🚧 Under construction — VoTechWorks is being built. A free tech program — complimentary, as our commitment to get America working.
voTechWorksGrade 13 · a vocational technology learning center, by urbanicity
newSchoolWorks.com › VoTechWorks › Technology & Design › Graphic & Digital Media › Video Production & Motion Graphics › Lesson 7
TEC-DIGI47-05 · Lesson 7 · 10 clock hours · lab

Audio Mixing, Music, & Captions in Premiere Pro

Objective. Students will mix a three-track audio sequence (dialogue, music, and ambient sound) to broadcast loudness standards and add burned-in captions using Premiere Pro's Caption panel.

⚠ Pending professional review

This lesson, its worked examples and its practice questions has not yet been verified by an industry-certified professional or by a certification or approval body. It is a study draft and may contain errors. Do not use it as instructions for work on a job site or on live equipment. Always follow your instructor, your supervising professional, and current official references and regulations.

VoTechWorks program approval and accreditation are still in progress. Credentials are issued only by the certifying bodies named in each program, after their own exams.

Audio Mixing, Music, & Captions in Premiere Pro

Producing a finished video means more than assembling great footage. Audiences notice bad audio immediately—dialogue that is too quiet, music that drowns out words, or captions that lag behind speech. This lab walks you through the three tasks every professional editor must master: assigning audio roles, matching loudness to delivery standards, and creating accurate captions.

Why Loudness Standards Matter

Loudness is measured in LUFS (Loudness Units relative to Full Scale), a psychoacoustic measurement that reflects how human ears perceive average volume over time. Because different platforms set their own normalization targets, you must master your audio to the platform's spec before upload—otherwise the platform's algorithm will turn your audio up or down automatically, often with unpleasant side effects.

Common integrated loudness targets include approximately −14 LUFS for YouTube and −16 LUFS for many web/social platforms, while linear broadcast in the United States follows ATSC A/85, which targets −24 LKFS (a near-identical unit). Always check the current specification for the platform you are delivering to, as targets can change. For this lab, aim for integrated loudness between −18 LUFS and −14 LUFS, which covers most web and social delivery.

True Peak (the highest instantaneous peak after digital-to-analog conversion) should be kept at or below −1 dBTP to prevent inter-sample distortion on compressed codecs such as H.264.

The Essential Sound Panel

Premiere Pro's Essential Sound panel simplifies complex audio tasks by letting you assign a role to each clip or track: Dialogue, Music, Ambience, or Sound Effects. Once a role is assigned, the panel surfaces the most relevant controls for that type of audio.

Using the Loudness Meter

The Loudness Meter (or Loudness Radar, available as a free extension or built-in meter in some Premiere versions) displays integrated loudness, short-term loudness, momentary loudness, and true peak in real time as you play the sequence. To check integrated loudness:

1. Open the Loudness Meter panel and reset it.

2. Play the entire sequence from start to finish without stopping.

3. Read the Integrated value when playback ends.

4. Adjust the master audio track gain or the sequence master clip volume and repeat until the integrated reading falls within your target range.

Make small adjustments—1–2 dB at a time—and re-check rather than making large, uncorrected jumps.

Keyframing Music Volume

To ride music volume under dialogue, use keyframes on the clip's audio rubber band (the thin horizontal line on the audio clip in the timeline):

1. Hold Ctrl (Windows) or Cmd (Mac) and click the rubber band to add a keyframe.

2. Add keyframes just before and just after each section of dialogue.

3. Drag the rubber band downward between the inner keyframes to reduce music volume during speech.

4. Maintain a natural-sounding ramp by placing transition keyframes a few frames before the level change begins.

This technique is called ducking. The Essential Sound panel includes an automatic Duck against Dialogue checkbox for music and ambience tracks, which can save time on long edits.

Creating Captions in the Caption Panel

Captions provide accessibility for viewers who are deaf or hard of hearing and improve comprehension for viewers watching without sound. Burned-in captions (also called open captions) are permanently embedded in the video frame and do not require a separate subtitle file.

To create captions manually:

1. Open the Caption panel (Window > Captions and Graphics).

2. Create a new captions track and select the appropriate format (SubRip SRT or CEA-608 are common choices).

3. Play the timeline and use the + button to add a new cue at the exact timecode where dialogue begins.

4. Type the on-screen text for that cue and set the out-point where the dialogue ends.

5. Repeat for each line of dialogue. Aim for cues of roughly one to three seconds; avoid single cues that run more than two lines of text.

6. Review each cue by scrubbing the timeline to confirm the text appears and disappears in sync with speech.

For burned-in export, check Burn Captions into Video in the Export Settings dialog before rendering your H.264 MP4.

Safety and Hearing Health

Extended headphone use at high volumes causes permanent hearing damage. Keep monitoring levels comfortable—industry practice recommends keeping reference levels at or below 85 dB SPL for critical listening, with regular breaks. If you must raise headphone volume to hear details, fix the problem in the mix, not by increasing your monitoring level. Never use hearing protection that is not rated for your specific environment; in a studio setting, the primary hazard is sustained loud output, not impact noise.

Workflow Summary

1. Assign roles in Essential Sound → auto-match dialogue loudness.

2. Keyframe music to sit 12–18 dB below dialogue.

3. Play the full sequence through the Loudness Meter; adjust master gain to hit −18 to −14 LUFS integrated.

4. Add at least 10 caption cues manually; verify frame accuracy.

5. Export as H.264 MP4 with captions burned in.

Key terms10

LUFSLoudness Units relative to Full Scale; a psychoacoustic measurement of average perceived loudness used by streaming platforms and broadcasters to set delivery targets.
Integrated LoudnessThe average loudness of an entire audio program from start to finish, measured in LUFS; the primary metric platforms use when normalizing uploaded content.
True PeakThe highest instantaneous level a signal reaches after digital-to-analog conversion, including inter-sample peaks; kept at or below −1 dBTP to prevent distortion on compressed codecs.
Essential Sound PanelA Premiere Pro panel that assigns audio roles (Dialogue, Music, Ambience, Sound Effects) to clips and surfaces role-appropriate controls such as auto-match loudness and noise reduction.
DuckingThe technique of automatically or manually reducing the volume of music or ambience tracks during dialogue so speech remains clearly intelligible.
Keyframe (audio)A point on an audio clip's volume rubber band that stores a specific gain value at a specific timecode, allowing volume to change smoothly over time.
Open Captions (Burned-in Captions)Caption text permanently rendered into the video frame during export, visible on any player without requiring a separate subtitle file or player support.
CEA-608A widely supported closed caption format used in broadcast and streaming; one of the caption formats available in Premiere Pro's Caption panel.
Auto-Match LoudnessA feature in the Essential Sound panel's Dialogue section that analyzes each selected dialogue clip and applies a gain adjustment to bring all clips to a consistent loudness target.
Loudness MeterA metering tool in Premiere Pro (or a compatible plug-in) that displays integrated, short-term, and momentary loudness plus true peak in real time during playback.

Worked examples4

1. Your dialogue track auto-matched to −23 LUFS and your music track is currently at −23 LUFS as well. How many dB do you need to reduce the music to place it 15 dB below dialogue?
  1. Step 1 — Identify the dialogue level: −23 LUFS.
  2. Step 2 — The target music level is 15 dB below dialogue: −23 − 15 = −38 LUFS.
  3. Step 3 — The music is currently at −23 LUFS and must reach −38 LUFS.
  4. Step 4 — Reduction needed: −38 − (−23) = −15 dB.
Answer: Reduce the music track by 15 dB so it sits at −38 LUFS, placing it 15 dB below the −23 LUFS dialogue track.
2. After playing your sequence through the Loudness Meter, the integrated reading is −20 LUFS. Your target is −16 LUFS. How many dB do you need to raise the master output, and will the adjusted level fall within the lab's passing range of −18 to −14 LUFS?
  1. Step 1 — Current integrated loudness: −20 LUFS.
  2. Step 2 — Target integrated loudness: −16 LUFS.
  3. Step 3 — Gain adjustment needed: −16 − (−20) = +4 dB.
  4. Step 4 — Raise the master output by 4 dB and re-measure.
  5. Step 5 — Check: −16 LUFS falls between −18 LUFS and −14 LUFS, so it is within the passing range.
Answer: Raise the master output by 4 dB. The resulting −16 LUFS integrated level falls within the −18 to −14 LUFS passing range.
3. Your 90-second sequence has a dialogue clip at a relative level of −12 dB and a music clip at −12 dB. After ducking, the music drops to −27 dB under the dialogue section. What is the difference in dB between the dialogue and the ducked music, and does this meet the 12–18 dB target?
  1. Step 1 — Dialogue level: −12 dB.
  2. Step 2 — Ducked music level: −27 dB.
  3. Step 3 — Difference: −12 − (−27) = 15 dB.
  4. Step 4 — Check: 15 dB is between 12 dB and 18 dB.
Answer: The difference is 15 dB, which falls within the 12–18 dB target range. The music duck is correctly set.
4. You have 10 caption cues. After instructor review, 2 cues are off by more than one frame. How many cues are frame-accurate, and does the sequence pass the lab's caption criterion?
  1. Step 1 — Total cues: 10.
  2. Step 2 — Cues that fail frame accuracy: 2.
  3. Step 3 — Frame-accurate cues: 10 − 2 = 8.
  4. Step 4 — The passing criterion requires at least 9 of 10 cues to be frame-accurate.
  5. Step 5 — 8 < 9, so the caption criterion is not met.
Answer: Only 8 of 10 cues are frame-accurate. The sequence does not pass the caption criterion; at least 9 of 10 accurate cues are required.
In class: Use the Essential Sound panel to assign roles (Dialogue, Music, Ambience) to each track and apply auto-match loudness to the dialogue track, then manually ride the music volume keyframes so music sits 12–18 dB below dialogue. · Run the Loudness Radar or Loudness Meter in Premiere and adjust the master output until the integrated loudness reads near −16 LUFS (suitable for web/social delivery) as per common platform guidelines. · Create at least 10 caption cues manually in the Caption panel, matching on-screen text to dialogue timing, then export the sequence with captions burned in as a final H.264 MP4.
Materials: Adobe Premiere Pro (current version), Essential Sound panel, Loudness Meter, Caption panel, provided three-track audio sequence from prior lessons, headphones, reference sheet of common loudness targets for major platforms (YouTube, broadcast), export settings guide
Check: Exported MP4 reviewed by instructor: integrated loudness must be between −18 LUFS and −14 LUFS, music must not mask any dialogue word (spot-check of three random moments), and captions must be present and frame-accurate on at least 9 of 10 cues; all three criteria must be met to pass.

Practice check8

Answer each question, then check it. Nothing is recorded; this is your own practice.

Video Production & Post-Production
1. Which measurement unit do major streaming platforms use to specify their audio loudness delivery targets?
Video Production & Post-Production
2. In Premiere Pro's Essential Sound panel, which role should be assigned to the background crowd noise recorded on location?
Video Production & Post-Production
3. A student's music track reads −20 LUFS and the dialogue track reads −8 dB relative. By how many dB must the music be reduced so it sits exactly 12 dB below dialogue?
Video Production & Post-Production
4. What is the recommended True Peak ceiling to prevent inter-sample distortion when exporting to a compressed codec such as H.264?
Video Production & Post-Production
5. A student exports a video and notices the captions do not appear on any player. They used the Caption panel to create cues but forgot one export step. What did they most likely omit?
Video Production & Post-Production
6. When checking integrated loudness with the Loudness Meter in Premiere Pro, when must playback be stopped for the meter to display a valid integrated loudness reading?
Workplace Safety & Health
7. Which of the following best describes the hearing-health practice recommended when using headphones for audio mixing?
Video Production & Post-Production
8. A student places music keyframes so the music drops during dialogue sections. The dialogue track measures −10 dB and the ducked music measures −25 dB. Does this meet the 12–18 dB separation requirement?
Record this lesson on my transcript. Sign in with your key and enrollment email to record lessons (enroll free if you have no key). This page stays open to read and practise.
← Previous lessonProgram planNext lesson →