Audio compressor

[Zeto]

I think that the show would be a better experience if it was run through some audio processing software before it was released. The lows are pretty low, and the highs and laughs are pretty loud. I think that, at the very least, a compressor would help immensely.

If none of your crew know about compressors, basically they take audio signals and do a pre-amp on the input in order to make the quiet parts more audible, and then you set a loudness level you wish to compress, and the loud parts ideally won't go over whatever you set it at. This process is also called normalization.

I love the show. I think it's fun and informative, and I think it could be even better with some audio compression :)

Zeto

[Amanda]

I'm not REALLY sure what Joe does, but I think he does run the audio through some sort of program first.

[Christopher Masto]

I agree, it can be a bit hard to listen to. I get the impression that the levels are pretty low going in. I think it would help to have an AGC or analog compression before it hits the computer.

[Amanda]

We purposefully set the levels low on the mixer (if that's what you're talking about). Not sure why, but Joe said to do it that way. Since we've gotten the new equipment, we're still trying to figure out what settings to use. I think we got it pretty good a few episodes ago, but do you think we wrote down what we used? Nope. *Head slap*

[Joe]

Audio is an ongoing problem for me. I'm trying to get things as level as possible, but I also don't have a ton of time to spend on it.

I use Audacity and usually run the Leveller on the audio first. I'll try adding in some Normalization too this time around and see how that works out.

Also, I've always though you want to keep the audio signal low so you don't max out the levels. I may be erring too much on the low side, but it's that better than being too high?

[Christopher Masto]

Is it possible you're recording at a low bitrate or converting it down somewhere in the process? I just listened to the most recent episode and it sounds a bit.. fuzzy? muffled? I'm certainly not an audio engineer, so I don't know exactly how to describe it, but there's some detail lost that makes it a bit hard to understand, especially when Brandice is speaking quietly.

I think you're using the same mics as me, and although I'm not getting supah-awesome quality either, I listened to the "TeaTime" I posted recently for comparison, and I think it's a bit clearer-sounding in some way. My audio path is rather pathetic, actually; I plug my microphones into a Y adapter and into my camera's 1/8" jack, and just leave the AGC on. I don't really have the hardware to do it any more professionally.

[Zeto]

Input amplitudes should be set such that the loudest sound during a recording is as close to 0db as possible, using the largest range of amplitude available, as the signal to noise ratio is best at this point. That might be another reason why things sound the way they do. I know it might be counterintuitive, but you want levels to be as high as they can be without something called "clipping." Usually you can test inputs on the fly with your recording software or directly on the mixer using a 'test' talk. Usually the values output as lights (green, yellow, red) where you want two things... the levels should be poking into the yellow at the loudest parts without audible distortion or clipping.

A normal postprocessing chain goes something like: remove static lows and highs from ground and background noise (usually there are built-in software filters for this,) normalize all volume to approximately 95% (maximizes the gain for all sounds,) then use a compressor to up the quiet parts and soften the loud parts, and finally EQ it if necessary, as not all codecs record all ranges equally, and sometimes male low tones get drowned out.

But, again, awesome show and it's nice that you have some new toys to play with and learn about :) sound engineering can be a pain, as it's really an art in and of itself.