I was wondering why I'm often bristling at some of the audio I hear from the internet. I guess it's not easy to get audio as good as video - is the easy answer: People generally don't want to turn down settings if they get a peak, because when there are no peaks, the sound is too low. A more correct answer is that I'm a bit of an audio Nazi and am used to traditional broadcast quality. Some of my fanaticism extends to Giantbomb and seems to afflict the East more than the West (ironically). As much as I love watching every GB video regardless of region, I'm often waiting for the inevitable distortion peaks and general "close mic hot" sound from GBeast.
The mics used are excellent (if they are the Elecrovoice EV20 - about GBP 500 a pop!) and posses an interesting feature called "Variable D". It is designed to give the microphone a more useful dynamic range while retaining a neutral tone (i.e. not sounding bassier as you get close to the mic). It's also why these mics have an unusual slotted grille running down their length.
I think perhaps this may be causing problems in the GB East studio. Rick Bell, the "EV mic guy", has made a video about this feature and demonstrates the principles of cardoid patterns, what Variable D does, and how axis and proximity affect the sound for various mics in the range that either have, or don't have "Variable D". It is well worth a watch to see how moving about in front of a mic affects how it sounds:
Rick mentions "working the mic" many times during the video, which suggests control over the sound is most affected by distance and axis to it, as he demonstrates. I remember being previously impressed with Austin Walker's vocal performance on GiantBomb, and his natural ability in "working the mic", since he rocked away from the mic when yelling and closer for more hushed voices, as most professional singers would control their instrument. It seems the practice with the current crew is to get as on axis and as near to the front of the mic as possible - something not really necessary in a studio, unless you wish to block out a lot of background noises, while recording your mouth noises.
Maybe the "problem" with variable D is it attempts to preserve the tonal qualities of the mic until it overloads, while a cardoid without this feature would start to get overblown in the bass, as a "warning" that the capsule was becoming overloaded. Therefore a skilled broadcaster would back off naturally (through monitoring output) before their audio went into distortion, or an engineer would intervene at a more energetic bass overload. The alternative for these EV20 mics is they harden up near their limit, especially with voices like Alex and Dan, (whose voices are at the mid-range end of the spectrum). In this way, Variable D might be analogous to going straight into a skid without first getting any sign of over-steer in a fast car. Just owning expensive microphones might not be enough for great sound, in a similar way.
It is also a human temptation, with a mic in front of you, to move closer until you kiss it (especially when you are not monitoring your voice). Carefully positioned pop shields can act as a good barrier against this, but do clutter the screen. I always thought the choice of this large type of mic for broadcast was an odd one, because of the tendency of presenters to constantly fondle it, the extra technique required for head placement (especially in a very close-mic'd set-up), and also because it's imposing bulk tends to obscure most of the broadcasters' faces as it constrains them. This seems rather counter-productive for a video format, especially when headphones are incorporated into the picture too.
Old-style TV studio broadcasting has never done this for good reason: Lavalier style mics are often better for broadcast because the performer forgets they are there, they require no "technique", they are off-axis to breath and mouth noise and they are discreet for an audience. They are also capable of excellent full range vocal audio. The GB West videos with such mics are far more pleasant to listen to for me (except for their removal and fitting noises) and there isn't the "hot" close mic sound, as if they're licking your ears: The close mic emphasis on "peh", "beh" and "teh" sounds can be fatiguing to hear and are more products of breath (put your hand 1" in front of your mouth, talk and feel your breath on it) and therefore unheard in normal conversation distances, but picked up on a close mic capsule.
The other human condition is of reticence at a sound-check, so the volume levels of any subsequent mindless japery and tomfoolery are not tested. The art here is for the sound engineer to coax each broadcaster into a re-birth regression scenareo, so that unrestrained screams of new consciousness remembered can be used as a useful guide to setting maximum levels conservatively.This is where the job of the GiantBomb engineer is hard, because a singer or broadcaster typically have a defined range, whereas reaction style content demands capture of a broader range of speech.
It's unlikely that broadcasters will wish to change their own style to sound great, so we need technical preparation instead: Pop shields can enforce a minimum distance. Reducing the microphone's height (with lower mic stands) may work too in requiring performers to go slightly off-axis, as might setting them further back on the desk (unless presenters stoop down towards them). Although audio levels would have to increase with distance from mic, the range of acceptable gain bandwidth would more than compensate.
Lavalier mics would be much better for video and audio stability and fixed distance and axis, but would create more of an issue with studio noise and audio bleed-though of games played, unless headphones are worn, or studio sounds kept low. The ear mounted mics used in E3 presentations were sometimes capable of good quality, but presenters seemed really conscious of wearing them and they often fiddled and moved the capsules closer to the paths of their breath.
Perhaps adding suitable compression, or changing existing compression settings might work, if the distortion peaks aren't from mic capsules getting overloaded. Then, the distortion is in the pre-amp circuitry and created by excessive voltages from the microphone. There are some rather clever compression tools around these days that can auto-adjust for this sort of thing. They exist both as parts of hardware audio interfaces and as software tools.
But this still means the "close mic" sound will remain unless a few inches can go between mouths and mics, and off-axis helps too. This, in turn means upping audio levels to get the same gain, allowing for a greater overload potential if the same sound levels come into close proximity the mics, but a wider loudness range before distortion. These audio problems should disappear in a steady-state studio environment, once a "butter zone" is found. Everything, from a sigh to a scream, will be preserved in all it's audio glory, with no overloading at all, for my listening pleasure.
I only write this because I love GB content so much and think I'd be able to love it even more if the audio was consistently as brilliant as their video quality and content ideas. Big props to all the work the GiantBomb staff do for us viewers! I mean this with the best of intentions and knowing I've got into trouble for writing about these audio problems before.