Deliverables Checklist: WAV Specs, Stems, M&E, Loudness, and Naming Conventions

Getting the creative side of ES-LATAM dubbing right is only half the battle. If your deliverables do not meet technical specifications, your content will be rejected at the platform ingestion stage, delayed during QC, or worse, published with audio issues that reach the audience. This guide is a practical reference covering the technical requirements that every dubbing deliverable must satisfy: audio format specifications, stem structure, loudness compliance, M&E handling, file naming conventions, and quality control checkpoints.

Bookmark this page. You will come back to it.

Audio Format: WAV 24-bit / 48 kHz

The industry standard for dubbed audio deliverables is WAV (Broadcast Wave Format) at 24-bit depth and 48 kHz sample rate. This is non-negotiable for virtually all major streaming platforms and broadcast distributors.

Why These Specs Matter

24-bit depth provides approximately 144 dB of dynamic range, giving mix engineers ample headroom for processing without introducing quantization noise. This is particularly important for dialogue, where quiet passages and loud exclamations can span a wide dynamic range.
48 kHz sample rate is the standard for video production and broadcast. It aligns with the audio clock used in professional video editing environments and avoids the sample rate conversion artifacts that would occur if you delivered at 44.1 kHz (the CD standard) and the platform needed to convert.

Common Mistakes to Avoid

Delivering MP3 or AAC files instead of WAV. Lossy formats are never acceptable for master deliverables.
Upsampling from a lower bit depth or sample rate. If your recording chain captured at 16-bit/44.1 kHz, upsampling to 24-bit/48 kHz does not recover the lost information. Always record at the target specification from the start.
Using 32-bit float WAV files without confirming platform acceptance. Some platforms accept them, many do not. When in doubt, deliver 24-bit integer.

Stems: DX, MX, FX, and the Full Mix

Most dubbing deliverables require isolated stems in addition to the final mixed file. Stems allow the platform or downstream facility to remix, adjust, or reversion the content without going back to the original session.

Standard Stem Structure

DX (Dialogue): The isolated dubbed dialogue track, free of music and effects. This is the core dubbing deliverable.
MX (Music): The isolated music track. In most dubbing workflows, this comes from the original production's M&E or from a separate music stem provided by the content owner.
FX (Effects): The isolated sound effects track, including Foley, ambience, and hard effects. Like MX, this typically originates from the original production.
Full Mix (FMIX): The final combined mix of DX + MX + FX, balanced and mastered to specification.

Stem Delivery Requirements

Each stem must be the exact same duration as the program, including any handles (typically two seconds of silence at head and tail).
Stems must be phase-aligned. When summed, the DX + MX + FX stems should reconstruct the full mix precisely.
Each stem should be delivered as a separate WAV file at 24-bit/48 kHz.
Channel configuration must match the specification: stereo (L/R) is most common for streaming; 5.1 surround stems may be required for premium content.

M&E Handling

The Music and Effects (M&E) track is the backbone of any dubbing project. It contains everything except the original dialogue, and the quality of the M&E directly affects the quality of the final dubbed mix.

What to Check in Received M&E

Before starting any dubbing work, verify the M&E:

Completeness: Play the M&E against the picture. Are there gaps where the original dialogue was improperly removed, leaving holes in the ambience or effects?
Sync: Does the M&E align perfectly with the video? Even a single frame of drift will cause audible artifacts in the final mix.
Quality: Listen for artifacts, distortion, or remnants of the original dialogue bleeding into the M&E.
Format: Confirm the M&E matches the target specification (24-bit/48 kHz, correct channel count).

When the M&E Is Incomplete

If the M&E has quality issues, the dubbing facility may need to perform fill work: recreating ambiences, Foley, or effects to cover gaps left by imperfect dialogue removal. This adds time and cost. Flagging M&E issues early in the workflow prevents surprises downstream.

Loudness Standards

Loudness compliance is a hard requirement for broadcast and streaming delivery. Two standards dominate the industry.

EBU R128 (European Broadcasting Union)

Target integrated loudness: -23 LUFS (Loudness Units Full Scale)
Permitted deviation: +/- 1 LU
True peak maximum: -1 dBTP
Loudness range (LRA): typically not mandated but monitored

EBU R128 is the prevailing standard for European broadcasters and is widely adopted by international streaming platforms.

ATSC A/85 (Advanced Television Systems Committee)

Target integrated loudness: -24 LKFS (Loudness K-weighted Full Scale)
Permitted deviation: +/- 2 LU
True peak maximum: -2 dBTP

ATSC A/85 is the standard for North American broadcast and is used by several major US-based streaming platforms.

Practical Considerations for ES-LATAM Dubbing

Always confirm which loudness standard applies before beginning the final mix. Different platforms serving ES-LATAM markets may use different standards.
Measure loudness on the full mix, not individual stems. Stems may have different loudness profiles that are only correct in combination.
Use ITU-R BS.1770 compliant metering tools. Approximations or non-standard meters will produce unreliable readings.
Document your loudness measurements and include them in the QC report delivered with the final files.

File Naming Conventions

Inconsistent file naming is one of the most common causes of delivery rejection. A clear naming convention eliminates ambiguity and ensures that every file can be identified, sorted, and ingested without manual intervention.

Recommended Naming Structure

A robust naming convention typically includes these elements in a fixed order:

[SeriesCode]_[EpisodeNumber]_[Language]_[StemType]_[Version].[ext]

Example:

SHORTDRAMA01_EP003_ESLA_DX_V01.wav
SHORTDRAMA01_EP003_ESLA_MX_V01.wav
SHORTDRAMA01_EP003_ESLA_FX_V01.wav
SHORTDRAMA01_EP003_ESLA_FMIX_V01.wav

Naming Rules

Use uppercase letters and underscores. Avoid spaces, special characters, and accented characters in file names.
Include language codes consistently. For ES-LATAM, common codes include ESLA, ES-419, or es-LA depending on the platform specification.
Version numbers are mandatory. Every revision must increment the version number (V01, V02, V03).
Never overwrite a previous version. Maintain all versions until the client confirms the final accepted delivery.

Versioning and Revision Tracking

For serialized content like short drama, where dozens or hundreds of episodes move through the pipeline, version control is critical.

Maintain a delivery log that records every file delivered, its version number, delivery date, and acceptance status.
When a revision is requested, document the specific issue that triggered the revision alongside the new version number.
Archive superseded versions for a defined retention period (typically 90 days post-acceptance, though some clients require longer).

QC Checklist for Deliverables

Before any file leaves the studio, run through this checklist:

Format verification: WAV, 24-bit, 48 kHz confirmed via file properties (not assumed).
Duration check: All stems and the full mix are identical in duration, including handles.
Phase alignment: Summed stems match the full mix with no phase cancellation artifacts.
Loudness compliance: Integrated loudness and true peak measurements fall within the target standard. Measurements documented.
Sync verification: Dubbed dialogue aligns with picture. Spot-check at least three points per episode (beginning, middle, end).
File naming: Every file follows the agreed naming convention exactly. No deviations.
Completeness: All required stems and mixes are present. No missing files.
Playback check: Full listen-through of the final mix against picture, checking for audio artifacts, dropouts, or glitches.
Metadata: Broadcast WAV metadata fields (originator, description, time reference) populated if required by the client.

Build Reliable Deliverables Into Your Workflow

Technical deliverables are not an afterthought. They are the final expression of every creative and production decision made during the dubbing process. Getting them right consistently requires established specifications, trained engineers, and systematic QC processes.

Sound Ally builds deliverable compliance into every ES-LATAM dubbing project from day one. Our services cover the full chain from recording through final QC and delivery. If you need a partner who treats technical specifications with the same rigor as creative quality, reach out to us.