getting correct volume of narration

[webberphan]

I have Sony movie studio and have created a MP3 file of my narration. I also have Sound Forge audio studio.

I have converted to an MP3 file and would like to normalize it but would like to know what settings I should use? -3db? I've seen some recommend only -.1 db

Any advice would be appreciated.

[johnmeyer]

The presets are -10dB for speech and -16dB for music. If you use -3dB or -0.1dB I think you'll find a lot of distortion and clipping.

Have you tried it? It only takes a second or two to modify 10-15 seconds of audio and listen to the result. What were the results of your tests?

[webberphan]

Yes, I did
Here is a picture of the waves before normalization, as recorded

and also one after normalization which seems much too low. I am making a youtube with slides and video with narration under it. Will this be loud enough?

[JVRaines]

The problem is what's called intersample peaks. If your samples are very close to 0 dBFS, they may be rendered higher than zero by a digital-to-analog converter, producing clip distortion. I find that peaking no higher than -1 dBFS is usually fine and there is broad consensus that -3 dBFS is a conservative limit. Where is your audio file headed? In some contexts, such as broadcasting, the average level should be much lower according to loudness regulations.

EDIT: Now I see you are going to YouTube. If you normalize to -1 dBFS everything will be fine.

[webberphan]

Thank you, I'm recording on my Movie STudio/or Sony Vegas program (depending on which computer I'm on) so if I set voice to -.1 db for volume, I should be good?

[johnmeyer]

You didn't mention what normalization level you used. Your "after" looks pretty weak to me (I assume the second pic is after normalization).

What is the nature of the transient spike? If it is some sort of pop, clank, or other similar impulse event, you might want to remove it prior to doing normalization. Sound Forge's de-click module might be able to handle it.

If you want to upload a few seconds of the before, I'd be happy to take a quick look at it and make further suggestions. I too use Sony Vegas, Sound Forge, and also iZotope RX. The latter is quite expensive but is a very professional restoration tool that might be able to do a little more with your file than the Sony products, depending on what other issues you are dealing with.

[JVRaines]

Thank you, I'm recording on my Movie STudio/or Sony Vegas program (depending on which computer I'm on) so if I set voice to -.1 db for volume, I should be good?

My recommendation is -1.0, not -0.1.

[webberphan]

Sorry I did mean that -1.0