-write every stem. AI music generation tools—still maturing, but moving quickly—are beginning to let designers define parameters (tempo, key, instrumentation, emotional intensity) and generate stems on the fly, so an adaptive music system could, in principle, compose a unique underscore for each play session. This is not yet the norm in shipping titles, but it is the direction middleware vendors and game audio toolmakers are moving.

Ethics: voice-actor consent and SAG-AFTRA. AI voice replication raises the hardest question in this space: whose voice is it, and who gets paid? In January 2024, SAG-AFTRA negotiated a digital voice replica agreement with Replica Studios, the first of its kind: performers license their voice for AI-generated game dialogue under per-session minimum fees (built on the Interactive Media Agreement rate framework), with written consent required for every new project. The broader issue escalated into the SAG-AFTRA video game strike, which ran from July 2024 through mid-2025 over exactly this question—how to define, limit, and compensate AI replicas of union performers. The strike ended with the ratification of the 2025 Interactive Media Agreement, approved by 95% of voting members, which establishes consent and disclosure requirements for digital replicas, usage reporting obligations for producers, and the right of performers to suspend consent during labor actions.

The practical rule for any studio: you cannot create or deploy an AI replica of a named performer's voice without separate, written consent and compensation, regardless of how the lines are ultimately used. “Temp VO” built on a real actor's voice model—even scraped without a formal agreement—is an unfair labor practice under the 2025 IMA. Build your placeholder pipeline on non-union synthetic voices or purpose-built libraries, and bring real performers in for final characters.

Look at the names that defined cinema sound and you find a pattern. Walter Murch reinvented film sound on Apocalypse Now—helicopters that move through a 360-degree space because the audience needed to feel surrounded. Ben Burtt built the entire sonic universe of Star Wars from real-world recordings, projector motors, and the hum of live electronics. Randy Thom has spent four decades at Skywalker Sound shaping films from The Empire Strikes Back to The Wild Robot. Mark Mangini won an Oscar for Mad Max: Fury Road and another for Dune. Hildur Guðnadóttir scored Joker on cello and built Chernobyl from real ambient recordings of a decommissioned power plant. None of them started in post. They started where you started—microphones, signal flow, an interest in how sound makes a person feel something.

The skills you built across nine chapters of practical exercises—microphone choice, signal flow, EQ, dynamics, time-based effects, mixing, mastering, picture sync—are the same skills these careers were built on. Music chops give you the ear; the picture chops give you the eye. Together they open every door audio has to offer.

And it is not just picture. The same signal chain and the same ear run the podcast booth, the FOH console at a two-thousand-seat room, and the game-audio middleware session. Podcasting alone is a billion-dollar industry; live sound is one of the largest employers of audio professionals alive; the games industry now outgrosses film and music combined. The engineer who can move fluidly between all of them never runs out of work.

After more than twenty years in this work, here is what I will tell you about the engineer who can do both: the engineer who can only do music waits for music gigs. The engineer who can also do picture works every week of the year. Films come, podcasts come, live events come, games come—and the technical work for all of them is the work you have already learned. What is left is reps. Take a short film. Score thirty seconds. Foley a scene yourself. Mix a podcast for a friend. Run sound at a venue. Each project teaches you something the textbook cannot. The textbook gets you to the door. The reps get you through it.

You may think your work for this project is over, but the project is bigger than the project. Every record you cut, every score you place, every mix you bounce, every Foley pass you record—it is part of one career, and the career is built one project at a time. In the next chapters we cover the business side that protects all of it (Chapter 21) and the AI side that is reshaping all of it (Chapter 22). The technical foundation is yours. Now you build the career.