VideoHelp Forum
+ Reply to Thread
Results 1 to 9 of 9
Thread
  1. i want to make the audio in a movie louder by using audacity and then exporting the audio as aac. someone told me that i can do that by first applying the normalize effect and then applying the compressor effect

    i know what to do in the compressor tab. keep the default values and tick "compress based on peaks"

    i am unsure what to do do in the normalize tab. i need to set a value for "normalize peak amplitude to". the default value is -1.0 db. that same person told me to use -0.1db

    i did some comparisons for the wavelengths

    https://forum.videohelp.com/images/imgfiles/CSx47N7.jpg is the original wavelength of the movie

    https://forum.videohelp.com/images/imgfiles/eFrm7eh.jpg is the wavelength with peak value of -0.1 in normalize. the wavelength looks the same as the original

    https://forum.videohelp.com/images/imgfiles/tinEIVy.jpg is the wavelength with peak value of -1 in normalize. the wavelength looks smaller than the original

    i want to ask what did normalize by using -0.1 do? did it make the audio louder?

    what did normalize by using -1 do? did it make the audio quieter because the wavelength is smaller?

    is there a better combination of effects that i can use to make the audio louder like for example first amplify without allowing clipping and then compressor?

    is there the need to use normalize?
    Quote Quote  
  2. I'm a Super Moderator johns0's Avatar
    Join Date
    Jun 2002
    Location
    canada
    Search Comp PM
    The peaks will cause the wave not to change in volume due to the fact the peaks would clip if they increased in volume.You could try loudness normalization but softer sounds will get louder as well.
    I think,therefore i am a hamster.
    Quote Quote  
  3. Originally Posted by johns0 View Post
    The peaks will cause the wave not to change in volume due to the fact the peaks would clip if they increased in volume.You could try loudness normalization but softer sounds will get louder as well.
    i didn't understand what you said. what is the difference between -0.1 and -1?
    Quote Quote  
  4. Originally Posted by cns00 View Post
    i didn't understand what you said. what is the difference between -0.1 and -1?
    You should really google by yourself and not always ask us holes in the belly.
    Quote Quote  
  5. Originally Posted by ProWo View Post
    You should really google by yourself and not always ask us holes in the belly.
    i did google. i couldn't find anything
    Quote Quote  
  6. Originally Posted by cns00 View Post
    what is the difference between -0.1 and -1?
    0.9 decibels. That would be barely noticeable to the listener.

    https://en.wikipedia.org/wiki/Decibel

    The setting in Audacity's Normalize filter sets the desired peak volume level (below 0 dB which is the loudest a sound can be without distortion). Your source's peak is already near -0.1 dB so normalizing to -0.1 will not change it much. Normalizing to -1.0 dB will decrease the overall volume.
    Last edited by jagabo; 1st Sep 2022 at 18:52.
    Quote Quote  
  7. Compress it dynamically
    Quote Quote  
  8. Note: your images are "waveforms", not "wavelengths".
    Quote Quote  
  9. Member Cornucopia's Avatar
    Join Date
    Oct 2001
    Location
    Deep in the Heart of Texas
    Search PM
    Yes, the minimum difference that most people can tell as a "change in volume" is 3.0dB.

    To get a noticeable difference in most recordings, normalize CAN be used when the overall highest peak (of the WHOLE clip) is well below the digital clip level (aka 0dB). Thus all parts of the WAVEFORM are shifted upward (both quiet and loud sounds) to bring the overall level higher, and by nature of its reference to the peaks, bringing the overall highest peak up to, or near 0dB. If this is a raising of the level of 3dB-5dB it will be barely noticeable, at 6dB-10dB it will be noticeable. Above that, it will be VERY NOTICEABLE. Remember that it is equally raising all parts of the clip by an equal amount. This includes the noise floor.

    A signal that has average level of -12dB, but with occasional peaks all of which only reach to -1dB will only change to an average level of -11dB with occasional peaks which reach 0dB if you attempt to use "Normalize to full 0dB". And again, this is a minute and not discernable difference.

    If you want non-equal operation on different level areas of the signal (for example you want quiet sounds to stay quiet, but medium and loud sounds to get much louder, or you want everything to get louder EXCEPT the very loudest sounds which you want to keep the same), you must use the dynamic "compressor" family of processors: Compressor, Peak Limiter, Peak Unlimiter, Expander, Noise Gate. Possibly also in concert with forms of EQ and band filtering.

    The person who told you to use -0.1 doesn't know the whole story. There ARE other reasons for not going fully to 0dB, but I won't go into them now, and for your purposes, it would make no difference to you.

    If your signal originally had peaks that maxed at 0dB, "Normalize to 0dB" would do nothing. And "Normalize to -1dB" would lower your overall level (only very slightly). It may be a visible difference in the waveform, but maybe not.

    Remember, the waveform shows the signal levels LINEARLY (at least in most simple, consumer based apps), but our understanding of sounds and listening is LOGARITHMIC (thus the use of the dB scale).

    FYI, WAVEFORM = "the image of a wave". WAVELENGTH = the length of a wave's cycle of peak-to-trough-to peak (which can vary, and it is inverse to its frequency)


    Scott
    Quote Quote  



Similar Threads

Visit our sponsor! Try DVDFab and backup Blu-rays!