ATEM Mini Sound Settings - Compressor, EQ, Limiter, Expander

in this episode with our ATM mini or a 10 mini Pro we're going to look at some of the audio settings you can use for live streaming [Music] now there are a few things we need to keep in mind here number one the first question I get all the time is what settings should I use so I'm going to give you some settings here but they're just a starting point you're going to have to tweak them to your particular situation your voice your microphone and your recording space so keep that in mind you can start with these may need to tweak them number two when you're listening to this you need to use the best quality headphones you can get your hands on if you're walking around out in some noisy place and listening on your phone you're not going to get a lot out of this and also you're going to need to ride the volume a little bit because I'm gonna give you the raw audio that's coming out as we process it and get things set up so there will be some volume fluctuations be prepared for that we are over here on the HM software control down here at the bottom I have my audio tab I'll click on that and the first thing I'm going to do is go ahead and just reset all of this so that we can start from scratch just like you would be reset and then reset now the first thing you'll want to do is that when you come in here your audio should be peaking somewhere around -12 DB so first of all however you send your audio in you want it peaking somewhere around there no hotter than -6 at the absolute most but probably somewhere closer to -12 would be ideal in the control panel here you can see our audio is coming in from cam one we haven't applied any sort of offset in terms of the input gain here so we've left that just as it is and we're peaking here our absolute highest peak was at minus 9.8 but around minus 12 generally all right so first of all to kind of tune the sound we're going to use our equalizer and this does a couple of things first of all turn on band number one this is a high-pass filter or a low-cut filter same thing different names and what we're doing here is just applying this so that we get rid of any sort of low-frequency Rumble so for example if you have an air conditioner or heater running and you need to cut some of that low frequency sound this will do that I do this as a matter of course on all audio because you don't record spoken word generally below 80 Hertz and so I just set this at its default is that 46 Hertz and I just turn it on and it has a little bend here which indicates low-cut or high pass all right next up to kind of tune the sound to my particular voice and my particular microphone and the space I'm recording in I do a little bit of EQ typically but very very mild and let me show you what I mean here I'm gonna come over to band number 3 and I'm gonna boost the gain on that and you can see here I can move this back and forth this is called sweeping so I can sweep it back and forth and what I'm going to do also is narrow that up a little bit I just want to kind of be a little bit more surgical about it now you can hear my voice is changed some notice here as I get right up in this area the erotic becomes really honky that means there's potentially a little too much of this energy so we're gonna pull it back and what I'm gonna do here is just take the game and we'll pull that down probably about 5 DB and I can widen that so that's what it sounds like now so this is gonna be very subtle here's before and if I talk enough there's sort of this harshness in the mid range of my voice around 1 kilohertz and then let's go ahead and turn it back on and now we've turned it back on and you can see it just smooths my voice out just a little bit so that's one thing you can do now your voice will probably not be at 1 kilohertz maybe it is maybe it's not hard to tell what you need to do is just sweep that frequency back and forth until you start hearing something that doesn't sound good and then pull it down and widen it up just a little bit now use your ears this is where you use your ears to make the judgment if it sounds good it is good a lot of times people kind of second-guess themselves if you've been working on it for more than 15 minutes you probably need to step away from it and come back but that's an idea of how you can find frequencies in your voice it can be a little bit harsh and pull them back just a little bit so we're going to go ahead and leave via equalizer there and now we'll move on to the dynamics alright the dynamics gives us a couple of different tools an expander a gate a compressor and a limiter now I've come into the dynamics for channel number one which is where camera number one is coming in now you'll need to do this wherever your microphone is coming in or wherever your audio is coming in so keep that in mind and then also we're going to put our limiter here on the master buss because we don't want the audio to clip at the end but we'll do that later so first let's come into dynamics we just click on this little square here for dynamics for camera 1 and that brings up this panel here so over on the left hand side we have our input level and that's how it's coming in from the camera the audio and then over here we have our output now what we're going to be aiming for is we probably want our output to be averaging somewhere between minus 10 and minus 5 but that can be a little bit dangerous if you do it over on this side because it could start to clip and things of that nature so we'll do that when we come to the compressor and the make up game for now what we're going to do is let's come into the expander actually I'm gonna go ahead and apply a little bit of makeup gain here just so you can hear me reasonably well all right so let's turn the expander on now what an expander does for us is you can actually probably hear it already an expander is going to reduce any noise that you might hear when I stop talking so when I stop talking watch what happens over here we have this little gain reduction thing this is the expander here as soon as I stop it actually pulls the noise floor down now it's not doing it while I talk or not much but as soon as I stop talking it engages so let's go ahead and set this up here now you can set the threshold this is the level when the audio falls down to this level in this case - 45 dB that's when the expander kicks in and what the expander does is it it expands the quiet things and makes them even quieter so it takes up quiet things and it basically pulls down the fader and makes them even quieter so that's gonna be your noise floor any noise you may have again once you stop talking so our threshold is when that kicks in the level at which it has to drop to before it starts to expand and you can go down as low as minus 50 I usually stick it in the minus 45 to minus 50 and again that's assuming that my levels here are peaking somewhere around minus 12 so let's just say - 45 shall we and call that good okay now the range is how much it will reduce that audio between phrases now I usually drop this to about minus 9 its default is minus 18 or 18 you can think of it as reducing by minus 9 or minus 18 or whatever you set this to decibels so the reason I set it to minus 9 is that it's less obvious that way what you don't want to do whenever you're doing any sort of audio processing for dialogue spoken word things you don't want to go really heavy-handed because it will just sound obvious and strange and so we don't want to be super obvious and strange we just don't want that noise in between phrases to be really distracting so 9 is usually a good starting point do what you need to do to get your sounding right in ratio I'm just gonna leave that at one to one point one and then for here on the attack I usually like to make that really quickly this is how quickly it kicks in actually we don't have to do super quickly we can do a little bit of a delay you can see here as I make it longer it takes it longer before it kicks in but I don't want it to bake take too long so usually I'm gonna leave that right around its default here so I think somewhere in the I think it's defaults about 1.5 milliseconds and that's just fine now I'm not going to worry about the hold the release I typically put around 100 milliseconds there we go and this is what that sounds like now so you can hear what things sound like now as I stop talking you're going to notice that we don't hear a lot of the noise floor so all right there's our expander again you're gonna have to fine-tune that to make it as sound as natural and [Music] non-distracting as possible so keep that in mind and then next we're going to come over to the compressor now the purpose of the compressor is to well that's extreme huh I'm gonna pull this up for now okay the compressor the purpose of a compressor is to take those spikes in the waveform the audio sometimes we get these very prickly kind of spikes sticking up and that's when certain sounds are made usually it doesn't really contribute a lot so that's gonna be like at the start of a phrase maybe it's gonna go really loud for just a second or really high amplitude for just a second and then come back down and we want to actually manage some of that and pull it back down so that we can loudness normalize we can bring up the entire audio level all of it together if we don't do any sort of compression first the problem is is if we just loud miss normalize if we brought all of the audio up those little spiky things which are called transients would then clip and they would distort so what we're gonna do is we're gonna use this compressor to pull those transients down first and then we can use this make up game to make our audio overall louder so how do we do that first all we have our threshold now what you want to do in the threshold this is the point at which the compressor starts kicking in so once the audio gets up to -12 DB it's going to start compressing it and so you can see here if we watch this meter right here it'll come down from the top doing gain reduction it's not engaging at all now if I go down here now it's engaging pretty much every time I say something what I want to do is I want to adjust this until it's just barely tickling that meter that is to say it just shows up every once in a while especially if I do a laugh test so I'm gonna do a laugh test here ha ha ha ha ha ha ha so you can see there it brought it down a little bit but it shouldn't be engaging all the time just when I speak a little bit louder so we found that in our case about minus 18 is gonna do that job I leave the ratio at 2 to 1 because I don't want to do I don't want this to be super obvious I want this to be pretty subtle and what that refers to is it as soon as the audio does reach the threshold it comes up to the threshold level I've set here - 18 DB any audio that goes above that or gets louder than that will be compressed at a two-to-one ratio so if it goes over - 18 DB by 2 decibels this compressor will now compress it down to just one decibel that's what two-to-one ratio is all right and then our attack we're gonna make that really fast so we're gonna put that at 0.7 and I want this to react very quickly because again I'm just trying to use this basically like a limiter I just want it to catch those transients and push them down just a little bit so a fast attack will do that job for us we're not going to apply any hold and then for our release we'll put that at about 100 milliseconds again all right so this is what we get now now notice here that our output is a little bit louder than our input because we've applied some makeup gain what I can do now is that as I'm talking I can increase this makeup gain a little bit here so that I'm coming in between minus 5 and minus 10 overall and if I do a laugh test here let's just see what happens ha ha ha ha ha ok we got a little bit above 5 we might pull down just a tiny bit somewhere in that range so that's gonna put us at a good strong healthy level for our streaming audience to listen to and I think we're in good shape now so let's go ahead and close that now the last thing we want to do is make sure that if things really get out of control and something gets really really loud unexpectedly that we don't clip so what I'm going to do is going to come over here to the master buss and click on dynamics there and over here we will apply the limiter and I'm going to take the limiter and move its threshold up to about let's go minus 3 dB somewhere around there now the reason I go to minus 3 DB is a few things that's when the limiter will engage so when the audio gets up to minus 3 DB yet will start to engage and push that audio down and I want to push it down really quickly so I'm gonna set the attack to 0.7 no hold and again the release to 100 milliseconds again may need to tweak this for your settings so let's go ahead and close that now I'm gonna do a laugh test and see where we get to on the meter here I want to make sure that I can't clip so watch your volume on your headphones it's gonna get kind of loud here here we go ha ha ha ha ha ha ha all right you can see it stopped us at amount - 2.25 DB so it's definitely doing its job and that means we are not going to clip now why did we leave that much Headroom we've left -2 DB of headroom a little bit more well the reason we did that is that once you are live-streaming you are actually encoding the audio to a lossy format and those lossy formats actually eat up some of that Headroom they'll use that up so if you don't leave yourself 1.5 to 2 DB of headroom you can still get clipping even though you're not technically clipping here on your side before the encoding takes place so that's why we do it that way and that is an overview of the settings that I typically use when I'm using something like the a10 mini or the a10 mini Pro to do live-streaming and kind of process up my audio a little bit make sure it's loud enough make sure it's pretty clean in between phrases and that it sounds just a little bit better with that EQ so now let's go back to the 8m mini and the a10 mini Pro and the sync issues I have a request for Blackmagic Design so when the a-10 mini first came out there was actually a seven frame offset hard-coded into the unit and that was too much I guess you guys realized that especially when you came out with the a-10 mini Pro and the ability to control your Blackmagic pocket Cinema cameras from your a 10-minute or your a-10 mini Pro and you change the offset so that there was only a one frame offset from the time the audio came in to one of the audio inputs until it was joined up with a video coming from the camera now that works great for the pocket cinema cameras it doesn't work great for a lot of other cameras because a lot of them have more latency on their HDMI output so what I would like to request from black magic is a configurable setting for the audio offset so if we could configure between zero and ten frames of offset for the audio coming in to one of the inputs on the a-10 mini Pro that would be amazing I know it's possible at least to do seven because of course you're going to do some sort of buffering in there but it's possible to do seven frames and if you could implement that as a configurable value that would be a fantastic addition and make the ATM mini on the a-10 mini Pro even more valuable they're great devices love them that's a one thing I'd love to see as well so I don't necessarily have to feed audio directly to camera to keep everything in sync so hope that was helpful if you have any questions go ahead and leave those down below if you've not already subscribed make sure you do that and we'll be sure to get you more great videos on how to improve your lighting and sound or video talk to you soon [Music]

Channel: Curtis Judd

Views: 76,198

Keywords: ATEM Mini, ATEM Mini Pro, Sound, Audio, Settings, Compressor, EQ, Limiter, Expander, howto, how to, how-to, tutorial, setup

Id: j4RLCBd3Q64

Length: 16min 18sec (978 seconds)

Published: Sun May 03 2020