Both PCM and DSD file conversions can cause overload. DSD files may initially have reduced loudness to provide sigma-delta modulator stability. Read how to automatically normalize audio level by average value or maximal peak (maximize album loudness), Frequently Asked Questions.
To normalize audio peak volume level:
* Try in FREE version. Audio normalization is fully supported in configurations
Read how it works >
To normalize the average loudness (ITU-R BS.1770) of audio files:
* Try in FREE version. Audio normalization is fully supported in configurations
Read how it works >
The headroom (as rule 6 dB) is used for stable work of sigma-delta modulator (DSD encoder).
We can imagine audio signal as water in a pool.
We can pour water into the pool until its level achieves the pool edge.
After it, the water level may not be increased via the water pouring.
It's called overload or level clipping.
An audio signal (music) is like a wave on the pool surface.
If the wave is too high, it will pour out of the pool.
So, we can limit the wave height, to avoid water loss due to overflowing (overload).
The limit is called "headroom" for audio.
Sigma delta modulator due to overload can lose stability. It can be heard as continuous tone. Stability can be restored by resetting the encoder.
However, sometimes loudness increase is possible without instability of DSD encoder.
If input audio signal has an amplitude near maximum, upsampling can cause overload.
Let look at the picture:
After upsampling (oversampling) between real samples of original signal appears "virtual" samples that appear due interpolation.
Amplitudes of some "virtual" samples can be upper than original samples. If peak level of original samples near maximal allowable value, "virtual" samples can be upper the value.
It can cause overload.
For avoiding same overloading during audio conversion need perform correction of output file levels.
Such conversion executed in 2 stages:
1. Measurement real output peak levels after audio conversion
2. Correction levels of output files.
Example 1 (normalizing audio with overload avoiding):
Output file #1 have peak level -1 dB.
Output file #2 +2 dB.
Maximum peak level here +2 dB.
For avoiding overload need decrease level of output file #2 at 2 dB. It will have peak level 0 dB (= +2 dB - 2 dB).
Output file #1 level need decrease at 2 dB too. It will have peak level -3 dB (= -1 dB - 2 dB).
Example 2 (normalizing audio with loudness maximizing):
Output file #1 have peak level -1 dB.
Output file #2 -2 dB.
Maximum peak level here -2 dB.
Here possible increase level of both output files. Output file #1 at 1 dB. It will have peak level 0 dB (= -1 dB + 1 dB).
Output file #2 level need increase at 1 dB too. It will have peak level -1 dB (= -2 dB + 1 dB).
Sometimes need total normalize peak loudness for several albums. Album here is either folder with songs or ISO-file or CD.
AuI ConverteR found peak levels for all tracks of all albums.
There are 2 ways maximizing loudness with avoiding of overload:
1. Manual
Volume don't auto adjusted during conversion. Read more...
Analyzing level of output audio files without recording it to hard disk. Results will displayed at table in Levels tab of main window of AuI ConverteR 48x44.
To display average level (EBU R128),
in Settings > BWF > check "Calculate loudness EBU R128 (ITU-R BS.1770) [WARNING: works slower]".
NOTE: The checked state cause lower speed of level calculation.
3. To max peak
Auto maximizing peak levels output files without overload. How it work see here and here.
4. To average
Auto normalizing to average level (EBU R128). Read how it works.
AuI ConverteR has several price/functionality options that allow saving your budget for the required functionality. You can expand Modula-R abilities later by purchasing new modules. Unlimited-time license, major updates (... 10.x > 11.x > 12.x > ... versions) are paid.
* All prices on this page are in U.S. dollars without V.A.T. and other applicable taxes and fees. The prices are recommended. Information on this page is not a public offer.
[MINIMUM PRICE] for full functionality of AuI ConverteR, in comparison with Modula-R edition that is including all modules.
![ALL-OPTIONS are included [PROduce-RD edition]](/soft/images/full-functionality2.png)
Decription
All abilities are included in PROduce-RD version. It allows saving budget, comparing purchasing of the full abilities (all modules) of Modula-R edition. Unlimited-time license, paid major updates (... 10.x > 11.x > 12.x ...).
[MINIMUM PRICE] configuration for convert
FLAC, WAV, AIFF, mp3, other PCM (up to 192 kHz, 24 bit, 32 bit float)
to
FLAC, WAV, AIFF, mp3, other PCM (up to 192 kHz, 24 bit, 32 bit float) with auto level normalizing [consists of 2 modules]
![Normalize PCM audio files [Modula-R edition]](/soft/images/lev-norm-pcm-files.png)
Decription
The configuration allows normalizing PCM audio files.
Read the user manual...
The configuration abilities may be expanded by purchasing other modules.
Unlimited-time license, paid major updates (... 10.x > 11.x > 12.x ...).
[MINIMUM PRICE]configuration for convert
DSF/DFF (up to DSD128), FLAC, WAV, AIFF, mp3, other PCM (up to 192 kHz, 24 bit, 32 bit float)
to
DSF/DFF (up to DSD128), FLAC, WAV, AIFF, mp3, other PCM (up to 192 kHz, 24 bit, 32 bit float) with auto level normalizing [consists of 4 modules]
![Normalize DSD audio files [Modula-R edition]](/soft/images/lev-norm-dsd-files.png)
Decription
The configuration allows normalizing DSD audio files (.dsf, .dff).
Read the user manual...
The configuration abilities may be expanded by purchasing other modules.
Unlimited-time license, paid major updates (... 10.x > 11.x > 12.x ...).
Remarks:
Allowable maximal peak level of signal encoded to DSF can cause stability troubles sometimes.
Stability depend on impact time of maximal level to input of DSD encoder.
If AuI ConverteR have stability troubles to you files recommended use additional attenuation: in Settings window (push Settings button in main window) > General tab need set for switch Sigma delta modulator attenuator value discrepant zero.
More attenuation (more negative number) - less probability of breaking stability of DSD encoder.
However, it decrease loudness of converted file too.
You can set auto restoring of average default level (reset).
In Settings > Interface > check "Save level normalization state after restart" switch.
In Settings > General > set value of "Default normalization level, dB [average normalization audio]" input field.
In the first approach, normalizing like any gain control does't affect audio quality. It the gain control is done properly.
However, level decreasing degrade signal/noise ratio for sound recording. This could have implications for very quiet music fragments.
Read more...
Normalizing audio do level equal target value. In instance, all album tracks have different average levels. We are select level that we want (target level). And, make equalization of all track levels to target one.
Read more...
Audio normalization is elementary procedure - gain altering. If it done properly, it don't cause distortions that we can detect even via analysis.
Read more...
For some applications of professional audio, requirements exist.
As example:
TV advertizing should provide certain level. Otherwise, the advertising may sound too loud.
For DSD audio, peak level should have headroom -6 dB. Otherwise, it can cause overload and broken stability of some DSD modulators.
When you listen playlist, different tracks may have different levels. It can cause discomfortable listening. You can equalize all average levels of the tracks, and listening may become more comfortable.
Read more...
Yes. VLC player can normalize audio. More...
To normalize audio volume, average-level normalizing is required rather. Read more...
To "flatten" or normalize audio volume level, use normalizing by average level.
Read more...
As far as is known, YouTube normalize too loud audio. Read discussion...
You can compress audio, apply equalization (EQ), reduce noise, etc.
Peak level of an audio track should not be above 0 dB. Sometimes, headroom (peak level reserve) is required. In instance, -3 dB is target maximum peak level.
Read more...
January 30, 2023 updated | since October 31, 2015
