DTC 491 - Advanced Digital Cinema

Week 7 — Post-Production Lab: Montage & Voice

Loops

Interview Lab: Post-Production

This week is a direct continuation of the Studio Interview Lab. You now move from recording presence to shaping meaning through montage.

Using the shared interview footage, you will begin weaving together voices, images, sound, text, and rhythm—discovering structure in the timeline rather than executing a fixed plan.

Editing is where the interview becomes cinema.

Editing Workflow

Assembly Edit

The assembly edit is exploratory. It is about getting material into the timeline and discovering relationships.

Rough Cut

The rough cut introduces structure.

Final Cut

The final cut refines intention.

Montage Tools & Strategies

Interviews

B-roll & Cutaways

Sound & Music

Text & Titles

Sound Waves to Digital Signals

Sound is physical: a vibration moving through air as waves of compression and rarefaction. When those waves reach a microphone, the diaphragm moves in response, converting mechanical motion into an electrical signal. That analogue signal is then sampled — measured thousands of times per second — and converted into binary data. This is the analogue-to-digital conversion at the heart of all digital audio.

Essential Sound Terms

Frequency / Hertz (Hz)
The rate of vibration of a sound wave, measured in cycles per second. Frequency is what we perceive as pitch — low Hz is bass, high Hz is treble. Human hearing spans roughly 20 Hz to 20,000 Hz (20 kHz). A voice sits between 80 Hz and 8 kHz; presence and intelligibility live around 2–5 kHz.
Amplitude / Decibels (dB)
The strength of a sound wave — perceived as loudness. Measured in decibels. In digital audio, 0 dBFS (full scale) is the maximum before clipping — distortion caused by a signal exceeding what the system can record. Dialogue typically sits around –12 to –6 dBFS. Never let peaks hit 0.
Dynamic Range
The difference between the quietest and loudest sounds in a recording or a mix. A wide dynamic range is expressive and cinematic; heavily compressed audio with a narrow dynamic range sounds flat and fatiguing. Film mixes preserve dynamic range; broadcast and streaming platforms compress it.
Sample Rate (kHz)
How many times per second the audio signal is measured during digital conversion. 48 kHz is the standard for film and video — always match this in your project settings. Mismatched sample rates cause audio to play back at the wrong pitch or speed.
Bit Depth
How precisely each sample is measured — the resolution of the audio. 24-bit is standard for production recording. 16-bit (CD quality) is acceptable for delivery. Higher bit depth means more dynamic range and cleaner recordings at low levels.
Gain vs. Volume
Gain is the input level — how much signal you capture at the microphone or bring into the edit. Volume is the output level — how loud something plays back. Gain affects the quality of the signal; volume only affects playback level. Always set gain correctly first; adjust volume in the mix.
EQ (Equalisation)
Adjusting the balance of frequencies within a sound — boosting or cutting specific Hz ranges. Used to remove unwanted low-end rumble (high-pass filter), add clarity to dialogue (boost around 3–5 kHz), or shape the tonal character of an atmosphere. In Premiere: Audio Effects → EQ or Parametric EQ.
Room Tone
The ambient sound of a specific space — the "silence" of a room when no one is speaking. Every environment has its own noise signature: HVAC hum, traffic, reverb. Recording 30–60 seconds of room tone on every location gives you neutral fill material to smooth cuts and patch gaps between dialogue.
Lufs (Loudness Units Full Scale)
A measurement of perceived average loudness over time, used for broadcast and streaming delivery standards. YouTube targets –14 LUFS; film streaming platforms typically –23 LUFS. Premiere's Loudness Radar effect and the Match Loudness panel in Essential Sound can measure and match this automatically.

Key Premiere Audio Settings

In-Class Post Lab: Cutting 14 Interviews to under 5 Minutes

You have 12 interview clips — each featuring a different student speaking about their interests and ideas for the group project. Your goal is to reduce this material to appoximately 5 minutes. The edit should surface each student's distinct voice while suggesting the shared territory of the project.

Lab 1 — AI Text-Based Editing

Premiere's AI transcription tools let you read and edit an interview the way you might edit a document — selecting and deleting text to shape the timeline. This is especially powerful when working across many clips: you can scan all 12 interviews for recurring ideas, strong phrases, and moments of contrast without scrubbing through video.

Steps

  1. Open Window → Workspace → Text-Based Editing. Select Select the clips for Project panel (top right).
  2. With a clip in the Source Panel, click Transcribe Sequence in the left Text panel. Transcription runs in the background — a progress indicator appears in the panel.
  3. Once complete, read through each interview transcript. Highlight any phrase that you want for the subject and then in the Source, click Insert. This will take the selected text sequence and insert into the Timeline. Premiere builds a new sequence using only those selected clips. Repeat for each subject.
  4. Use the search bar in the Transcript panel to find recurring words across all 12 clips — look for thematic threads you can build around.
  5. Optionally, select Filter for certain words, pauses, filler words, etc.
  6. Once you have a text-driven assembly, return to the timeline to refine cut points, add handles, and adjust for rhythm.

Lab 2 — Sound Fixes

The interview clips have two consistent problems: background noise and low overall volume. Before editing and mixing, we need to clean the audio so the dialogue is clear and consistent. Work through the following steps in order.

Step 1 — Repair: Reduce Noise and Rumble

Start with the Essential Sound panel to remove common background problems. Select a clip that contains dialogue and tag it as Dialogue in the Essential Sound panel.

  1. Select one dialogue clip or all clips.
  2. Open Window → Essential Sound.
  3. If necessary, click Dialogue to assign the audio type.
  4. Expand the Repair section.
  5. Enable Reduce Noise and start with a low amount (around 3–5). Increase only until the background noise is reduced without harming the voice.
  6. Enable Reduce Rumble. This removes very low frequencies below about 80 Hz that usually come from HVAC systems, traffic vibration, or microphone handling.

Step 2 — Optional: Enhance Speech

Premiere also offers an AI-based tool called Enhance Speech. This attempts to isolate and clarify spoken dialogue by boosting voice frequencies and suppressing background noise. Be careful, this could add weird artifacts in the sound.

  1. In the Essential Sound panel, locate the Enhance Speech section.
  2. Enable the feature and adjust the strength slider.
  3. Listen carefully while adjusting.

Use this tool carefully. At higher settings it can make voices sound metallic or artificial. For many interviews, the Repair tools alone produce a more natural result.

Step 3 — Adjust Volume Across All Clips

Once the audio is cleaned, adjust the dialogue volume so it sits at an appropriate level in the mix. In digital audio, 0 dBFS is the maximum possible level. Dialogue peaks should normally remain below –6 dBFS to prevent clipping.

  1. Select all interview clips in the timeline (Cmd/Ctrl + A, then deselect any clips you don't want to affect).
  2. Right-click one of the selected clips → Audio Gain.
  3. Choose Normalize Max Peak to: –6 dB.
  4. Click OK.

This raises each clip so its loudest moment reaches –6 dBFS while preserving the relative dynamics of the recording.

As a final check, open the Essential Sound panel and use the Loudness → Auto-Match feature. This adjusts the average perceived loudness of all dialogue clips so they feel consistent when played together.

Sending Audio to Adobe Audition (Advanced Noise Removal)

If background noise remains after using Premiere's tools, Adobe Audition provides more advanced spectral repair tools. These allow you to sample the noise itself and remove it from the recording.

  1. Right-click the clip in the Premiere timeline → Edit Clip in Adobe Audition.
  2. Audition opens with the clip loaded.
  3. Find a short section that contains only background noise (no speech).
  4. Select that region and go to Effects → Noise Reduction/Restoration → Capture Noise Print.
  5. Select the entire clip (Cmd/Ctrl + A).
  6. Open Noise Reduction (Process) and apply the captured noise profile.
  7. Adjust the reduction settings until the noise is minimized while the voice remains natural.

Save the file and return to Premiere. The cleaned audio will automatically update in your timeline.

Lab 3 — Color Grading

The interview clips currently look flat — low contrast, muted color, and some subjects show noticeable light glare or hot spots on their faces. Color grading addresses both the technical baseline and the expressive tone of the material.

Minimizing Facial Glare and Hot Spots

Glare on faces — bright, shiny patches on the forehead, nose, or cheeks — usually comes from specular reflection caused by direct lighting on skin. The best solution is proper lighting during filming, but several tools in Lumetri Color can reduce the problem in post-production.

  1. Open the Lumetri Color panel (Window → Lumetri Color) and select the affected clip.
  2. Start in the Basic Correction section. Reduce Highlights (try around –40 to –70) and lower Whites slightly. These controls darken the brightest parts of the image and often recover detail in shiny skin areas without affecting the rest of the exposure.
  3. For glare on a specific part of the face, use a mask inside the Lumetri Color panel. Click one of the mask icons (oval or pen), draw a mask around the bright area, and apply the Highlights adjustment only within that region. Use mask feathering and tracking so the correction follows the subject as they move.
  4. If glare is extremely strong and confined to a small area, more advanced repair may be required. Tools in Adobe After Effects, such as clone or healing effects, can remove reflections more precisely. These methods are usually reserved for important shots rather than routine editing.

Correcting Interview Color with an Adjustment Layer

Because all of these interviews were recorded in the same lighting setup, we can apply a base color correction across the entire sequence using an Adjustment Layer. An adjustment layer acts like a transparent filter placed above the timeline. Any Lumetri Color changes applied to it affect every clip beneath it at once.

  1. In the Project panel, click the New Item icon → Adjustment Layer. Accept the sequence settings and click OK.
  2. Drag the Adjustment Layer to a video track above all interview clips in the timeline. Stretch it so it covers the entire sequence.
  3. Select the Adjustment Layer and open the Lumetri Color panel. All corrections made here will apply to every interview clip below it.
  4. Start with Basic Correction:
    • Increase Contrast slightly (around +20 to +35).
    • Lower Shadows slightly to add depth.
    • Raise Whites modestly to brighten highlights without blowing them out.
  5. Open Lumetri Scopes (Window → Lumetri Scopes) and check the Waveform. Make sure highlights do not exceed the top of the scale and shadows are not crushed into pure black.

This adjustment layer creates a consistent base grade across the entire interview.

Fine-Tuning Individual Clips

Even though the lighting setup is the same, different skin tones reflect light differently. Some faces may appear slightly darker or lighter than others. After establishing the base grade with the adjustment layer, you can fine-tune individual clips.

  1. Select the clip that needs adjustment in the timeline.
  2. Open the Lumetri Color panel.
  3. Add a second Lumetri Color effect directly to that clip. This correction applies only to that clip and will sit on top of the adjustment layer correction.
  4. Use small adjustments in Basic Correction:
    • Raise or lower Exposure slightly to balance skin brightness.
    • Adjust Temperature if skin tones look too cool or too warm.
    • Use Shadows or Highlights for subtle balance.

The goal is not to make every face identical, but to ensure that all subjects look natural and evenly lit within the overall interview sequence.

Workshop: Group Project Introduction (15%)

This is your transition into the major group production for the semester. Teams of 3–4 will make a short in one of three modes:

Every short must actively consider framing, color, lighting, sound, and editing.

Forming Groups

Bring one rough idea you're curious about. We'll form groups around shared interests and compatible roles — who's directing, who's running camera, who's sound, who's lighting, who's editing — and plan rotations so everyone learns the full workflow.

Note on AI Integration

Experiment with workflows that combine live-action shooting with AI tools: visuals, voice, texture, remix, or hybrid scenes. For in-class demos, AI video generation will be demonstrated using a course account.

Workshop Tasks (Today)

  1. Decide which mode: Fiction, Documentary/Essay, or Hybrid AI Short
  2. Form in groups of 3–4. Identify compatible roles as a starting point: director, camera, sound, lighting, editor. Plan to rotate so everyone touches each role.
  3. Pitch to group initial ideas. What is the subject, and what draws you to it? What props or locations are available?
  4. If a shared constraint, genre or theme has been descide. Begin outlining a script and storyboard.
  5. Draft a brief visual and sound style statement: what does this film look like, sound like, and feel like?
  6. Share your group, mode, and pitch with the class before the end of session.
  7. Assign role to work on production: script > storyboard > production manager > performers > camera and directing.