Programme audio for UK broadcasters under the DPP AS-11 delivery specification follows EBU R 128: integrated loudness of -23.0 LUFS, within plus or minus 0.5 LU for pre-recorded programmes (plus or minus 1.0 LU for live and as-live), and a maximum true peak of -1 dBTP. The fastest way to know your mix passes before it is wrapped into the AS-11 MXF file is to measure the exported mix and stems against that spec and keep the numbers. WavSpec, a Windows app from Automata Labs, does exactly that with a built-in UK DPP AS-11 broadcast spec: it measures every file, corrects what can be corrected without touching the mix, re-measures the delivered copies, and writes a report.
This guide is for freelance editors and small post houses who mix in a DAW, hand the audio to an online editor or finishing house, and then hear back from the broadcaster’s QC weeks later. Catching a loudness fail at your desk costs minutes. Catching it after the file has been rejected costs a re-delivery.
Check a whole batch against the UK DPP broadcast spec and keep a report for the QC desk.
Get it from Microsoft
What is DPP AS-11, and where does the audio fit?
The Digital Production Partnership (DPP) is the UK industry group, a collaboration of UK broadcasters including the BBC, ITV and Channel 4, that publishes the common technical and metadata standards for file-based delivery of TV programmes. Programmes are delivered as AS-11 UK DPP HD files: an MXF wrapper holding the video, the audio tracks and a defined set of metadata. When the DPP moved audio compliance from traditional PPM levels to EBU R 128 loudness, the audio part of the spec became a loudness spec (TVBEurope summary of the DPP update).
So there are two jobs. Building the MXF file with the right metadata is the online editor’s job, in their finishing software. Making sure the audio inside it meets the loudness and peak rules is the mixer’s job, and it can be checked on the mix and stems before they are laid back. That second job is where a batch checker earns its keep.
What are the UK DPP loudness and peak numbers?
The figures below come from the DPP technical specification as reported when version 5.0 tightened the tolerances, and from EBU R 128 itself.
| Measurement | DPP / EBU R 128 requirement | Notes |
|---|---|---|
| Integrated (programme) loudness | -23.0 LUFS | Measured over the whole programme per ITU-R BS.1770 |
| Tolerance, pre-recorded | plus or minus 0.5 LU | Tightened from plus or minus 1 LU in DPP v5.0 |
| Tolerance, live and as-live | plus or minus 1.0 LU | Previously plus or minus 2 LU |
| Maximum true peak | -1 dBTP | Lower ceilings may apply where lossy coding follows |
| Short-term and momentary loudness | Mixing aids for long-form | Pass or fail limits apply to short-form under R 128 s1 |
Sources: the Production Expert analysis of the DPP v5.0 audio changes, which quotes paragraph 2.2 of the spec (“Non-live -23.0 LUFS ±0.5LU; Live (including as-live) -23.0 LUFS ±1.0LU”), and EBU R 128, which sets the -23.0 LUFS target level and states that the true peak level “shall not exceed -1 dBTP”. The EBU’s QC item for true peak also notes that -3 dBTP may apply where bitrate compression is used (EBU QC 0084B).
Two practical points follow. First, plus or minus 0.5 LU is tight. A mix that lands at -22.4 LUFS fails, and the difference is not audible on most monitors. Second, true peak is measured with oversampling, so a sample-peak meter that reads -1.0 dBFS can still hide inter-sample overs above -1 dBTP. If you want the terms in more depth, our explainer on LUFS, LKFS, dBTP and LRA covers each one.
Why do mixes that sounded fine fail DPP loudness QC?
The usual causes are mundane:
- A late edit after the final measurement. A music cue replaced or a scene trimmed after the mixer metered the programme moves the integrated value.
- Different meters, different gating. A plug-in set to the wrong standard, or a meter reading a selection instead of the whole programme, gives a number that does not match the QC tool.
- Limiting on sample peaks. A brickwall limiter set to -1 dBFS with no oversampling leaves true peaks above -1 dBTP.
- Stems that do not sum to the mix. When the deliverable includes M&E or split stems, each track is checked, and a stem bounced at a different gain fails on its own.
- Normalizing by gain alone. Raising a quiet mix to -23 LUFS by adding gain can push peaks over the ceiling unless a true-peak limiter follows.
None of these are judgement calls. They are measurements, and they can be made on the exported files before anything is sent.
How do you check DPP audio compliance before delivery?
A repeatable check takes five steps:
- Export the final mix and every stem the delivery schedule lists, at the sample rate, bit depth and track layout the delivery schedule asks for.
- Measure each file over its whole duration with a BS.1770 meter: integrated loudness, true peak with oversampling, loudness range, and the maximum short-term value.
- Compare each value with the spec, including the tolerance, and note which files fail and by how much.
- Correct only what the spec allows you to correct without remixing: a static gain change to the target, and true-peak limiting where the gain change creates overs. Anything else goes back to the mix.
- Re-measure the corrected files from disk, not from the session, and keep the before and after numbers with the delivery.
You can do steps 2 and 3 with a loudness meter plug-in, one file at a time. Step 5 is the one people skip, and it is the one a QC desk effectively repeats.
How does WavSpec check a UK DPP delivery?
WavSpec runs the five steps above as one batch on your PC. Drag the mix and stems, or a whole folder, into the window and choose the built-in UK DPP AS-11 broadcast spec. Each built-in spec in WavSpec is cited to its source document and dated, so you can see what it is checking against.
| Step | What WavSpec does |
|---|---|
| Read | WAV, Broadcast Wave, RF64, AIFF, FLAC, MP3, AAC and M4A, in mono, stereo, 5.1 and 7.1 |
| Measure | Integrated, short-term and momentary loudness, loudness range, and 4x oversampled true peak per ITU-R BS.1770-5, plus sample rate, bit depth, channel layout and container |
| Grade | Each criterion marked pass, fail or advisory against the chosen spec |
| Correct | Gain to the exact target, look-ahead true-peak limiting only where needed, sample-rate conversion with a windowed-sinc resampler, bit-depth conversion with TPDF dither |
| Verify | Re-measures every delivered file from disk |
| Report | PDF, HTML, CSV and JSON with before and after values, a loudness timeline per file, a batch consistency line, your client and project details, and a SHA-256 digest of every delivered file |
Corrected files go to a separate delivery folder; your originals are never overwritten. Broadcast Wave files keep their bext, iXML, LIST and cue chunks, and the verified loudness is written into the bext chunk. Our guide to normalizing Broadcast WAV without losing metadata explains why that matters for timecode and slate information.
Measure, correct and re-measure the mix and every stem in one run, with the originals untouched.
Get it from Microsoft
What WavSpec does not do for an AS-11 delivery
It helps to be precise about scope, because AS-11 is a whole-file specification:
- It does not build or check the MXF file. WavSpec delivers PCM WAV (or constant-bit-rate MP3 for specs that want it). Wrapping audio and video into an AS-11 UK DPP HD file, and filling in its metadata, is done in your finishing or wrapping software.
- It does not remix. If dialogue sits too low against music, a gain change will hit the integrated target and still leave a mix the broadcaster may reject on intelligibility. That is a mix note, not a correction.
- It does not replace the broadcaster’s QC. The report shows what was measured on your PC, with which spec, and the digest of each file you sent. It is evidence for a conversation with QC, not a certificate.
When does a batch tool beat a meter plug-in?
A meter plug-in on the master bus is the right tool while you mix. A batch checker is the right tool when the mix is done and you have a set of files to deliver:
| Situation | Meter plug-in | Batch checker such as WavSpec |
|---|---|---|
| Mixing a single programme | Best fit | Not needed until export |
| A series of episodes, each with stems | One file at a time, by hand | Whole folder in one run |
| Proving what you delivered | Screenshots or notes | Report with before and after values and file digests |
| Fixing a 0.6 LU miss | Re-bounce from the session | Gain and true-peak limiting on a delivery copy |
| Recurring client deliveries | Repeat setup each time | Saved spec per client, watch folder or command line |
If you also deliver to other territories, WavSpec’s built-in specs include EBU R 128 broadcast, EBU R 128 s1 short-form, ARD/ZDF/ORF TV 2025 and ATSC A/85, and you can build a named spec of your own for a client’s house rules. Our EBU R 128 s1 explainer covers the stricter rules for promos and trailers, and the audio delivery QC checklist lists the non-loudness checks worth doing at the same time.
Frequently asked questions
What loudness does the BBC, ITV or Channel 4 want?
For programmes delivered under the UK DPP specification, integrated loudness of -23.0 LUFS with a tolerance of plus or minus 0.5 LU for pre-recorded content and plus or minus 1.0 LU for live and as-live content, and a maximum true peak of -1 dBTP. Always check the delivery schedule for the specific commission, which can add requirements.
Is LUFS the same as LKFS?
Yes. They are two names for the same K-weighted loudness measurement defined in ITU-R BS.1770. European specs tend to write LUFS and US specs LKFS.
Can I just normalize the mix to -23 LUFS?
You can, but a gain change that raises a quiet mix can push true peaks above -1 dBTP. A compliant correction applies the gain and then a true-peak limiter only where the new peaks need it, and then measures the result again. WavSpec does those three steps in order.
Does WavSpec work for the stems as well as the mix?
Yes. Drop all the files in one batch; each one is measured and graded against the spec on its own, and the report has a line for every file plus a consistency line for the batch.
Does any audio leave my PC?
No. WavSpec runs offline on Windows 11, with no account, no cloud upload and no telemetry.
Try WavSpec
WavSpec is a one-time purchase in the Microsoft Store, and the Store runs a 15-day free trial with every feature. Run your last delivered programme through the UK DPP spec during the trial and compare the report with the QC notes you received. More detail is on the WavSpec product page, and the EBU R 128 normalization page covers the same workflow for European broadcasters.
Run your next delivery through the UK DPP spec during the 15-day free trial.
Get it from Microsoft
BBC, ITV and Channel 4 are trademarks of their owners. WavSpec is not affiliated with or endorsed by the DPP, the EBU or any broadcaster.
Last updated 19 October 2026. Published by Automata Labs, a software company in Canada.