[Sound]: How do you normalize/equalize?

[Maglok]

With SC2 coming up I am dusting off the old skills. SC1's sound editing tended to be optimized for 22.5khz mono playback. It had it's tricks, it had it's uses. What I've been hearing from the campaign bits of SC2 is that 44.1khz stereo is not an issue. Most recordings will still be 44.1khz mono o course.

So what I'd like to know is what kind of things do you apply to a recording to get it ready for any potential effects. Not including noise reduction.

I am thinking what kind of equalizing, dynamic processing presets, etc.

Of course every recording is a little different, but I am still trying to nail a good baseline with all these 'new' techniques around. A consistent one you can use to make sure recordings are clear during gameplay.

[Legion]

I always run a compressor over audio files that aren't loud enough. This does make it sound a bit unnatural, but it works well for protoss and zerg.

I've recently decided that I like normalizers as well ( a bit ), though I usually apply those after I've applied all effects.

Interesting topic, Maglok! Thanks!

[Maglok]

The nasty thing about compressors is that eventually they make everyone sound the same. If i compress using a compressor (which I do) I tend to be less able to produce different sounding voices. It's annoying. Then there is the equalizing, which is always... chaotic Nothing ever works 100%.

As for normalizing, you do it after effects. I wonder if that is 'proper' or not. Sure you want to normalize after effects, but might you not also want to normalize before sometime? There is a 'rule' in audio editing that revolves around always using the loudest recording you can make. If we can normalize before effects, shouldn't we do that to create the loudest recording? Without it clipping our effects afterward of course. Stuff like that interests me, what do you apply first.

[Maglok]

Here's what I mean:

3 samples
- Original
- Normalised to 95%
- Dynamic processed then normalized to 95%

[IskatuMesk]

Compressing? Is that the same as Dynamics Processing?

When I first record, even with my good mic that picks up almost no background noise, I always do a noise reduction first. Selecting an "empty" part and then making a profile and removing it from the rest.

I then throw on my effects first.

After my effects are in, or inbetween depending what I'm trying to do, I then use Dynamics Processing. I have like 5-6 profiles that suit my various needs. My DP effects tend to be extremely heavy to make up for my "accent" in which some words are too quiet and others too loud. It evens them out rather effectively. The side effect is that if there is any static or noise, that often gets amplified as well.

I use the DP effects to some extent on other people's voices as well, but these ones suit my voice the best.

After everything is done it then depends. If it's straight dialog, I normalize the entire thing. If it's unit responses, I normalize individually.

If it's a flat-out recording, like dialog for my Captain Planet RPG or a webshow for example, I don't use DP at all. Just a noise reduction/normalization and that's it.

[Maglok]

If you disregard what it is technically, a DP usually whirls down to the voice becoming easier to understand in a crowded environment like a SC map. Sound effects going off, background music, etc.

I find it interesting you apply DP after effects. Since DP makes the waveform 'louder' so to speak and generally people encourage you to work with the loudest recording you can. Though I definitely know that you can apply effects first, the difference in applying DP before or after is astounding. I wonder what is 'industry standard' so to say.

[IskatuMesk]

The difference is astounding. You probably know a lot of my voices use extremely heavy effects, which can sometimes "dampen" the voice and make it harder to understand. The DP helps alleviate this.