OK, I have a Shure V83F shotgun mike on top of our Sony A7R Mark III in our studio setup. With a 50mm F/1.4 lens at about 2.5 meters, it’s the perfect setup for two people. With a 25mm F/2, it works great with four people.
One of the hard things though is setting up the audio. It took a long time to figure it out and the acoustics in our open room are not great, so YMMV, but this is the setup that works for us:
- Shure V83F Gain. I set this to 5 and it seems to work pretty well. The main thing to make sure of is that when you are talking loudly, you are not clipping (the OBS volume bar goes all red, so it is over 0dB). I had it set to 35 originally and it was terrible but at. You want to make sure that at normal volume, the volume is at the end of the yellow (not the red). As an aside, you can hear what you are recording by going into the audio mixer and turning “monitor and output” on the gears icon so you can listen with your headphones.
- MacOS levels for microphone inputs are set at 100%. I also use SoundSource from Rogue Amoeba and make sure this is turned to 100%
- Open Broadcast Studio has a series of plugins and here are the settings that worked for me: 0db Gain. This is what you want, you don’t want to have digital gain, so make sure you are maximizing the input from the microphone itself
- Noise suppression. You want this right after the gain, so you get out the noise as early as possible. The RNNoise works very well with no real lag or performance hit at least on my M1 Max 🙂
- 3 Band Equalizer. OK, this is the one that changes the most. That’s because since this is a shotgun mike, you really lost quite a bit of base and don’t get the Radio sound (that is lots of bass), so I set this for bass at +10 dB, mids at -5 dB because I have a not very baritone voice and then +3 for highs to give it a little pop.
- The Expander is next, first, you want to set it for a 3:1 ratio (so not too much expansion), then attack at 1 ms so it picks up very quickly, and set the release at 100 ms. Now move the threshold all the way to the left at -60dB. Now talk as quietly as you can and turn up the Gain until that quietest sound is in the yellow. For this far +5 dB because we were pretty far from the microphone, so the quiet sounds really dropped off. There is less of this effect in the near mike (see below). Now you turn the threshold up and what happens is you make the lowest possible sound and then create a low sound like a gentle click, you want to not hear the click and you do want to hear your voice. Beware that if you set it too high, then you will have a weird loss of the first syllable of each word. That’s because the thing doesn’t turn on until after you start talking, so be conservative to prevent this kind of effect. I set it to -52 dB mainly because the noise suppression does such a good job. Remember the very low sounds are cut out from the noise suppressor, so this is for sounds that the suppressor is missing (that is it is loud and sounds like speech, but isn’t you). If you set it to high, you get this strange breathing effect, so I set it as low as I can stand.
- Compressor. Now you do the opposite at the other end, that is with the combination of these two filters, you are pushing all the sounds into the middle volume. That’s exactly the way a “radio” voice works. So again, you set the Ratio to 3:1, attack at 1 ms, and, release at 100 ms and now you shout and turn down the threshold to the RIGHT this time. That is, it never sets. Now shout really loud and keep turning the threshold down until it doesn’t go into all read no matter how loud you shout, you never go beyond -1dB. Note that you always set this at 0dB gain since you are trying to make things quieter. In this case, I set it for -8dB which seemed to do a good job with the shouts.
- Limiter. This is the last stage just to make sure you don’t overload the system. Set the threshold to -0.1db and then you will never overload the system.
Settings for the Rode NT1A and FocusRite
To show you how different these settings are, I used the same scheme for Rode NT1A which is close-miked in a much quieter room. Here are the settings (also so I don’t forget if I blow away my OBS Studio installation):
- Gain: 0dB as before. I’m using a nice FocusRite so I just set the levels over there. I found that on this setup, setting the FocusRite to 4 o’clock was about right.
- Noise Suppression: RNNoise as before. I have to say this noise suppressor is amazing
- Three band equalizer: 8dB Bass, -3dB Mids and 2dB Highs. So less bass as this is close-miked than the setup above. Just make sure your mouth is close to the microphone and you have the pop filter to make it work.
- Expander. 1ms Attack and 100ms release as before, Now again, turn the threshold all the way to the left and then turn the output gain up until the quietest voice you want to capture should be in the middle of the yellow. For this setup, the output gain is 6dB and the Threshold is -60 dB. Since I’m much closer to the microphone, just a light tap means a very low threshold as it’s a quiet room so even a light tap didn’t get through which is pretty cool. Things like keyboard typing are already removed by the suppressor. These settings are surprisingly close to the far miked Shure which is interesting
- Compressor. 1ms attack, 100 ms release and, 0dB gain as before, but the threshold at -5dB when the expansion starts. You don’t want the loudest sound to go above -1dB. This is less than the previous probably because I’m already so close to the microphone that there is less range anyway in the sound input.
- Limiter. This is the default -0.1dB and 60 ms release as before.
For a great video that explains this, check out the Video Nerd for a good explanation in detail
