@spuddy13

Hi Brian,

I think that yours should be the best question of the year!

Ever wonder what all of those audio type people are doing in films? Watch the credits for voice-over artists, recording engineer/soundman, cleaning/sound improvement specialist, mixing engineer, mastering engineer, and probably a few more. Well, for home video making, that's all you.

Where to start? How much time do you have?

Here is somewhat of an overview:

1. Equipment - good mic, audio interface. Recording environment - no outside noise, limited reverb from the room, and limited room noise - watch for noisy equipment, like fans.

2. Recording and Levels - adjust the recording level in Windows if necessary.

but my recording outputs are all different because i do over many days and i talk different and i am nearer or further from mike

You know the answer - find the sweet-spot, measure the distance mic to mouth, and try to repeat it, practice. Keep your mouth close to the mic, use foam on the mic and a pop filter (note to self, get a pop filter).

Before recording a new session, listen to previous recordings through headphones connected to the audio interface, mixed with live sound (your mic) and what comes from the computer. Adjust your voice to get it to sound as close to the previous session as you can. This way you can hear yourself.

Watch the levels on the meters. Keep out of the red zone. Try to keep your voice even.

Watch some videos on voice over to see how the experts do it. Important things like recording standing up (I don't), breathing exercises, throat clearing, drinking something or not, etc.

Look at the waveforms to see if you're getting a couple of very loud words per recording, but most audio is much lower. Try to avoid that. Look at your text. If there are words that you want to emphasize, doing so straight into the mic will ruin the rest of the audio or you could get distortion. The trick singers use (my wife) is to move the mic away from the mouth, in your case, you can move back a bit for louder words, or, as I try to do, turn your head slightly off centre of the mic.

I almost forgot, make sure that you're speaking into the proper part of the mic. Mine is on one side, not at the end.

Don't try to read too much in one shot. I can get 2 to 4 sentences out before making a mistake, so I limit each recording to about that. And I need to breathe.

3. Post-production

Normalize. I used to use the video editor (VPX) for recording narration, now I use SF Audio Cleaning Lab (ACL), one of the best or most useful Magix products no longer available. You have Music Editor 3 with Video Deluxe that is an older stripped down version of ACL that you can use for levelling out audio as shown below.

I normalize in ACL to 90%, which leaves the peak sound at almost 1db below maximum (0 db is maximum). In VDL, you can only normalize to 100% and then lower the middle handle of the audio by 1 db.

Look at the waveform in the image below. You can see a peak at the beginning - a loud sound - and another one further along.

What I do is go through each recording and reduce the peaks using the volume curve as shown below. You can do the same in VDL. The difference is that in ACL, the waveform changes as a point is lowered or raised, and I can see by how much.

Reducing the peaks allows for increasing the overall volume.

Here is what you would do in VDL, I'm using VPX.

Normalizing problems: if one recording has one very loud sound and the other doesn't, but the rest is all of similar loudness, then each recording will get a different level after normalizing. The first one may be raised by 1db, the other by 6db. This is probably part of the cause of your uneven levels. You need to find an optimal level and then use that for adjusting all recordings, but only after reducing peaks using keyframes to avoid clipping.

How to avoid clipping? You need something that shows you actual levels and where problems are. Download and install the free version of YouLean Loudness Meter and learn to use it. You add it as an effect in the mixer, for the narration track FX, and for the master track FX.

Here is a good introduction to YouLean:

When you play back, it will display levels, show you where peaks are, and the true peak value, which you need to keep below -1.0 db to avoid clipping. With that, you can see which peaks need to be reduced and if you can increase the overall volume of the clip.

DeNoise: Now is where I do some cleaning, usually just denoising to reduce that background noise.

Compression: Now I add in compression to make the narration more even. Ratio 2:1 or up to 3:1. Beyond that, it starts to sound like radio voice, no nuance. You can use the compressor in VDL for the track, not the one in cleaning for an object.

Careful here, as some compressors do not make up the gain. A compressor compresses the audio between the peak down to the threshold. This reduces the volume of the louder parts, bringing them closer to softer parts. So you may have to increase the volume to raise the overall level. However, raising the level, also raises the level of background noise, so monitor that as well.

Levelling between objects: You still may have to do some levelling as some recordings may still sound louder than others. Here you use your ears and YouLean Loudness Meter on the master track. You can play back each clip, note the maximum and other parameters, and adjust the volume of the clip using the centre handle of the clip on the timeline.

Note that the YouLean LM is pre-fader, so the volume that it hears is what is before the track fader or master track mixer fader. Once you have everything balanced out, you may want to reduce the master track fader so that nothing exceeds -1.0db, or as many users do, adjust the fader so that the maximum sound is at -3.0 db.

VoiceLeveler: There are some plugins out there that do voice levelling, like ERA6 VoiceLeveler from Accusonus. It's free somewhere on the internet, as part of a free kit of audio plugins. I haven't used any of these.

Plosive Remover: I have trouble with plosives - p's and b's - so sometimes I apply the Accusonus PlosiveRemover.

It all takes work, but a good video with bad sound will be bad. A mediocre video with good sound will be better.

I didn't mention EQ, which is another thing that you can do to improve your narration, even if it's just removing the low end, 0-50 Hz with a low pass filter and a de-esser preset. Example below:

If you have any annoying nasal sounds, you can find and remove them with a good equalizer, like the one above. There are plenty of tutorials about how to do this on YouTube.

I also didn't mention fixing other audio, like that with video clips, or adding music and sound effects, and mixing it all together so that it sounds good. That adds to the complexity.

There is more.

John CB