Circular No. 34/2017/TT-BTTTT Issuing "National Technical Regulations on Sound Volume Levels and Maximum Peak Levels of Audio Signals in Television Programs"

 

CIRCULAR

ISSUING THE "NATIONAL TECHNICAL REGULATION ON LOUDNESS AND TRUE PEAK LEVEL OF AUDIO SIGNALS IN TELEVISION PROGRAMMES"

Pursuant to the Law on Standards and Technical Regulations dated June 29, 2006;

Pursuant to the Law on Telecommunications dated November 23, 2009;

WHEREAS, Decree No. 127/2007/NĐ-CP dated August 1, 2007 of the Government detailing and guiding the implementation of certain provisions of the Law on Technical Standards and Regulations;

Pursuant to Decree No. 06/2016/NĐ-CP dated January 18, 2016 of the Government on the management, provision, and use of radio and television services;

Pursuant to Decree No. 17/2017/NĐ-CP dated February 17, 2017, issued by the Government, on the functions, tasks, powers, and organizational structure of the Ministry of Information and Communications;

Pursuant to the proposal of the Director of the Science and Technology Department,

The Minister of Information and Communications issues this Circular to stipulate the national technical regulation on loudness and true peak level of audio signals in television programmes.

Article 1. Attached herewith is the National Technical Regulation on Loudness and True Peak Level of Audio Signals in Television Programmes (QCVN 115:2017/BTTTT).

Article 2. This Circular takes effect from July 1, 2018.

Article 3. The Heads of the Office, Department of Science and Technology, Heads of agencies and units under the Ministry of Information and Communications, Directors of Provincial Departments of Information and Communications, and relevant organizations and individuals shall be responsible for implementing this Circular.

 

 

QCVN 115:2017/BTTTT

NATIONAL TECHNICAL REGULATION ON LOUDNESS AND TRUE PEAK LEVEL OF AUDIO SIGNALS IN TELEVISION PROGRAMMES

National technical regulation on Loudness and True Peak level of audio signals in television programmes

TABLE OF CONTENTS

Chapter 1. GENERAL PROVISIONS

1.1. Scope of Application

1.2. Applicability

1.3. Referenced Documents

1.4. Terms and Definitions

1.5. Abbreviations

Chapter 2. TECHNICAL PROVISIONS

2.1. Programme loudness

2.2. Loudness range

2.3. True peak level

3. MEASUREMENT METHODS

4. REQUIREMENTSỀ QUA5. RESPONSIBILITIES OF ORGANIZATIONS AND INDIVIDUALS

APPENDIX A (For reference) Guidelines for standardizing loudness and true peak level for television programme productionỔ APPENDIX B (For reference) Guidelines for standardizing loudness level for distribution systems according to EBU R 128

6. IMPLEMENTATION ORGANIZATION

QCVN 115:2017/BTTTT is based on standards EBU R 128, EBU Tech 3341, EBU Tech 3343, EBU Tech 3344.无效QCVN 115:2017/BTTTT was developed by the Centre for Research and Application of Science and Technology in Broadcasting (BRAC), Vietnam Television, and the Institute of Post and Telecommunications Science and Technology, reviewed by the Department of Science and Technology, and issued together with Circular No. 34/2017/TT-BTTTT dated November 22, 2017.

1.1. Scope

LIST OF REFERENCES

 

Foreword

This regulation applies to the loudness and true peak level of audio signals in television programmes transmitted and broadcast in Vietnam.

This regulation does not apply to internet broadcasting.

 

NATIONAL TECHNICAL REGULATION ON LOUDNESS AND TRUE PEAK LEVEL OF AUDIO SIGNALS IN TELEVISION PROGRAMMES

National technical regulation on Loudness and True Peak level of audio signals in television programmes

Chapter 1. GENERAL PROVISIONS

This regulation applies to entities and businesses transmitting and broadcasting radio and television programmes in Vietnam.i of regulation

ITU-R BS.1770-2 (03/2011): Algorithms to measure audio programme loudness and true-peak audio level.

EBU R 128 (08/2011): Loudness normalization and permitted maximum level of audio signals.

1.2. Applicability

EBU Tech 3341 (2011): Loudness Metering: 'EBU Mode' metering to supplement loudness normalization in accordance with EBU R 128.

1.3. Referenced Documents

EBU Tech 3342 (2011): Loudness Range: A measure to supplement loudness normalization in accordance with EBU R 128.

1.4.1. Programme

(programme)

A distinct content containing sound and images used in television broadcasting. An advertisement, introduction, commercial content, or similar content is also considered a programme under this regulation.

1.4. Terms and Definitions

1.4.2. Short-form content (Short-Form Content)

A programme with a short duration, typically less than 30 seconds (with repeated sections not exceeding 2 minutes).

1.4.3. Loudness meter Specialized measuring equipment compatible with EBU R128 and/or ITU-R BS.1770-2 used for measuring loudness.

1.4.4. Target loudness level

(target loudness level)

A specific loudness value used for loudness normalization.

1.4.5. Loudness normalization â(loudness normalization) Adjusting the loudness of different programmes to the same level.

1.4.6. Loudness metadata

(loudness metadata) âMetadata containing information about the loudness level of audio signals. 1.4.7. Comfort zone

(comfort zone)

This is a range from (+2.4 dB, -5.4 dB) of audio loudness used for research on some listener samples. 1.4.8. Dialnorm

A parameter representing loudness contained in metadata and transmitted in the AC-3 bitstream, with values ranging from 1 to 31.

1.4.9. DRC Profile Describes how metadata information is used to control dynamic range.

1.4.10. Sound reproduction level

(Sound Reproduction Level)

The sound reproduction level for home theatre devices is -31 LUFS or -27 LUFS.

1.4.11. Limiter

(limiter)

Processing to limit the peak level of audio signals. â1.4.12. Downmix (Downmix)

A coefficient used for reference and mixing multi-channel audio signals in the audio mixing technique of a multi-channel programme into a programme with fewer channels.

Downmixing is a term used for audio processing, mixing multi-channel audio signals in the audio mixing technique of a multi-channel programme into a programme with fewer channels. For example, when converting 6-channel audio (often referred to as 5.1 audio) to stereo (2 channels), the process is called downmixing. 1.4.13. Headroom

A buffer for the peak level of audio signals to avoid signal distortion.

1.4.14. Three measurement timeframes Measurement timeframes for loudness, consisting of three frames as follows:

- The shortest timeframe is called 'momentary', abbreviated as 'M'.

- The intermediate timeframe is called 'short-term', abbreviated as 'S'.

- The timeframe for a programme or segment is called 'integrated', abbreviated as 'I'.

DAB

Digital Audio Broadcasting Organization

Digital Audio Broadcasting

DAB+

DAB using the AAC codec

DAB using the AAC codec

1.5. Abbreviations

Chapter 2. TECHNICAL PROVISIONS

2.1. Programme loudness

The average integrated program loudness measured over the entire duration of a television program. Program loudness is calculated in LUFS units, denoted as L.

b. Requirement for program loudness_k.

= -23 LUFS ±1,0 LU_k

L_k For programs with short content (<30 seconds) (such as commercials, advertisements), in addition to the above program loudness L

requirement, the following condition must also be met:_k + Maximum short-term loudness:

= -18,0 LUFS (+5,0 LU on the relative scale).

L_k Distribution of loudness within a television program, denoted as LRA

2.2. Loudness range

The average integrated program loudness measured over the entire duration of a television program. Program loudness is calculated in LUFS units, denoted as L.

b. Requirement

LRA < 20 LU

NOTE: This requirement does not apply to programs with short content.

2.3. True peak level

The true peak level indicates the maximum (positive or negative) level of the continuous signal waveform in the time domain, and this value may exceed the highest sampled peak value, denoted as Max TP.无效Max TP = -1 dBTP.

The average integrated program loudness measured over the entire duration of a television program. Program loudness is calculated in LUFS units, denoted as L.

NOTE: This requirement applies to all types of programs.

LRA < 20 LU

3.1. Measurement equipment requirements

- The measuring device must be compatible with ITU-R BS.1770-2 and EBU R 128 standards.

3. MEASUREMENT METHODS

- The measuring device must support three time windows according to EBU Tech 3341.

- The measuring device must support at least the display functions according to EBU Tech 3341.

3.2. Determination method

- Measure "momentary" loudness using a sliding window of length 0.4s for the "M" time window. The measurement method does not use gating;

- Measure loudness over a short period using a sliding window of length 3s for the "S" time window. The measurement method does not use gating. The update rate for the direct meter is at least 10 Hz;

- Measure loudness for an entire program or a program segment using the gating method as specified in ITU-R BS.1770-2 for the "I" time window. The update rate for the broadcast program meter is at least 1 Hz; use a silent gating threshold of -70 LUFS to calculate absolute gating levels; use a relative gating threshold of -10 LU compared to the absolute gating level; the input is measured with gating applied to 400 ms frames with an overlap of 75% between consecutive windows.

The final part of the integrated loudness measurement process that does not fill a complete frame will be discarded.

- Measurement diagram

Measurement signal: If measuring monaural sound, only one audio channel input is required.

Stereo sound measurement: Measure on two channels L and R and calculate loudness.

NOTE:

5.1 audio sound/measurement measurement: Only measure loudness for five channels L/R/C/Ls/RS. Ignore the LFE channel (see diagram in ITU-R BS.1770-2).

4. MANAGEMENT REGULATIONS

The loudness and maximum true peak level of the audio signal in television programs transmitted and broadcast in Vietnam under section 1.1 must comply with the requirements set forth in these Standards.

5. RESPONSIBILITIES OF ORGANIZATIONS AND INDIVIDUALS

Organizations and enterprises responsible for transmitting and broadcasting radio and television programs in Vietnam have the responsibility to ensure that the transmitted and live broadcast signals comply with these Standards, implement conformity declarations, and be subject to inspection by state management authorities in accordance with regulations.

The Telecommunications Administration, the Radio, Television and Electronic Information Administration, and Provincial Departments of Information and Communications have the responsibility to guide and manage organizations and enterprises to implement these Standards.

During the implementation of these Standards, if any issues arise or difficulties occur, relevant organizations and individuals should report them in writing to the Ministry of Information and Communications (Science and Technology Department) for guidance and resolution.

6. IMPLEMENTATION ORGANIZATION

6.1. Guidelines for normalizing loudness and peak levels in television program production

6.2. In cases where the provisions set forth in this Standard are changed, supplemented, or replaced, they shall be implemented according to the new document.

6.3. A.1. Normalization methods

 

A.1. Measurement Server

(For reference)

A.1.1. Production and post-production phase

There are two methods for normalizing loudness during the production and post-production phases: âMetadata containing information about the loudness level of audio signals.

+ Method 1 involves maintaining the actual signal level and processing gain adjustment;

+ Method 2 involves controlling loudness and normalization (in which case gain adjustment is not necessary or only very small adjustments are needed), see Diagram A.1.

Diagram A.1 - Two methods of normalizing loudness during the production and post-production phases

An acceptable error margin of ± 0.1 LU around the expected level of -23 LUFS, except in special cases (silent drama programs...).

Method 1 (maintaining actual levels): In most cases, it is necessary to reduce gain. Therefore, subsequent processing steps such as reducing dynamic range and/or limiting the maximum true peak level are usually unnecessary. Since most adjustments involve reducing gain, the metadata solution is not suitable for Method 1.shallMethod 2 (loudness normalization): After the program has been measured and tested, a loudness meter is installed in parallel with the regular meter (usually a QPPM) to monitor loudness. However, the dynamic range may increase due to additional sound effects added to the signal (such as background noise from fans in sports programs, sound effects in game shows with audiences); or dialogue recorded in a studio is often compressed for artistic reasons, balanced by retaining the original dynamic range of multiple takes.

A.1.2. Measuring loudness during production and post-production

Method 1 (Maintain Practical Level): In most cases, it is necessary to adjust the gain with a negative value (reducing). Therefore, the next processing step of reducing the dynamic range and/or limiting the maximum peak level is usually not required. Since most adjustments involve reducing the gain, the use of metadata is not suitable for Method 1.

Method 2 (Volume Normalization): After the program has been measured and tested, a volume meter is installed in parallel with the standard meter (usually QPPM) to monitor the volume. However, the dynamic range may increase due to additional audio effects (such as background noise from fans in sports programs, sound effects in audience-participatory game shows); or dialogue in recording studios is often compressed for artistic reasons, balanced by multiple recordings retaining the original dynamic range.

A.1.2. Measuring Volume During Production and Post-production

The volume meter uses the "EBU mode" measurement mechanism with three measurement time frames, where the M and S window times are used for measuring instantaneous levels and processing mixes of audio signals. Setting the initial level can be best achieved using the M meter to adjust the levels of main components, the "anchor" parts of the audio signal (speech, music, or sound effects) to approach the target level of -23 LUFS. The mixer itself can know the loudness level at any point in the actual signal as per the M and S measurement modes.

Due to inconsistencies between ITU-R BS. 1770 and ITU-R BS. 1771, this standard proposes the following symbols:

• The symbol for the loudness level, K-weighted curve (Loudness Level, K-weighted) is "L"_k".

• The unit symbol "LUFS" indicates the value of L_k across the entire digital full-scale range.

• The symbol "LU" indicates the value of L_k describing relative reference results or differences in loudness levels compared to the reference level.

The interface of any volume meter compatible with "EBU mode" will define two scales: "EBU +9 Scale" commonly used for most programs, and "EBU +18 Scale" only used for programs with wide LRA. Both scales may display relative loudness levels in LU or absolute values in LUFS. Specifically, "0 LU" in "EBU mode" corresponds to the target level of -23 LUFS.

A.1.3. LRA for production and post-production

Normalizing the volume to meet requirements also means controlling the LRA value (when the dynamic range may expand through processing). This is crucial in ensuring that the audio signal is suitable for the intended audience and distribution chain. For production and post-production, a common mixing model may be chosen (with relatively high LRA values and Max TP at -1 dBTP), while other stages may use lower LRA and Max TP values (but still ensuring the L_K level is -23 LUFS).

When measuring LRA, the system can determine appropriate measurements to compress the dynamic range of a program within an acceptable level. To achieve the target level of -23 LUFS, parallel leveling adjustments along with gain control of the compression process may be performed based on the original volume level.

A.2. File-based Production and Playout

In file-based workflow applications, volume normalization must adapt to each process. The fundamental principles remain the same: normalizing volume and controlling the dynamic range of the audio signal, particularly for new content. However, when metadata information is integrated as part of the file-based system, metadata usage solutions have their own advantages.

An initial promotional file containing audio signals can be sourced from ingest processes, transmitted from another server, or retrieved from a file-based archive. Current programs (stored content) have four processing options to meet volume normalization requirements as follows:

- Adjusting the volume level of all audio files to the target level;

- Only adjusting the volume level when required;

- Using measured volume level results to adjust playback levels without changing the original volume level;

- Transmitting metadata information to customers so that their devices normalize volume based on this metadata.

The practical use of metadata information depends on many factors such as infrastructure, workflow, media asset management (MAM) resource management, availability of suitable equipment, financial resources, time, etc. Initially, during the early stage of a file's lifecycle in the system, it must be measured to obtain three parameters: L_K, LRA, and Max TP (the S and M frame time maximum levels are also measured for short contents < 30s). Depending on the measurement results and the method chosen for volume normalization with permissible LRA levels, the volume processing solution will handle each parameter separately (Figure A.2) or simultaneously (Figure A.3).

The three block diagrams in Figure A.2 play a core role in file quality control processing based on technical measurements from audio content. The process starts with measuring the three parameters L_K, LRA, and Max TP, and from these measurement results, each parameter adjustment is carried out as shown in the flowcharts in Figure A.2.

HìFigure A.2 - Processing normalization of individual parameters

Figure A.3 illustrates possible scenarios with parameters, specifically:

Case (a): All three parameters meet the requirements. This is the ideal case when the L_K level is -23 LUFS, the LRA level is within the broadcaster's limit (depending on the genre and/or distribution platform), and the Max TP level is equal to or below the predetermined maximum level for the designed distribution system.

HìFigure A.3 - Volume processing solutions handling multiple parameters simultaneously

Case (b): L_K> -23 LUFS. The solution is to reduce the gain by:

Gain (dB) = L_K (expected level) - L_K(measured level)

Example: Measured L_K = -19.4 LUFS, expected level = -23 LUFS, the gain adjustment needed is: [-23 - (-19.4)] = -3.6 dB, at which point the Max TP level also decreases proportionally to L_K.

Case (c): L_K< -23 LUFS. The solution is to increase the gain (gain offset), in which case the Max TP level must be recalculated: Max TP (original) + Gain Offset = Max TP (adjusted) to control the possibility of the adjusted Max TP exceeding the allowed limit. If the adjusted Max TP exceeds the limit, peak limiting processing (TPL) is implemented (Figure A.2 (c)). Another solution that can be executed (without using real peak limiting) is to maintain the original L_K level and supplement with appropriate volume metadata information. To do this, a fully functional audio encoding system supporting and transmitting metadata (such as Dolby Digital or MPEG-4) is required.

In both cases (b) and (c), the gain value is stored as metadata and can be used for subsequent processing steps if the Max TP exceeds the limit (case (c)). This gain value can also be used to control the playback level of the file to the expected -23 LUFS level.

Case (d): The L_Klevel < -23 LUFS and the LRA is wider than the permitted level for the program genre or distribution channel. The L_K may be processed as in case (c), while the level of LRA may be adjusted downward using the block diagram in Figure A.2 (b), this processing step may reduce the level of Max TP. When adjusting the level of L_K then the level of Max TP may exceed the permissible limit, but further processing of the increased Max TP level may not be necessary because the subsequent LRA processing reduces the Max TP level again. Therefore, it is only necessary to calculate and adjust the level of Max TP during the reduction of the LRA level.

Case (e): The LRA exceeds the permitted level for the type of program or distribution channel (while L_K and Max TP meet the requirements). As previously described, a compressor with a low threshold and appropriate compression ratio can narrow the LRA (Figure A.2 (b)). For files, automatic processing to achieve the LRA level triggers dynamic range compression with appropriate parameters. The adjusted Max TP level is usually lower than the initial level, thus additional processing is not required.

Case (f): The Max TP level exceeds the permissible limit. Exceeding the permissible Max TP level in the distribution system may cause distortion in the downstream signal (during D-A conversion, sampling frequency conversion, bit rate reduction, etc.). Processing is carried out according to the diagram in Figure A.2 (c) to reduce the Max TP level. Even if the L_K changes significantly, the Max TP level is only affected by the number of peaks detected in the signal.

Other possible cases of the parameters L_K, LRA, Max TP may be categorized into one of the above cases.

A.3. Metadata

A.3.1. Metadata information L_K

With the requirement to standardize the audio signal production to the expected level of -23 LUFS, the related metadata parameter is set to indicate -23 LUFS.

Cases where values other than -23 LUFS may occur include:

- Programs that do not comply with the -23 LUFS and -1 dBTP requirements. This occurs with films having significant dynamic range variations and broadcasters wanting to transmit these programs with high loudness/peak ratios.

- Old programs from archives that have not been adjusted within the system to meet the requirements of this standard.

- Live programs with varying levels and metadata.

- A complete metadata system in the signal chain, where the signal and metadata are transmitted to customers, and their devices automatically adjust the levels based on the received data according to this standard.

A.3.2. Metadata for dynamic range control

Dynamic range processing similar to loudness normalization can be performed either directly on the audio signal itself or through metadata. If metadata is used, dynamic range compression information is sent as part of the data stream in the form of gain-words (gain-indicating words). In home customer equipment (such as home theater systems), this information is used to reduce the dynamic range of the signal (either by default from the device or activated by the user). Metadata-based dynamic range control provides "additional information" when users want a lower dynamic range.

There are two compression mechanisms currently used in Dolby Digital encoding: "Line mode" and "RF mode" corresponding to each specific compression mechanism setup.

With this standard-compliant system, audio normalization is based on the -23 LUFS loudness level and uses the LRA parameter to determine the processing method, and the "None" setting is used. In practice, Dolby Digital encoding can be configured with "Line mode" or default to "RF mode".

Dynamic range control of the audio signal during transmission will be based on the upstream stream. For a given program, the broadcaster may select a profile category for the RF-mode system or choose "None" for the Line-modesystem. Broadcasters can also create settings other than "None" to support specific workflows (though such settings may not be fully compatible with listener devices).

A.3.3. Level reduction coefficients (downmix)

These metadata parameters (for the Dolby Digital system) apply only to surround sound signals. They support controlling the gain (dB) of the center channel, mixing the surround channels with the left and right channels to produce a two-channel (stereo) signal. The volume of the two channels measured after automatic level reduction (channel mixing, e.g., from surround sound to stereo) depends on:

- The actual downmix coefficients: +3 / +1.5 / 0 / -1.5 / -3 / -4.5 / -6 / -∞,

- The content of the center and surround channels,

- The safety margin to avoid overload conditions.

Here, signal overload after level reduction processing occurs when using dynamic range processing for the upstream stream. A flexible scale should be used because the volume of the two-channel signal after level reduction can differ greatly from the original surround signal.

In the Dolby Digital system, the downmix coefficients follow two downmix profiles. Initially, there was only one profile providing coarse parameters, specifically -3/4.5/-6 dB for the center channel and -3/-6--∞ for the surround channels. Currently, there is a second profile containing extended bitstream information (BSI) to provide finer parameters (refer to DVB TS 101 154, section on downmix coefficients). However, note that not all devices can understand the downmix coefficients using the second profile's extended BSI information, so if the decoder cannot understand the information from the second profile, it will use the coefficients from the first (coarse) profile.

In the case where the metadata for the downmix coefficients is lost or unreliable, it is recommended to use the coefficients according to ITU-R BS.755-2 as follows:

- Left, Right (L, R front): 0 dB,

- Center, Surround Left, Surround Right (C, Ls, Rs): -3 dB.

NOTE: Surround channels have a weight of +1.3 dB throughout the measurement process (according to ITU-R BS.1770). However, if automatic downmix processing is used, this weighting is not applied since the result is only two-channel (left and right) audio. Programs with more surround channels will lead to greater volume variation when reducing channels to two-channel audio compared to programs with fewer surround channels.

In certain cases, metadata information provided from separate files without attaching the audio file may not be accurate, or may lose metadata information... In such cases, the metadata for L will take default information from the manufacturer, which is -27 (the default value for dialnorm in the Dolby Digital system) or -31 (the lowest allowed value in the system)._K The default information from the manufacturer is -27 (the default value for dialnorm in the Dolby Digital system) or -31 (the lowest allowed value in the system).

Recommendation: If external metadata information about volume or dynamic range control is not guaranteed to be reliable, it should be removed (except in truly trustworthy cases). Downmix coefficients from metadata are only accepted with fully supported or closed systems. If this cannot be ensured, re-measurement of the three main audio parameters must be conducted to ensure subsequent processing accuracy.

 

ANNEX B

(For reference)

Guidelines for normalizing audio levels in distribution systems according to EBU R 128

B.1. Normalizing audio levels in digital distribution systems

B.1.1. Differences in audio levels during distributionâIn practice, when distributing television signals worldwide, audio signal levels have been quite different across regions over a long period due to various reasons such as competition among TV stations, TV programs...

Users increasingly demand higher audio signal volumes as they have more access to high-quality TV programs.

Normalizing audio levels during distribution will largely eliminate differences, allowing users to switch between services comfortably without encountering annoying volume issues.

B.1.2. Active normalization for distributing television services

Figure B.1 illustrates a block diagram in an integrated digital distribution system (head-end) that can be applied to different platforms like IPTV, satellite. âAudio level normalization is carried out through three main functional blocks:No.- Measurement unit;No.

- Steering unit;

- Adaptation unit.

The adaptation unit can be integrated into the DVB multiplexer or equivalent digital processing equipment. The measurement and steering units can be combined into a single application. Adaptive processing of audio levels may also be influenced by the input gain control process of encoding devices (encoders). However, re-encoding using the same compression format is not selected due to potential quality degradation and cost.

The audio level normalization system can support one or multiple types of encoding (codecs). The adaptive processing solution for audio levels depends on the type of codec used:

- Directly within the audio signal stream for MPEG-1 Layer II encoding.

- Within the accompanying metadata for DD/DD+ and HE-AAC encoding.

Decoded signal volumes are continuously measured throughout the day (24 hours), divided into 24 segments, each lasting 1 hour. The start time of segment 1 is 03:00, and of segment 24 is 02:00 the next day. The reason for using this time is its minimal impact on daily programming. Volume measurements are performed according to the "I" frame time window for each segment.

Figure B.1- Integrated distribution system model

For cases using DD/DD+ and/or HE-AAC encoding, volume metadata information is determined throughout the measurement process for target querying when restoring volume levels. The measurement system applies a reference volume level of -31 LUFS for DD/DD+ encoding and -23 LUFS for MPEG-1 Layer II and HE-AAC encoding. With DD/DD+ systems, volume metadata is recorded through the Dialnorm descriptor of the DD/DD+ data stream and uses the standard SRL reference level of -31 LUFS SRL. For HE-AAC systems, volume metadata is recorded by applying the Programme Reference Level (PRL) descriptor (using the prog_ref_level parameter according to ISO/IEC 14496-3) of the HE-AAC data stream and the target decoding level descriptor (target_level according to ISO/IEC 14496-3) at -23 LUFS. If the HE-AAC stream does not contain volume metadata, the service will insert new data or re-encode if the service volume information is incompatible with -23 LUFS ± 0.1 LU.

The 24 segments (blocks) in a day are measured, and the values of the segments are not less than 2 LU lower than the highest retained value (this is also equivalent to a tolerance of ±1 LU in EBU R128). The average of the maximum values of the segments is represented for the station's operation and considered as the service volume. This value may slightly differ for each program measured according to the target level. The audio level and allowable tolerance of programs must be measured before normalization. The measurement unit can be selected for this task. Volume balancing for individual services is applied to all programs based on the overall daily measurement results.

HìA database listing all TV stations measured and normalized (including special audio services for multilingual purposes, audio description services) will help identify them, and this database will also contain the volume measurement results. The steering unit compares the service volume information (including corresponding adjustment factors in metadata) with the expected levels for services using MPEG-1 Layer II and HE-AAC encoding; and with the SRL for services using DD/DD+ encoding.rime Minister cSports âMetadata containing information about the loudness level of audio signals.

After obtaining data from all services at 03:00, the steering unit compares the data with the target level and adjusts the offset if the measurement tolerance exceeds ±0.1 LU compared to the expected level. In this way, all services are kept close to the expected level.

For consistency, the steering unit will use a step size of 0.5 LU over 24 hours. Depending on the system encoding, the step size value may vary, specifically using the following step sizes:

• 2 LU for signal stream adaptation systems using MPEG-1 Layer II encoding

• 1 LU for encoding adaptation systems using MPEG-1 Layer II encoding, and for metadata adaptation systems using DD/DD+ and HE-AAC encoding.

For consistency, the control block will use a step size of 0.5 LU over 24 hours. Based on the system encoding, the step size value may vary, specifically using the following step sizes:

• 2 LU for systems adapting to signal flow using MPEG-1 Layer II encoding

• 1 LU for systems adapting to encoding using MPEG-1 Layer II encoding, and for systems adapting to metadata using DD/DD+ and HE-AAC encoding.

An option may measure additional peak real maximum values of decoded signal. Measuring the peak real maximum values of a 2-channel stereo signal reduced from multi-channel service is most meaningful when the volume levels of services increase. The measurement results are also transmitted to the control block. The control block may support the additional measurement option for monitoring purposes, or special monitoring with switching and sharing services.

To support dynamic volume indicators as shown through parameters in the metadata of encodings (such as Dialnorm and PRL), a compensation process will be applied with a deviation to the obtained value. The deviation applied for systems using DD/DD+ and HE-AAC encoding is usually to reduce the volume of the service and the deviation tends to increase; while the deviation applied for systems using MPEG-1 Layer II encoding is also usually to reduce the volume of the service, but the deviation tends to decrease.

Figure B.2 illustrates the volume measurement diagram and normalization method for services with different encodings.

HFigure B.2 - Block diagram of the process of measuring services corresponding to the encodings used (a) MPEG-1 Layer II; (b) DD/DD+ and/or HE-AAC

B.1.3. New Enhanced Services

New enhanced services added or cases where there is a disturbance in the step value and schedule can be configured manually, for example, direct volume measurement at the end day or several hours before. It is best to implement this with new services such as sound services for listeners. The control block will be responsible for handling in this case.

B.1.4. Registration Information andn l3.2.7. National Institute of Hygiene and Epidemiology, Pasteur Institutes

Volume registration and notifications are executed in the control block for automatic normalization monitoring. Information about the deviation value, metadata values about the actual reception level, adjustment level are also reported. Some conditional warning messages may be generated:

- If the peak real value of the decoded signal exceeds -1 dBTP after normalization.

- If the peak real value of the 2-channel signal due to the reduction of the decoded signal exceeds -1 dBTP after normalization.

- If the measured volume service results over two consecutive days differ more than a predetermined threshold, for example, 3 LU.

- If the deviation value for systems using MPEG-1 Layer II encoding is lower than or higher than predetermined thresholds, for example, +6 and -14 LU.

- If the deviation value causes the metadata information about the volume of encoded DD/DD+ or HE-AAC audio signals to adjust below -31 LUFS or above a predetermined threshold (for example, -10 LUFS).

- If the volume metadata information in services using DD/DD+ or HE-AAC encoding is invalid or missing.

Volume normalization must be continuously and automatically processed. System measurement warning notifications must be monitored daily, and volume measurement at the central control room or system (network) operation center must allow random checks when necessary.

B.1.5. Local Distribution Systemsâlocal scaling factor

HFigure B.3 -ng Diagram of standard audio distributionshallfrom the central distribution system to local distribution systems

Local distribution systems distribute standardized audio content from the central distribution system. Figure B.3 illustrates the process of distributing normalized audio from the central distribution system to local distribution systems, which can be used in IPTV distribution methods.

Figure B.4 - Provision and Control of Volume Data âfrom the Central Head-End to Local Head-Ends

In the case where the central distribution system does not provide normalized audio, local distribution systems may normalize as in the model described in Section B.1.2. Optionally, volume data from the central distribution system can also be used to remotely control volume at local distribution systems through data connections. In that case, it is necessary to ensure that the signal provided to the measurement system does not have volume differences between the received signals from the local distribution system and the signals provided by the central distribution system as shown in Figure B.4. If the received signal from the local distribution system, such as terrestrial reception, has a different volume level, this signal must be simultaneously provided to the central distribution system's measurement system (typically, to control volume from the central distribution system, other received signals from the local distribution system must be simultaneously provided to the central distribution system regardless of volume differences, even if this signal is not reused at the central distribution system for further distribution). Usually, only the audio part from the local distribution systems is provided to the central distribution system's measurement system, while the video part is separated and not distributed to save bandwidth. Figure B.5 presents this control model.

B.1.6. Non-uniform Audio Signal Sources"b) In addition to the lists of public services issued according to the provisions of Clause 2, Article 4 of this Decree, specialized agencies under provincial People's Committees shall report to the provincial People's Committee for decision-making on amending, supplementing, or issuing the list of public services funded by the state budget within their jurisdiction and consistent with the local budget capacity within the approved budget by the Provincial People's Assembly, and send it to the Ministry of Finance and relevant ministries and sectors for supervision during implementation."Distribution systems may be provided with non-uniform audio signal sources, some services need to be increased in volume. Providing a service to listeners with a "soft" volume level will not irritate the listener but will cause problems when the listener switches channels to another service. In some cases, background music loops used in news or weather reports require consultation between broadcasters and distributors to ensure listener satisfaction with volume levels, or a fixed normalization method can be used for these unusual services. A fixed normalization method can be understood as manual adjustment based on frequency repetition information from logging in the measurement block, while the processing of fixed normalization will be handled by the control block.

B.1.7. Ad Insertion

B.1.7. Inserting Advertisements

For systems with their own ad insertion signals, the insertion signal may be located behind the volume normalization system. Ad signals can be normalized to a volume level that uses software algorithms for standardization. At this point, the average volume level of the broadcast content is balanced with the expected level at the ad insertion position. In sound systems using DD/DD+ or HE-AAC encoding, local ad insertion metadata will accurately indicate the actual volume level, and the service transmitted through the ad playback system will not be adjusted for gain or attenuation from the system. Algorithms in the playback system are designed to follow the average volume level of the main program (without processing previously standardized data). The post-insertion ad volume monitoring process is optional.

B.1.8. On-demand television systems (VoD) and other playback systems

VoD service providers are also considered broadcasters (distribution systems) and their audio signals can be normalized. Stored content in playback systems is also checked and normalized (for example, by software), meaning the average volume level of promotional content is balanced with the target level. In systems using DD/DD+ or HE-AAC encoded audio signals, volume metadata from VoD programs will always accurately reflect the actual volume level. Continuous volume measurement of VoD content for monitoring purposes is recommended.

Figure B.5 illustrates the volume normalization model in VoD and ad insertion systems. This model can be applied to IPTV distribution methods.

Figure B.5 - Block diagram of an integrated volume normalization digital distribution system for VoD and ad insertionshalla volume normalization for VoD and ad insertion

B.1.9. Regionalized services

HFigure B.6 - Block diagram đof a head-end integrated volume normalization system and a local head-end implementing a regional service swap in the service stream"b) In addition to the lists of public services issued according to the provisions of Clause 2, Article 4 of this Decree, specialized agencies under provincial People's Committees shall report to the provincial People's Committee for decision-making on amending, supplementing, or issuing the list of public services funded by the state budget within their jurisdiction and consistent with the local budget capacity within the approved budget by the Provincial People's Assembly, and send it to the Ministry of Finance and relevant ministries and sectors for supervision during implementation."released on the product packaging (bottle cap, box lid, bottle stopper, wine spout or similar position) ensuring that when opening the lid, the seal will tear and cannot beine service stream

A main service can be replaced by another service depending on the distribution region (spliced into the DVB data stream) and there may be differences in volume levels between the signal sources. If the regional service is played back from a file-based system, the volume normalization method similar to that used for VoD services can be applied. In the case of a live regional service, the distribution studio must ensure that the distributed volume level meets the requirements of this standard. Regional services can be volume-adjusted if necessary by measuring the regional signal separately and making adjustments during the transmission period. The control block will handle this real-time process. Figure B.6 describes the volume normalization when swapping a regional service into the main service stream, and this model can also be applied to IPTV distribution methods.

B.2. Volume normalization in analog distribution systems

B.2.1. Volume level differences âin the pre-emphasis processâIn practice, when distributing television signals worldwide, audio signal levels have been quite different across regions over a long period due to various reasons such as competition among TV stations, TV programs...

Similar to section B.1.1, volume level differences are influenced by pre-emphasis processing.

After normalization, long-term measured volume levels will balance out, but peak true maximum levels, QPPM, and pre-emphasis processing of services may differ, and volume normalization does not affect these measured values (clipping does not occur when appropriate limiting is applied at the modulation stage).

B.2.2. Limiting processing

After volume normalization, peak levels and pre-emphasis limiting may be too high for analog transmission systems. Therefore, true peak and pre-emphasis limiting processing will be applied (with AM L systems, only true peak limiting is needed). This process can be performed with dedicated equipment or integrated directly into the analog modulator. (limiting) true peaks and pre-emphasis limiting will be applied (with AM L systems, only true peak limiting is required). This process can be performed with dedicated equipment or integrated directly into the analog modulator.

For television distribution systems using FM modulation, pre-emphasis limiting needs to be implemented (according to ITU-R BS.642) to reduce the headroom for the audio signal peak level to prevent distortion. The optimal location for the pre-emphasis limiter is either directly connected or integrated within the modulator itself.

B.2.3. Active volume normalization for analog television distribution

Analog distribution systems prioritize distributing audio signals that have been volume-normalized from central digital distribution systems so that all distribution systems meet volume requirements effectively. Analog modulators can be set with default adjustment levels to control stable audio levels, simplify operations, optimize costs, and achieve consistency across the system.

B.2.4. Local analog distribution systems

For analog distribution systems using unnormalized signal sources, volume normalization should be performed at the local distribution system. An option is to use volume data from the central distribution system to remotely control the volume at the local distribution system via data connections, similar to digital distribution systems.

B.2.5. Design of analog distribution systems for televisionân distribution systems for television

Figure B.7 illustrates a general design of an analog distribution system for television where volume normalization has been performed in a digital distribution system or at an earlier analog stage. The signal source can be an MPEG-1 Layer II encoded audio data stream, reduced-level DD/DD+, reduced-level HE-AAC, or directly from the studio. This figure also includes a 15 kHz low-pass filter compatible with transmission standards placed before the pre-emphasis limiting processor to prevent limiting processing of contents above 15 kHz.

B.3. Level matching in analog and digital distribution systems无效B.3.1. Level matching between systems and interfaces

B.3.1. Adjusting Levels Between Systems and Interfaces

To avoid inconsistencies in volume levels, the volume adjustment mechanisms between transmission systems and output interfaces must be implemented. The parameters used to enhance sound quality for currently used systems are illustrated in Figure B.7.

Figure B.7 - A central designân distribution systems for television

B.3.2. The modulation methods used for analog television systemsức modulation methods used for analog television systems

If the audio signal is normalized to a level of -23 LUFS before being supplied to the analog modulator, the device will operate in its default setting without requiring any sound level adjustments.

A 1 kHz sine wave is used for reference purposes (in accordance with CENELEC EN50049), the thresholds of the limiting processor include the pre-emphasis limiting gain, which are determined based on actual peak values.

Table B.1 - Parameters used for television systems

B.4.3. Level adjustment for television systems

Figure B.8 illustrates the level adjustment between systems and interfaces and is suitable for systems that simultaneously use both left and right audio channels. The red line represents the maximum peak level during transmission, the purple lines indicate the level of the limiter for the transmission system and corresponding adjustments at the input and output interfaces. The area to the left of the red line represents the maximum allowable peak level at the interface or encoding system. If the signal is a 1 kHz sine wave, the red line indicates the measurable level.

The program level may exceed the red line at the output interface due to overshoots in the encoding system. However, these overshoots will not reach the red line threshold because they are reduced in level. Based on direct transmission measurements, it may be necessary to reduce the level of the limiter processor in front of the codec system if relatively low bit rates are the common cause of overshoots. The gray lines emphasize the relationship when adjusting to characteristic levels:

-12 dBTP corresponds to the reference level defined according to CENELEC EN50049.

-18 dBTP is the calibration signal defined according to ITU-R BS.645, the signal level can be found at the input and output of the encoding system, the RF modulator system, and the interfaces from figure B.8.

Figure B.8 - Level adjustment in systems B, B1, D, D1, G, H, K, K1, I and无效o implement mứI1 I1

Some annotations in the drawing are as follows:

Production level: this is the digital audio signal level measured in dBTP of the input and playback sources in the studio. According to standards, the limiting processing for production level is -1 dBTP, this is the maximum allowed peak level (using a true-peak meter with an interpolating filter four times the sampling rate).

Encoders: MPEG-1 Layer ll/HE-AAC/DD(+)/DAB/DAB+: this is the digital audio signal level measured in dBTP of the encoded audio using one of the listed encodings. The red line represents the highest peak level. In practice, this means that the signal will pass through the final stage of the peak limiter set at -3 dBTP before entering the encoder, it may be necessary to reduce the level of the limiter in front of the encoding system if a relatively low bit rate is used.

Decoders: MPEG-1 Layer ll/HE-AAC/DAB/DAB(+) and DD+ in RF Mode: this is the digital audio signal level measured in dBTP of the decoded audio using HE-AAC, MPEG-1 Layer II or DD/DD+ encoding. If the Dialnorm value or PRL (Programme Reference Level) does not match the expected -23 LUFS, the signal level relationship will differ. For a program, the decoder output level may be higher than the input level due to overshoot effects, however, these overshoots must be maintained below the predetermined red line level. If a relatively low bit rate is used, Figure B.8 can be used to compare measured levels with the predetermined maximum levels to reduce the level of the limiter when necessary.

For decoders using HE-AAC, MPEG-1 Layer II, and DD/DD+ in RF Mode, different maximum peak levels are specified. Since the internal volume expectation of the DD/DD+ decoder is -20 LUFS reduced to -23 LUFS (using software attenuation reducing 3 dB), the actual cut-off level if a 1 kHz sine wave is used is -3 dBTP. The recommended maximum output level of the DD/DD+ decoder in RF Mode for a program is set at -4 dBTP (1 dB lower). The recommended studio limiter level is -3 dBTP, however, this does not cause any issues. Typically, systems using DD/DD+ encoding incorporate built-in overload protection, so the maximum peak level for the encoding process does not need to be reduced.

Output: HDMI/SPDIF/HDMI ARC in home mode: this is the digital PCM (Pulse Code Modulation) signal level measured in dBTP at HDMI, SPDIF, or HDMI ARC interfaces if the device is an IRD, IDTV, or multimedia player operating in home theatre mode. IRD/IDTV devices operate in data stream output modes (audio streams encoded according to specific types), internal decoders for DD/DD+ or HE-AAC within the device are bypassed, and the encoded data stream is transferred to SPDIF and HDMI (ARC) interfaces for data output (unless there is a special application such as audio description information where the audio stream must go through the decoders to extract application information, or there is E-EDID (Extended Display Identification Data) information identifying the playback device only supports basic audio). Typically, IRD, IDTV, or multimedia devices have two basic outputs depending on the home mode mechanism in the device's installation menu.

HDMI output with TV/SPDIF/HDMI ARC/AES3 mode in two-channel sound: This is the digital PCM (Pulse Code Modulation) signal level of decoded audio, measured in dBTP at the HDMI interface if the IRD operates in TV mode, or at the SPDIF, HDMI ARC, or AES3 interface if the IRD/IDTV operates in two-channel mode.

NOTE: AES3 (also known as AES / EBU) is a standard for exchanging digital audio signals between professional audio devices. AES3 was developed by the Audio Engineering Society (AES) and the European Broadcasting Union (EBU).

Output: SCART or RCA interface: This is the RMS analog signal level, measured in mV, of the decoded audio signal at the SCART or RCA output of the IRD, IDTV, or multimedia device. It also represents the RMS analog signal level, measured in mV, at the SCART/RCA output of the television receiver using the HDMI or SPDIF interface.

Output: XLR or analog interface: This is the RMS analog audio signal level measured at the XLR (balanced) output of the professional IRD. For example, the relative dBu0s level is understood as the absolute dBu level if the normalization factor 0dBrs or -3 dBrs is applied in sequence (as defined by ITU-R BS.645).

RF modulation: Systems B, B1, D, D1, G, H, K, K1, I, and I1: This is the peak FM deviation measured in kHz of the decoded audio signal (with pre-emphasis limit processing) provided to the RF modulator input. The red line indicates the highest peak level. In practice, the signal may exceed the -6.7 dBTP level (the threshold of the pre-emphasis limiter processor from the previous processing stage) before reaching the RF modulator. The purple line indicates the adjustment level of the limiter processor. The limiter processor may also be designed to be integrated within the RF modulator itself, thus requiring a reduction in the limiter processor level.

IRD output: SCART or RCA interface: This is the RMS analog signal level measured in mV of the demodulated audio signal at the SCART or RCA output used for internal TV RF tuning or recording device.

 

TABLE OF REFERENCES体制机制物质技术条件、人力资源报告（根据2016年7月1日第105/2016/NĐ-CP号政府决议附表02规定，该决议对计量器具和测量标准的检定、校准、检测活动条件进行了规定，并经2018年第154/2018/NĐ-CP号决议第二条第十二条修正）[1] EBU Tech 3205-E The EBU Standard peak-programme meter for the control of international transmissions.

[2] ITU-R BS.645 Test signals and metering to be used on international sound programme connections.

[3] ITU-R BS.1770 Algorithms to measure audio programme loudness and true-peak audio level.

[4] EBU Tech 3341 Loudness Metering: 'EBU Mode' metering to supplement loudness normalisation in accordance with EBU R 128.

[5] EBU Tech 3342 Loudness Range: A measure to supplement loudness normalisation in accordance with EBU R 128.

[6] EBU Tech 3343 Guidelines for Production of Programmes in accordance with EBU R 128.

[7] EBU Tech 3344 Guidelines for Distribution and Reproduction of Programmes in accordance with EBU R 128.

[8] EBU R 128 Loudness normalisation and permitted maximum level of audio signals.

[9] EBU R128 s1, 2014 Loudness Parameters For Short-Form Content.

[10] EBU R128 s1 V2, 2016 Loudness Parameters For Short-Form Content.

[11] ITU-R BS.642-1, 1990 Limiters for high quality sound programme signals.

[11] ITU-R BS.642-1, 1990 Limiters for High Quality Sound Programme Signals.