kw.media » Blog » The Audio Guide to Happiness, or: How to make your Streams & Videos sound good

The Audio Guide to Happiness, or: How to make your Streams & Videos sound good

Audio is the most important part to get right of a video. This may be counterintuitive at first – it ought to be the actual visual part (ie, the video), right? – but: A lot of viewers watch your content on mobile, and on mobile phones, 480p often is clear enough, especially if you are on mobile data.

This guide is mostly a beginner’s guide, so it just goes over “what mic do I even need as a game streamer?”, “where do I place my mic?”, “what should my recording set-up look like” and just one of the many possible filter chains you can set up in OBS. Which doesn’t mean that it’s not somewhat in-depth at 2000+ words long.General

Before we get started with the deep techy part, let’s start with something fairly obvious: Garbage-in, garbage-out applies to audio. Buying an expensive mic may make the audio sound better, but it won’t make you any funnier, fix your pronunciation, make you talk more or make you less rambly. In order to get better at those things, you’ll need to practice. This practice automatically happens just by making videos, but I’d highly recommend actively looking to improve these areas, as that’ll make the learning (and thus, improvement) go much faster.

Alright, let’s actually talk about mics now.

Which mic to buy?

In this section:

  • XLR vs USB
  • Pickup patterns
  • Frequency response

To get the TL;DR first: 
Any mic between 50 and 300 USD probably is gonna work fine for you. This even goes for professionals; the Shure SM58 has been the industry standard for vocal performances since the ’60s – and you can get that one for 100 bucks or so. This isn’t to say that you should all go for the SM58, as it is an XLR mic and thus requires quite a bit of extra hardware (though I won’t judge if you do go for it).

XLR mics as a whole are a bit of a minefield. While in theory, you can adapt an XLR mic to your average 3.5mm jack, in practice, you’ll probably need a preamp to use them, and in the case of a condenser XLR mic, you’ll need a preamp that can deliver phantom power.These things aren’t impossible to figure out and do offer a lot of flexibility once you add multiple devices to your chain, but if you are a sole gamer just looking for a good mic, they add a lot of unnecessary complexity when you can have all of that in one convenient package as a USB mic.

As for USB mics, every gaming company and their dog makes one these days, most of them somewhere in the 100-150 bucks price range, marketed “for gamers”, “for streamer” and/or “studio mic”. They all come with varying degrees of features, most of them being somewhat useless, such as switchable polar patterns.

polar pattern is from what angles the mic picks up noise from, and while it can be nice to be able to pick up sound both from the front and the back of the mic if you’re interviewing someone across the desk, you’ll realistically probably have a wall or your screens behind your mic anyway, so you really only need a good cardioid pattern and won’t benefit too much from being able to switch around.

Various polar patterns; the mic is pointing towards the top of the page. Image credit: Galak76 & Nicoguaro at Wikipedia, CC-by-sa 3.0; CC-by 4.0

That said: If you are a fan of clicky keyboards, cardioids will pick up your keyboard quite well unless you have the keyboard placed directly behind the mic. If you can’t do that, you may want to try your hands on shotgun-type mics, i.e. ones with a super-cardioid or lobar (“uni-directional”) pickup pattern. Further, if your videos involve you being outside somewhat often, a shotgun or even a lav mic suddenly starts being way preferable over a studio mic as studio mics don’t fare too well outside.

Now, I still haven’t answered which mic I actually recommend, and that’s because there isn’t a definitive answer that’s more concrete than “a cardioid ‘studio’ mic is probably the right tool for the job”.

For example, Blue’s USB mics tend to be somewhat flat but respond super strongly to the proximity effect, while Røde’s USB mics tend to focus the mids more, so what sounds amazing for one person may not sound good for your voice (or in more formal terms, different mics have different frequency responses).

Frequency response of the Røde NT-USB. Being a cardioid/pressure gradient mic, the low end falls off a bit. The bump in the mid-highs is there to add clarity – whether this sounds good on you I can’t tell though. Image credit: Røde

So really, the ancient wisdom of “try before you buy” applies here. Or at least “read reviews and listen to comparisons before you buy”. If you already have a mic, do share in the comments what you use and what you like and dislike about it – I can imagine that hearing from other creators is gonna be quite helpful here.

What to look at when recording?

In this section:

  • Mounting (mic arms)
  • Pop filters
  • Critical distance
  • Sound treatment & insulation
  • Proximity effect

So, you now have bought some kind of mic, and if you bought one of the more common ones, it probably is standing somewhere on your desk in front of you. And while that probably already works somewhat well, it is somewhat problematic: Your mouth is now further away than your keyboard. So it’ll pick up a lot of keyboard noise – regardless of which type of mic you bought – and not that much of you yet.

Ideally, you want your mic to be placed

  1. an elbow’s length away from your mouth at most, while being
  2. almost but not directly in front of your mouth and
  3. as far away from any other noise as possible.

Outside of an actual studio, this probably isn’t achievable, but you can buy a mic arm which raises your mic up so that it sits on your head level and not nearly as close to your keyboard.

The reason why you don’t want your mic right in front of your mouth is that sound is tiny wiggles in the air, while blowing is much, much, much more powerful – try raising the volume of your speaker until it can blow out a candle. And plosives (that is, “p”-noises) are pretty much just blowing. So in order to avoid blowing at the mic, you do want to keep it a bit to the side from your normal speaking direction. In addition to that, you also can buy pop filters (or socks and some wire) to add some extra diffusion in case you do speak directly into the mic.

The critical distance is the distance at which the direct sound (i.e., you speaking) becomes quieter than indirect sound (reverb, echo). The “elbow rule” from above is a good rule of thumb to stay comfortably within this critical distance almost anywhere, but it of course depends on the pickup pattern of the mic you’re using, as well as the room you’re in – a room with a lot of laundry in it has much less reverb going on than cold, hard tiles. Now that’s an excuse to not clean up your room!

If you want sound treatment but don’t want to live in a messy room, you can also buy various kinds of things which are meant for that purpose. In general, anything that’s soft and squishy will help reduce reverb and eat any sound that comes close to it: carpets, couches, cushions, curtains, the heavier, the better, the more, the merrier. Hanging clothes over bookshelves helps, as well, but of course actual acoustic panels, specifically built for the purpose is the best you can buy. You almost can’t have too much sound treatment outside of a few use cases (eg if you want music to sound more like a live recording rather than a studio recording). Sound proofing (preventing sound from outside getting in) is less important, but can be necessary if you live next to a noisy road or otherwise have a lot of noise outside.

The proximity effect is the effect that causes the bassy part of your voice to be amplified the closer you are to the mic (on pressure-gradient mics anyway, which includes most of the mics that you’ll probably be using). If you are within licking distance, you’ll sound like a radio host, if you go too far away, you’ll sound somewhat tinny. Again, the elbow distance is a good rule of thumb as to where it sounds mostly balanced.

Software

In this section:

  • Denoising
  • Compressing
  • Clipping & limiting
  • Mixing

For software, I like to have a chain of denoiser → compressor (→ equalizer) → limiter in OBS. OBS luckily comes with those things built-in just click the gear next to the audio source and select “filters”, so I’ll just use that as an example. It works both for live and recordings btw – if you use this chain in your recording, you generally won’t need to do too much cleanup after the fact.

The first thing is the denoiser. If you are in a fairly quiet space, you may not need this one at all, but if you need to have a fan running, the denoiser can be fairly helpful. Don’t turn it up all the way until you don’t hear any noise, as that will make you sound as if you’re sitting in a fish tank.

OBS’ noise suppression goes from “full fish tank” (-60dB) to “no noise reduction at all” (0dB)

If you have some heavy noise going on around you, you can instead also try using RTX Voice (which also works on GTX cards with some tinkering) or krisp (which is a paid service). Both of them have in common that they do AI “magic”, which may catch more noise than the standard OBS denoiser does. Alternatively, you can use the ReaPlugs VSTs, specifically ReaFIR – in “subtract” mode, it can create a noise profile and denoise based on that specific profile, which also generally introduces fewer artifacts than OBS’ denoiser. ReaPlugs works on pretty much any hardware, all the way to Windows 98 and Linux (via WINE) machines.

Though again: If you can avoid background noise altogether, go for that instead of trying to fix things afterwards.

Next up is the compressor. The compressor brings loud and quiet parts closer together, so that you won’t cause bleeding ears after suddenly screaming after whispering for minutes in a horror game. In general, you want the threshold to be somewhere around your normal talking voice, with the compression being somewhere between 2:1 (for an almost not noticeable difference in sound) and 10:1 (for a full-on radio announcer style). Note that the compressor makes things quieter (a 10:1 compression means that something with an input level of 10dB above the threshold will get compressed to just be 1dB above it in the output), so you’ll need to increase the gain afterwards. Try to aim for a level which is roughly -12dB to -15dB in your normal talking voice, after the compression.

Example of the OBS compressor

If you want equalize your audio afterwards, you can use an equalizer. This is somewhat optional as your choice of mic will already have some effect on how you sound, but if you dislike the sound of your mic or dislike the proximity effect, the equalizer can help. OBS doesn’t ship with one, but ReaPlugs does, ReaEQ and ReaFIR.

Last in your chain should always be a limiter. A limiter is your lifesaver if you screw up anything; it’ll prevent anything from clipping and sounding like an Apollo 11 broadcast. Just leave it somewhere high up, e.g. somewhere between -1dB and -0.1dB, and make sure that it’s the last thing in your chain (that is, the last entry on the filter list in OBS), and you’ll be fine.

Example of the OBS limiter

Mixing. Where the background music should be in relation to you is quite a tricky question, as volume (how much sound pressure reaches your ear) isn’t the problem here, but loudness (how your brain interprets this). Even worse, different devices will sound differently, with something sounding perfect on desktop causing mobile phone viewers to not hear your voice over the background music.

kliksphilip recommends making your background music 20 dB quieter than your normal talking voice. I’d overall agree with that, though I will add that if your background music has a lot of dynamics in it (that is, quiet parts and loud parts), you can mercilessly compress them as well, just you can do with your voice. That way, at least the background music will be more consistent in itself.

The same goes for game audio: Compressing it kinda helps, so loud sword clanks and quiet footsteps both are closer to one level, and you’ll probably also need to make it way more quiet than you think you do if you’ve got people watching on mobile.

Note that volume sliders in games often aren’t the same sliders that OBS uses – OBS is decibel based (which is a logarithmic unit), but many games (and other programs, for that matter) use a linear slider. This means to lower something in 6dB steps, you need to half the linear slider each time, ie from 100% → 50% → 25% → 12.5% → …, so you’ll need to pull down those linear sliders quite far if you do use them.And that’s it

Well, almost. There’s still a lot more that I could touch upon, but intentionally left out because I didn’t want to make it too overwhelming. I’ll leave some more links here though, in case you’re interested

  • YouTube had a workshop on good audio recently.
  • Sound comparison between the Elgato Wave 3, Blue Snowball, Audio Technica ATR2100USB, Røde USB Mini, Blue Yeti Nano, Samson C01u Pro, Blue Yeti, HyperX Quadcast, Razer Seiren X, Beyerdynamic Fox, Audio Technica AT2020, Blue Ember, Shure SM7b and Neumann U67 — aka most of the mics which currently are currently targeting gamers/streamers/podcasters, plus some very expensive pro stuff.
  • The Wikipedia page on Microphones has quite a lot more technical stuff; good starting point if you are interested in the tech behind it.
  • We have a discord server if you want to chat with us, whether it is audio, games or something else!

2 thoughts on “The Audio Guide to Happiness, or: How to make your Streams & Videos sound good”

  1. Pingback: AVerMedia NEXUS AX320 & Streamer MIC 330 - Review | Koytek Wattenberg Media

  2. Pingback: Be a YouTuber, not a NewTuber. Make great Content. | Koytek Wattenberg Media

Leave a Reply

Your email address will not be published. Required fields are marked *