FAIRLIGHT / MASTERING — PRACTICE
Reviewed: 2026-10-09
Guide: https://videomontazher.com/en/fairlight/mastering

Goal: prepare and verify a finished file by connecting the delivery brief,
the measured programme, constant-gain headroom and the actual Output Track.

The lab is an authored calculation from entered measurements. It does not
measure audio, reproduce Resolve Normalize/Optimize, a limiter or a codec,
or certify an exported file. Every target below belongs to an exercise;
none is a universal platform requirement.

1. DESCRIBE THE DELIVERY
Choose a complete short programme. Record the requirements first.
Edit version and language: ___________________________________________
Full In/Out range and expected duration: ______________________________
Measurement standard/mode: ___________________________________________
Integrated target and tolerance: ______________ LUFS / _______________
Maximum true peak: ____________________________ dBTP
Format, codec, Sample Rate, PCM bit depth/bitrate: _____________________
Channel format, order and required versions/stems: ____________________
If a requirement is missing, label your own exercise assumption clearly.

2. CHECK THE PATH AND BALANCE
Name the final bus and follow dialogue, music, effects and returns.
Check Mute/Solo, automation, sends and any duplicate source paths.
Final bus name and format: ___________________________________________
Selected monitoring source: __________________________________________
Planned Deliver Output Track: ________________________________________
One balance issue and its cause: _____________________________________
Control Room serves listening. Changing speaker level does not replace
an adjustment in the delivered mix signal path.

3. MEASURE AND CALCULATE
Reset before a fresh complete pass of the required programme. Record
actual Integrated and maximum true peak from the same version and range.
Integrated: __________________ LUFS; true peak: ________________ dBTP
Measurement range and settings: ______________________________________

Enter these values in the lab, then the target and ceiling from step 1.
Required gain = target LUFS − measured LUFS.
Maximum peak-permitted linear gain = ceiling dBTP − measured dBTP.
With the constraint enabled: chosen gain = the smaller of those values.
Estimated LUFS/true peak = original measurement + chosen gain.
A negative permitted gain means attenuation is required.

Compare the three examples at target −16 LUFS and ceiling −1 dBTP:
- Peak conflict: −22 LUFS, −4 dBTP. Required +6 dB, permitted +3 dB.
  Constraint on: +3 dB; estimate −19 LUFS and −1 dBTP; 3 LU shortfall.
  Constraint off: +6 dB; −16 LUFS and +2 dBTP; 3 dB above the ceiling.
  The model does not clip the peaks.
- Enough headroom: −19 LUFS, −7 dBTP. A gain of +3 dB gives −16 LUFS,
  −4 dBTP and 3 dB of peak headroom.
- Reduce level: −12 LUFS, −1 dBTP. A gain of −4 dB gives −16 LUFS,
  −5 dBTP and 4 dB of peak headroom.

Original measurements displays 0 dB gain while retaining the settings.
LUFS/dBTP/Gain only highlight the drawing. Reset restores the initial
measurements, requirements, constraint, result view and All focus.
What constrains your own example: ____________________________________
The resulting decision for the real mix: ______________________________

Calculation boundary: a LUFS shift equal to gain assumes unchanged gated
block membership. The fixed absolute gate can change that membership.
True-peak prediction assumes linear gain alone. Signal shape, a limiter,
codec, dither and resampling are not modeled. Repeat the real measurement
after changes; the shortfall is not a limiter gain-reduction prescription.

4. PREPARE AND RENDER
Correct a specific cause in the mix and measure again. In Deliver, check
Export Audio, required video, Single Clip, range, Output Track, channel
count and order, Sample Rate, format, codec and bit depth.
Normalize to Standard and Optimize to Standard are different modes.
Record the mode/standard/targets and enabled outputs, or its disabled state.
Export normalization: ________________________________________________
Filename, version and Render Queue job range: _________________________
Verify a short passage first, then render the complete programme.
If final word-length reduction to integer PCM is required, coordinate
dither with that last conversion using the selected tool's documentation.
Do not add another dither or subsequent gain without understanding the path.

5. VERIFY THE RESULT
Import the finished file through a verification path without reapplying
processing, automation or normalization. Inspect structure, duration,
sync, L/R and the required format. Listen through the full programme:
dialogue, attacks, transitions, opening and tails. Run a fresh complete
loudness and true-peak analysis.
Actual Integrated and maximum true peak: ______________________________
Format/channels/duration: ____________________________________________
Issues found and corrected: __________________________________________
Review date, version and reviewer: ___________________________________

For stems, agree common boundaries, contents and processing. Independent
normalization changes their balance; their sum need not reproduce a master
that passed through common nonlinear processing. Specify the capture point
and reconstruction path.

PRIMARY SOURCES
Blackmagic Design, DaVinci Resolve 21 Reference Manual:
https://documents.blackmagicdesign.com/UserManuals/DaVinci_Resolve_21_Reference_Manual.pdf
Pages 166; 4130–4140; 4143; 4177, 4179; 4193, 4196; 4205–4206; 4210–4212.
iZotope Ozone 8 Help, Dither and Dithering Tips (general conversion order,
not documentation of a native Resolve switch):
https://downloads.izotope.com/docs/ozone8/dither/index.html
https://downloads.izotope.com/docs/ozone8/tips-and-examples/index.html#dithering-tips
Gating and measurement units: see the primary ITU/EBU sources in
Loudness Meter — https://videomontazher.com/en/fairlight/loudness-meter