Transcription
Something that strikes me about the best engineers in the world is their humble curiosity. Their willingness to ask questions and not know the answers, but seek to discover them. And that was really eye-opening for me.
It's been an eye-opening experience for me because these people I've looked up to ask questions. They're meant to know everything. No, they know a lot and they're talented at what they do, but they never let that define them. It's their curiosity and their passion for their craft which defines them.
And last year, and even this year, I haven't documented it as much, but last year I aimed to document every single time I made an update to my process in mastering, in mixing, whatever it might be. And that's what this video is. All my mastering updates for last year, all the little 1% improvements, the aha moments, the things I fixed, the things I changed—the complete edition, that is this video.
Before we dive in, if you're somebody who's passionate about mastering your craft as an engineer, 5-day mastering sprint. It's free. Links in the description down below. Now, let's dive in. It's May 13, 2024. I'm going to be updating you every week with whenever I change something little in my mixing and mastering.
But this week was crazy because I thought I didn't think that's the problem when it came to clippers. And I was stubborn and I didn't want to believe something I saw a friend of mine, Drew Vespers, put out. Um, and I was wrong about that. But also, I'm going to show you something I'm doing with compressors, which is just fun. It's it's a little tweak to my TDR Katonov workflow. So, let's jump into it.
So, with clippers, what happened last week? Drew Vespers put out a video on oversampling with clippers. And I didn't want to believe it, and I couldn't believe it, cuz it wasn't happening to me. And what was happening to him was when he oversampled his hard clippers, um, or when there's a ceiling, the oversampling algorithm would increase the peak value when it dropped down in samples. Now, it wasn't happening all the time for me, but now I've figured out it actually does happen sometimes, and I can explain why.
So, what I've got here is I've got a mix here, um, by surname Ampersands. Uh, that's the name of the track. And if I go to the very highest peak. Okay. And this is where things start to get interesting. Okay. I got to zoom all the way in. Z. Okay. If we zoom zoom zoom zoom. Perfect. We're there. All right. Now, there are only two samples. So that peak level there, if I go to waveform stats is at 0.1. Okay, it's at 0.1 of a decibel. But those two samples about up to five. Actually, if we go up to this top one here, those two samples are 6 decibels difference, but but before it hits that next peak just on the right-hand side there of the waveform. So, if I set my clip at 0.5, it should just clip there.
So, what we're going to do is we're going to pretend we're oversampling. So, and and I'll show you why this is just so crazy. So, we're going to resample this up. So, we're going to do four times oversampling. It's going to sample up, which is great. No fidelity, no alias in distortion. But this is where things get funny. And I'll Where are we? Where are we? Uh, the clipper. I need the plugin. Let's get this up. We're going to clip it with no oversampling. So it's just one times oversampling. And we'll render. Okay. So, so far you're going to see the peak level be at what's it going to be at? -6. That's good. That's great. But then inside here when you're oversampling and then re-downsampling when you go back to 48, that peak level that should be -6. Take a look at what happens to it. Okay. Because it has to introduce this low-pass filter. It actually flicks it out. And then what's our new peak level? -5. So this oversampling, by adding oversampling here. Okay. I just emulated that oversampling by putting the resampling algorithms between the processes. You actually increase the peak value of hard clipping.
But this I didn't believe. And the reason why I didn't believe it is because it wasn't happening in my masters. And there's a reason why it wasn't happening in my masters. It's because the masters I was working on at the time, the top end wasn't super harsh. This only happens with very intermittent peaks that are very high frequency because this filter here. Okay, this oversampling and downsampling filter uses a low pass, and the low pass only rotates the phase of very high frequencies. All right, so now let's take a look at it. No oversampling and we can look at the peak value over here. That's the sample peak value. [Music] Now let's have a look at what happens when I enable oversampling. All right, now we're starting to get some peaks over six. 5.5, that's half a decibel louder. Okay, what if we put, and this is where we can have a bit of fun, we put a low-pass filter before the clipper. What happens to those peaks when they're oversampled? Perfect. It disappears.
So what this actually shows is that oversampling is introducing new samples in the high-end when the peaks have a lot of high-end information. Now what does that mean I've been doing in my sessions? So for the most part, I've kept oversampling. Okay, for the most part, I've kept oversampling because there are very specific circumstances when I have a lot of peak information in the top end, like high-frequency peak information, but for the most part, the mix is coming in pretty controlled there. And when I'm oversampling, it's not producing new sample peaks louder than the ceiling I'm setting on the clipper. That said, when I do have very sporadic high-frequency peaks coming through and it's causing my oversampling algorithm to create new sample peaks above the clipped level, that's when I choose to either go to one times oversampling or I'll remain at four times oversampling and try introducing high uh low-pass filters in the top end if they can be gentle enough to mitigate the amount of high-frequency content causing those overs. Or I just stick with one times oversampling.
So, that's one thing I'm doing here. That's an updated sort of revision to my clipping technique. It's a very subtle thing. This is like 0.005% of a change to my workflow. But, nonetheless, it's still a change that I want to share with you because as I learn information, I want to be helping you guys out, sharing information with you. And you'll even find more information if you're part of my newsletter. You can subscribe to that down in the link below. It's the best newsletter for working audio pros.
Now, I've sort of given my little spill there. Now, compressors. What am I doing different? This is TDR Kotelnikov. Funny story before I show you what I'm doing different here. They had the gentleman's edition on sale. I bought it like a month ago or two months ago whenever it was on sale. I haven't even installed it because I'm still using the free version. It's that good. So, if you don't have TDR Kotelnikov, make sure you download it. Um, I even got my presets here. You can get them if you're part of the newsletter. But what I'm focusing on here is this peak crest because there's two detection passes, a peak and an RMS detection circuit. And you can offset the peak threshold. So if you bring this positive more positive, it lifts the threshold. So it only captures the very top peaks. Otherwise, you bring it down and it becomes more aggressive on the peak detection circuit. Um, the way it compresses and you can got little indicators here that that dictate that. The reason why I've been fiddling with this more is we'll go into the delta signal. This song. This is more for demonstrative purposes, but with electronic tracks, I've been pulling the peak down. So, it's it's pumping a little bit more with the transients and you get a cool movement. I'll set up an A and a B and then level match it all. Because what when you hear it, you'll be like, "Oh, wow. This is actually really cool what you can do." Otherwise, what I'm doing is I'm pulling this peak crest up. Well, usually it's around there and like native 2. And I'm allowing the vocals or the main RMS to gel the track almost like SSL style compression where it just glues everything as it moves through the macro dynamics of the master. But we'll dial in two versions, a peak pumping and an RMS moving one level match. And you'll get to hear and feel how that changes the sound. I listen to the delta when I when I move this. So I'll listen to the delta and move this. And then I'll also monitor those release um uh indicators there. It's a really long trip through the up and down and me feel like a no. When life is a bull and I can't take that duality, it's a mystery. Is it really or is it only through the up and down when life is a duality? Okay, beautiful.
So, what I basically set up here is this A version. It's pretty much just peak compressing and you'll hear it in the bass that pumps a little bit. It's not really for this style, but it's just a nice tool to have when you're working on electronic music and you want to have a little bit of movement, especially when you're driving something loud. You don't want it to feel like it's just brick wall. You want the compression to sort of move with the music a bit. And then the B version is probably what I'd end up choosing on a final master. Um, just because it allows the drums to open up a bit more and the vocals to sort of drive the jelling through this passage. So, take a listen to the A and the B. I've been on a long trip through the up and down and feel like a life. And when I switch it over to the B version, it's going to feel a lot more open. [Music] So yeah, that's just something I've been playing around with the TDR Kotelnikov, that peak and crest factor.
So anyway, I hope you really enjoyed that, and I'll be back more with next week. Anything I've changed up at the studio, whether it's mixing or mastering, I'll share it with you. And until next time, take care.
The mid-range is so stupidly powerful. Not for the mid-range itself, but what it does to everything around it. I want you to listen to this before and after because something happened in another session and I'm like I got to demonstrate this to you guys because it's nobody talks about it cuz it's not exciting, but it freaking makes a huge difference just to listen to this before and after the EQ I used on this master. This track is by Musafili and um Saint the Gambian Dream. It is a cool tune. It's recently out and it's popping. All right, take a listen to the before. This is OG and the bottom is the master EQ. [Music] Now, yes, they're actually level matched in LUFS. The mid-range is where this is one. Now, backstory. I did a master for another clone. I'm going to show you that session down the track, but I had to do a huge scoop in the low mids. More than most people would expect a master was like 3 decibels. But what it did to the master, it made the low end feel big and round and the top end feel open. The it wasn't a muddy mix per se. It was just all perspective. I'm going to show you the EQ I did on here and why I chose that EQ. And then we're going to go over a scooped and a peak version level matched. But namely, what I want people to understand first is that your mid-range, okay, your boosts and your cuts are simply a perspective shift from your highs and your lows. Okay? So, you want less lows and highs, boost. You want more lows and highs, cut. Again, we know this smiley EQ, but there's there's a method to the madness. I got to show you the method to the madness. That's where it's going to all come in. So, obviously, you see here, level matched, a bell, and two shells. Okay, cool. All right, I got that out of the way. I just wanted to show you that for anybody that didn't understand what did I choose to do here. Now, this EQ here is a boost to the mid-range and a cut to the lows. So, let's monitor what's going on with this OG signal. All right. All right. Why are we doing this? And we'll turn this EQ off here. And we'll just have a look and listen. Okay. So, first things first, that low end, it goes down low and it's it's definitely too big. In my opinion, the low end's probably just a tad on the hair too big. You try and push that through a limiter, that sub is going to be the first thing that just starts to make that limiter pump a bit too much. It doesn't need a low, a high pass, sorry, because it doesn't go down that low. So, what I've opted to do here, okay, is add a shell. Okay, 100 hertz, one and a half dB shell. The real juice though, the top end's snappy already. It's already snappy. I actually end up taking half a dB off. The top end snappy. The bottom end's too low. Okay, it's too much. I need to drive the mid. I need to get some energy out. So, that 600 hertzish, all right, is right in the heart of it. So, as soon as I turn it on, take a listen. [Music] and that's the magic of it. That is why the mid-range is so strong and powerful. Okay, that mid-range affects your perception of the low end. It affects your perception of the top end. So, I thought for the sake of demonstrating this to you, it's important we run some tests here. Okay, we've got the OG one at the top. We got a master EQ down the in second. Then we've got a scooped one. And these are all level matched. I decided to scoop scoop a whole bunch of mid-range. Okay, the same mid-range we're boosting. Okay, and see what that does. It makes it quite dark because there's so much low end coming out and there's not much presence and the high end sort of just feels sizzly. It doesn't feel snappy. Okay, then we've got the peaks where I do the opposite. I just crank the top, but I don't touch the lows. And that feels congested because the lows are already quite big. And you're just adding more mid-range to compete with those lows. And it feels squashed, not open, and nice. So, what we've got is let's actually make all these a bit smaller. How do we do this? God, man, I'm bad when it comes to um interfaces. Let's scale this down to 50%. Beautiful. You go there, buddy. Should we just do a crossways? Okay, good job. Take you to there and then we'll get you, buddy. All right. And you're already at 50%. That is beautiful. Now we can go through each one and solo them. So we'll go OG, master EQ, scooped, and peaked. You ready? Let's roll this. [Music] [Music] You could you could hear it straight away. The scooped one, it's like it completely disappeared. The peaked one was like it was really congested. The master EQ, this one here, okay, with intention, you can use the mid-range, there's a little boost and a little duck in the lows. You can use the mid-range to create power in your masters. It was fun. I've got another master which I've done which as soon as it comes out I'll get permission from the producer and the label. I want to show you it because it uses this concept but at a freaking extreme which is wild because the before and after is like did you remix it? Not really. I I love the mixing engine. I love the producer. Super beautiful people, but that drive that enhancement is just ballistic. So, if you want to be around for that, make sure you subscribe to the channel, but more importantly, the newsletter. The newsletter is where I geek out on all this stuff. All right, so make sure you subscribe to the newsletter. Link in the description. And until next time, take care.
So, an email subscriber pinged me an email. They pinged me this email, and I saw it as a challenge. And it's a challenge that by the time I got to the end of it, I felt like an idiot. Now, I'm going to go you through my process of elimination, trying to understand this person's problem, um, or issue they were having with their limiters. And I'm not going to tell you what the solution is till the very end. But by doing that, I think some of you will be able to pick out the problem either straight away or somewhere halfway along my tests because now knowing what actually is causing this, I feel like a bit of an idiot. But it has taught me some things about a particular parameter in limiters which is very important to the tenality of them which I hadn't considered or I knew about but hadn't been brought to the forefront. So follow along here. Now, I'm going to exploit the person's name because I I don't I don't know if they want that information disclosed in a YouTube video, but basically they said situation. They got a simple square wave with unison generated by Serum to get some stereo information. So, they must have like done some synth wave modulation. I've got the file here that we can listen to. And they've monoed the low end using Tone Projects Baseline. And they've given me a graph here showing that everything is mono. But what's happened here is when they put FabFilter Pro L2 on the end of the chain, it and it gain reduces let's say 1 dB, 2 dB, there's stereo information being reintroduced back into the low frequencies and they said the same situation happened with Smart:Limit on Sonnox, the same situation with Music Hat Master Plan and Ozone 11 maximizer, so to solve this I use a limiter then Baseline to make the low low frequencies mono and then another limiter which does not introduce this issue when using like one decibel of gain reduction. Now this perplexes me. It's like how can a limiter introduce or skew the mono low end, um, period. If if you've made a mono then you're limiting it. It shouldn't be doing that. However, you're going to be listening, and some of you might have already picked up what the error is, but I'll take you through the test.
So, here's the square wave he sent me. I've just cut a section off where this issue was more prominent and persisting. So, let's have a listen to this. And if we go here, um, we can see there's definitely stereo information down there. If we monitor the side signal, um, that's on side. There's still low-end information there. So that's just the fundamental synthesis that is created for this test. Um, and what I've done is then I've added ProQ3 with a brick wall filter on the side signal because I didn't want to, you know, like have anything coming down into there that could affect the limiter at all, period. I wanted to be very clinical. Not that I would ever do this in a master, but um, let's have a look at the side signal now from that. And as you can see, like negative 115 decibels down there. It's pretty much non-existent, those low frequencies. Now, here's where things got interesting, and that's when I bring in this limiter. Um, we'll even just go to the factory default setting. Um, because I did do a bit of tweaking, but I found out that none of the parameters actually made this do this specifically. So, let's do this. And we're just going to do one and a half dB of limiting. What's going on here? See there? I'm getting low frequencies coming in. Why is that the case? So, I had to do some digging here. So, I rendered out the raw, the filter, the filtered, and the limited files. I've got them over here in audio editor. So, this is the raw one. Uh, 2.5 dB is the ceiling, which I'm using on the limiter. Then I got the filtered version without the limiter. And that's, you know, pretty symmetrical. And then I got the the limited version. Now, some of you are going to pick up what is actually causing that low end to shift into the side signal again after the limiter. If you haven't, you will soon. But fundamentally, what is happening here, and we can visually see it, it isn't a bad thing, but it's something we need to be conscious about because there's some lessons to take from it, which we'll learn in a second, is that let's go back to our limiter settings here. Let's go here. We're at 2.5 channel linking. Okay. Okay. This is where the issue was. Cuz I think I had the channel linking down here or so. Channel linking and attack. Basically what is happening is one channel is limiting and the other one is limiting less or not at all. And the reason is because of this channel linking and the attack time that is causing so signal our stereo field is the difference between the left and right channel. So obviously if the right channel limits more it's going to drift slightly one way or slightly the other and that difference signal even in the low end is going to be represented in that drifting. It's something we know. It's something we understand as part of using limiters. So if we just speed up the attack for the sake of demonstrative purposes, not that this is a solution, we're just looking at it and we get this all linked up. Let's have a look now at what happens to the low end here. Let's let's let's play this through. Okay, let's just make sure the transients are linked and we're good. There's not much low end as much low end sort of rearing its head there, which is good. Now, if we choose to bring this release back, we're going to get a lot more low-end coming in because the release stage is independent on the right channel as opposed to the left. But if we bring this up, wait a second. We're pretty good. Lessons, cuz I said there's something to take from this. So, we understand why it's happening. It. In a perfect world, of course, you don't want the the signal to drift. That's why we have that tool there of channel linking, so we can decide whether we want the stereo field linked with the limiter or not. Now, that's perfect world, but there's practical world in terms of the application and why this is important. If things are going out to vinyl and you don't want your needle in the subs to drift, you probably want to be pretty critical of how your limiter and all your processor processes are actually affecting the balance between the left and right channels. So, this is just something I'm learning here now through these tests. Even though I'm aware of it, I wasn't aware of how far these things could drift depending even at a small amount of limiting to 2 and 1/2 decibels. It's only doing 1 and 1/2 dB of gain reduction here. Um, it's just an interesting assessment and I want you guys to let me know in the comments, did you find this earlier on? Um, did you like I want to know what your thoughts are on this because it's just something I wanted to flesh out and come to understanding. So attack time, I'm not sure why the attack time so dependent on this, but I think it's just because um the velocity or speed in which the channels independently might be moving even when they're linked could affect it possibly. But it's mainly the channel linking that is causing the stereo field to shift or move or sway or drift um the low end or all the frequencies really, but it's just drifting that stereo image. I want to know your thoughts. This was a fun thing to come in. So, thank you to the subscriber who pulled it in. Um this is the sort of geeky stuff we do in the newsletter. So, if you're not a part of it, you need to be a part of it because these people, all professionals like you and me, we love the detail. We love pushing our craft. And what I thought would be such a complicated, convoluted issue with the limiter was actually something as simple as knowledge I already had in terms of channel linking. Duh. That's what would cause it. But it's only until you look into it that you're like, "Oh, actually, that's much simpler than I actually thought. I thought it was going to be some sort of weird DSP thing happening in the background doing something funky with the phase, but it wasn't." So, there you have it. Until next time, take care.
So, you've made it this far through the video, which means you love mastering. Perfect. 5-day mastering sprint in the link in the description below. Now, let's get back to it. I just filmed one of the wildest tests for my master class I'm putting out, and it was so wild that I
Had I have to share it with you. So, here it is. And if you're interested in advanced mastering details, the mastering class I'm putting on, there's a newsletter link down below, so you don't miss out when it gets released. I hope you enjoy this and until next time, big love.
All right, midside EQ. All right, in mastering you must use linear phase midside EQ, nothing else. Period. Okay, now this was brought to my attention by Ian Stewart when he was taking me through basselane pro, and it all made sense because I'm going to show you some things which are gnarly. Like gnarly.
Now midside, what is it? It is basically an encoding and decoding process of the left and right channels to compartmentalize the signal from a left and right, and then it gets encoded into midside by adding the similarities. So the left channel plus the right channel's phase creates the mid signal, and the left channel minus the right channel—so the differences between those two—create the side signal. That's the encoding of midside, and then it inverses back on itself in order to create the left-right output.
So what I've got here is the left-right output of this mix. It goes through a midside matrix which outputs mid and side. And I've got the mid and side in two separate mono channels which then goes back into a stereo bus which takes the midside input and makes it left-right. And to show you that this is all hunky-dory and nulls with the original stereo, let's have a listen to the stereo, and then I'll bring in this matricy, and you'll see with this null flip on the stereo that everything nulls. Catch me with no [Applause] perfectly nulls. [Applause] Okay, good. That is a fundamental.
Now, let's just dive in a little bit here. What creates and encodes and decodes the midside matrices? Phase relationships. Okay, that is why using linear phase with midside EQ is so critical. And I'm going to show you exactly why. What I've got here is an all-pass filter. Okay, let's put this at 1 kHz. Now, an all-pass filter passes all frequencies with equal gain. All it does is rotate the phase. And I've rotated the phase 180° at a center frequency of 1 kHz. And I've turned that on. Let's listen and see if this nulls. This is on the side channel of the midside matricy. Let's see if it nulls. Cuz technically the output of this is identical in frequency and tonal balance. All that is changing is the phase polarity at 1 kHz. Let's see. [Music] Okay.
So, why isn't it phasing? Why isn't it why isn't it nulling 100%? Well, obviously the phase relationships are off. And I'm going to really test—I'm going to make it really hard—cuz I'm going to leave the the the phase rotation at that frequency. And then I'm going to make it really easy in the next test to see if you can pick up what the differences are between the left-right and then the left-right midside matrix with this phase rotation. See if you can pick up something in the stereo imaging. I really want you to, you know, push your ears here. [Applause] Girls cuz love who doesn't turn their back away. Yeah. Finally did me a favor and walked right out that door. And I have ever left. Okay. Maybe you picked it out. Maybe you didn't. It did something a bit weird to the stereo field, and it didn't feel quite right.
So, what I'm going to do is I'm going to make it really easy, really easy um to pick this out. So remember that when we did the notching and the resonance um surgical EQ in the previous [Laughter] [Music] [Applause] chapter—get that 41 41 uh or 4.1 hertz kHz or 4100 htz—we're going to rotate the phase of that frequency. All right. Now, take a listen. I'm going to have it in bypass. And that ringing frequency is going to be on the left channel, which it is on the stereo mix. And then when I put this in, it flips the stereo imaging at that phase rotation point. Exactly that point. Take a listen to this. This is wild. The ones we want girls cuz this came back. Is that not wild? Let me just let me just loop that ringing so we can really get it. [Applause] It flips it. So because we're decoding back to left, right, and you've changed that phase relationship, it actually shifts where that element is in the stereo field. Now, albeit when you're doing EQs, you're not doing 180 degrees of phase rotation um with a bowel notch—doing one or two decibels—but nonetheless, it will move the stereo image if you're just working—if you're working without linear phase EQ.
So, when we get our EQ up, okay, and we've got this side channel, okay, and we're working on this side channel or we're working on the mid channel, please make sure it is in linear phase, okay? We want to keep our stereo image strong. Keep it in linear phase. As simple as that.
Today, I've got this mix here. All right, it has been limited. And when the client sent it to me, they were so happy about the mix. They said, "I did a whole bunch of mix processing. I love the mix. Can you work with it?" And I said, "Yes, I'm going to show you my playbook so I make sure I can make sure that next time you get a limited mix, you don't throw in the towel. You can actually engage in the master and walk out going, 'Man, this sounds banging.'"
Now, before we dive in, do me a favor. Subscribe to the best newsletter for working audio pros. There's a link in the description. Over 500 people have given it fivestar reviews. There's a 60% plus open rate for each email because each email delivers real value. People are excited to to learn and and see what I've got to say on these topics because I'm really trying to help the audio community. Now, we're sending out a technical breakdown of these tactics I'm going to share in this video. So, make sure you're subscribed, okay? Panoramaing.com/newsletter. Now, let's dive in. It's limited. It's right up at the ceiling. And the overall integrated loudness is -10.3 LUS. I've got one critique for myself. My critique to me is that I should have recognized that this wasn't a 32-bit file. So, I might have been selling myself a bit short to not let the client know, hey, can you shoot me over a 32-bit float or 24 8-bit float? Because technically if there was any information above there that hadn't been getting hard-clipped already by their mix bus processing, it was already getting done, and I might have bought myself a little bit more detail, a bit more flexibility to dial in clippers. That said, the mix sounded great. Okay, they they signed off on it. So, let's take a listen to this mix cuz it's it's fire. [Music] Okay. Okay. [Music] Do you hear that? It moves. It smacks the low-end sustain. Those—It's beautiful.
Now, the first thing I noticed when I heard this is think of the beat. Like you want it to feel like somebody is stomping the out of the ground. Okayoom. And I didn't want to EQ this. I didn't want to throw EQs and like warp the balance too much. I wanted to bring some warmth to it though in a way. Now what I did was I'll show you over in plug-in doctor. Then we'll hear it so you can see it. Then you can hear it in terms of the technical side of what's actually going on so you know what to listen for. So, we'll bring up Michelangelo EQ because what um Tone Projects has done on this is just beautiful. The tone it brings is just magnificent. These are presets I made for people in the newsletter because, you know, they're the coolest cats out there. So, basically, we've got this preset here. This is the Michelangelo preset I made. Uh, and it's thumping. So, what I'm trying to do is this is a slight tilt. Okay. Slightly rolls off the top. About a 1 and a half dB boost to the sub. So that's uh 60 Hz, 1.7 dB above -15. Okay, so little bump to the subs, but then what it does on the very lowest. Okay, so we see here we got this expander here on the subs. Okay, on the transient of the subs, what's it doing? It's adding third-order harmonics. So every time that kick hits, it expands and it drives in to some saturation. So it's thickening out on the fundamental. That purple line is the fundamental. So the low-end subs get thickened out just by a natural sort of boost there because of the way this is designed. And then in the transient side of things when that range pushes—you see that blue light blue light there—we're getting that thickness—that thickness in the subs—that third-order harmonic up there which is almost equal to the fundamental and then the fifth order down below. Now how does this sound? That's what's important. So let's get rid of that. We'll open up Michelangelo EQ, and let's actually take a listen to what this is doing. So we'll put it on and off between each pass of the bass notes. Okay. So we'll start with it off. [Music] Okay. [Music] Okay. Just thickens out that low end. Remember I said at the start I'm listening to the mix, and I'm going—this has to be bombastic—and that's exactly what it brings to the low end. It just pulls that thickness in the mids—as in in the in the center of the channel—and brings it right up to your face which is beautiful. And this is about enhancing a limited mix. So we're listening to the good qualities of it. If this was crunching and sounding horrible and thumping, I'd probably give some advice to the to the client and go, "Hey, we need to readress a few things." But this is because they've kept their mix bus processing on. I'm going to remember this is what they've signed off, and we have to work with it. We have to embrace it. I love that what that does to it. But then I'm like, I've got that low end fleshing out nice in the lows. What about the top-end transients? What about that transient enhancement?
This is something I do pretty much whenever something's a a little bit pudgy or glued or I want to lift out the transients, especially in the mid-range, the upper mids, and the highs—is this here is dine one. Okay. And I'll actually show you exactly how I dial this in. So I'll just put this on the original preset which is here. Okay. And I start off at -20. I start off at -20 cuz it's quite a quite an obvious sort of sound. Then what I'm going to do is I'm going to bring this threshold which is all linked. Okay, that's all linked between the bands. I'm going to bring that threshold to the point where the low end is hitting about -12 dB of overall gain reduction. Then I'll explain to you why I do that. [Music] Okay. [Music] Okay. Okay. So, the reason why I dial it in like this and then I'm going to tweak the overall amount coming back in is because I want the low end to be tight. Okay. So, this is parallel compression. I'm blending in. -20 dibels of this of this compressed signal is getting blended back in with the original one. I want the low end to be tight. I don't want that to be popping. I don't want it to be popping. I want the top end to be popping. That's what I want to be popping cuz I want the transient and the snares and all the little details to come out. I want to compress it. But what's going to happen is when things get too loud in the top end, it's going to compress it and stop it popping too hard. If something isn't loud enough, it leaves it open. It doesn't compress it. It just boosts that signal. So, it's it's it's it's like variably changing based on the amount of amplitude in each of those bands. So those higher bands when there's a lot going on, it's going to compress and it goes, we don't want to put too much of this signal back in cuz it's going to get really nasty in the ears. But if there's not enough there, it's not going to compress at all. It's going to let it lift up. So you'll see that. You'll see that across these. And I'm just going to dial this back and then we'll AB it. Okay. [Music] Okay. Okay. Okay. So now this is where it's going to get beautiful. We're going to listen to these snares. All right. Before and after. Actually, you know what? We'll do the same thing as before. We'll loop twice. Okay. So, let's go. [Music] Okay. Okay. [Music] It just adds that little bit of pop to the transients. Pulls them slightly up, which is what I wanted. Okay, again, limited mix. You think there's not a lot you can move around, but I'm making small changes to enhance what's there.
The next thing is you got to remember with all the tonal and dynamic considerations we've had, everything is being pulled forward through the middle. They've been stereo processes, but a lot of that energy in the mix is in the middle. The low end in the middle, the kick drums in the middle, the transient of the snares, some of the sound design and and the body of it out the side, but the snap of it, the transient in the middle. So, what's happened so far? Sounds punchy. It sounds bigger, but it's just kept bringing the middle forward. The sides have stayed back. Now, I don't want to do a midside EQ where I have to start boosting things out of the side and cutting things in the mid because remember this mix is signed off on. It's limited. The client likes the way the tone is sitting. I have to think creatively. So, what I've done is this is my side enhancer I built in Meta Plugin—I'm going to add harmonics to the side. I'm not going to EQ and boost them. I'm just going to add harmonic information from the fundamentals there and add it to the side signal to help bring it around. It's really noticeable in the snares, but in a positive way. Basically, what this is, it's a midside matrix. So, the mid signal goes through to the audio output. And then I got a four-channel mixer. This first one is the fundamental side signal. So, that's just passing through passively. The second one is parallel compression. I'm not using that on this one. The third one is odd-order harmonic compression, which I am using. And the fourth one is filtered harmonic compression to the top end um emulating an AEX vintage excider. And the reason why I got these is cuz I want to add those harmonics out to the side, but not this fundamental here. I'm keeping completely at unity. I'm not changing the overall balance of the stereo field. I just want to add harmonics to it. So, let's take a listen to this before and after because this is just embellishing what's there. We're going to remember what is there, what works, how can we move it forward, how do we move the needle forward. Remember, the dynamics are limited. So, it's not like we can start pushing and pulling EQs cuz then things are going to start to crumble. That's actually what happens with limited mixes. If you try to overdo it in terms of trying to do some big cuts with an EQ or a bell and a boost and a shelf here, it starts to shift the waveform, and then you got all this headroom that you have to—that that it chews up into because you're adding all these—so you you're shifting the phase, you're rotating the frequencies, and then you're adding extra peaks. You're also taking away some. But for the most part, if you do aggressive filtering, it's going to shift that very far from where it originally was, and then you're stuck with trying to make it loud with a filtered signal, and you compromise the audio quite a lot. So, that's just a generalization. That is not as hard and fast rule, by the way. Take a listen to this before and after. [Music] Okay? Okay. [Music] Now, where we have to listen to other snares because that's what I said. Now, when you're listening to these snares, I do not want you to listen to the snap of the snare, but the body and tail of it. The—listen to it out the sides, okay? You're going to you're going to hear it straight away. [Music] It's very subtle, but it just adds a little bit of that depth in the body of the snare around the side.
Now, the final frontier is loudness. I did use the God particle in here cuz I like what it did to the sound, but that's not really a nor here nor there thing because it just isn't—it wasn't a major consideration—to embellishing what was there was sort of just another step in the process which nor here nor there. Like I said, the main things were the Michelangelo EQ, the dine one, and the stereo wideness. But now we got to get on to loudness. So, as housekeeping, what did I say at the start? You know, unfortunately, mastering engineers are sooks when they get a loud master because they crow about headroom. I don't know why some mastering engineers do because look what you can do. If a mix is up at the ceiling, you can use clip gain and bring it down -6 dibels. And now I've got all the headroom I need. And now I've driven into all my processes. I can do the final limiting. Now, there's a few things we're going to be doing here. Some really good housekeeping. Now, overall loudness, you're driving a track like this. I'm going to talk about the maximizer as an EQ, which is going to be a bit weird. You'll understand once I get through it, and the vintage limiter. Now, I'm going to start with the vintage limiter first because this pairing—the vintage limiter, the maximizer—Will Borsa told me ages ago to do, and I am uh sort of somebody who's stubborn and stuck in their ways, and I have to discover stuff for myself in order for it to be valid, and he told me and I sort of ignored it, which is just—he's one of the most beautiful people. I'm just sort of a stubborn like that in my own brain because I'm got my way of doing things, and unless I discover it for myself or figure it out for myself, I don't I sort of turn off, which isn't a good thing. Anyway, it is actually a really good way—a good technique—to um limit things using ozone. You doing the vintage limiter, the maximize, and I'll show you why. Um, vintage limiter. The beauty of this is that as you approach the ceiling, okay, as you approach the ceiling of the vintage limiter in analog mode, you get these third and fifth order harmonics which thicken out the sound. Now, the byproduct of thickening out this sound. Okay, so we'll go over to here. This ramp -255 will just make a 440 Hz tone. The beauty of thickening out this sound is that there's -2. everything underneath it gets like a 2.3 dibel lift cuz it's almost like a soft clipping transfer into that ceiling. It's very soft round into that ceiling, but it's a couple ten of a decibel of extra loudness and the sort of warmth up against the ceiling of those third and fifth order harmonics. So, that's number one. That's why I use the vintage limiter. And I'll just quickly show you—I'll just quickly show you that vintage limiter how I'm setting it up. So, I've got a negative-0 threshold. I'll put the delta mode in, and I'm going to bring the the signal—the threshold—down until I can hear the delta mode taking out a lot of low end, and then I'll back it off. The reason why is I don't want my low end to get trimmed. That's number one, especially with electronic music. Mastering engineers push it too hard into the limiter or not listening to their limiter, and then the low end gets thinned out, and they send it back, and the client's like, "Well, the low end sounds thin." It's like, yeah, it sounds thin because you've not overlimited, but you haven't listened to what you're doing with your limiter. So, let's have a listen [Music] now. Okay. Okay. Okay, it's just hitting those big downbeats. Everything else is not really getting touched. What you're hearing in the delta signal is actually that soft curve or that transfer into the ceiling where you got that 2.3 dibels of extra loudness. What you're hearing on the downbeat is just it clamping down on anything that's just a little bit too much. And I do that on purpose because my main limiter I'm using for EQ. Remember I said that? What do you mean for EQ? Let me show you something here. Let's bring in plug-in doctor again. And and and I like to show what's happening so that way when we're listening for it, we understand what we're listening for. I don't do this in the actual session where I'm pulling plug-in doctor. I'm doing this for demonstrative purposes for you guys. So, let's pull up—I don't want you vintage limiter. You are no good to me. I want my baby maximizer. Let's pull it up. Remember I said I want to use it as an EQ. Now, overall, we've made this big and bombastic. But the realistic thing with this mix is when it leaves into the real world, having this much low end. I I just like—you got to have a conscience where you're mastering, and you want to get excited about the music, but you also got to be realistic about where it gets played. I just wanted to tuck that low end ever so slightly back. Just a little bit back. I didn't want to use an EQ. Why? Cuz I just didn't want to use an EQ. Simple as that. Now, the reason why the low end gets choked a little bit. So, we'll look at this Hammerstein graph here, that fundamental there. Okay, we're in -2 dibel output. Um, and that's -2.7. So, it's like having a half a dB to 1 dB shelf on all the subs. Okay, the reason why those low ends get choked if we go into an oscilloscope, let's look at 35 htz, okay, is because the symmetry gets pushed out. It's not symmetrical anymore because the speed of the limit is moving faster than the oscillation or the frequency. Um, it's something I talk about in my advanced mastering details course which comes out on Thursday if it hasn't already come out by this video. Otherwise, I'll leave a link to it in the description below. Small little plug there. Now, let's get back to this. Because the limit is moving faster than the lowest oscillations of frequencies, it tends to shift. That's how we get harmonics, but also the fundamental—which is that purple line G1—that fundamental gets attenuated. So what I'm doing is I'm not only getting extra loudness by using this limiter. I know for a fact this style of limiting ducks that low end a little bit. Okay, depending how hard I push it, which means I can consciously use this mode. Now the balance mode is probably actually the pumping mode is the most neutral to the low end. So you get the flattest response. The balanced is definitely balanced; you you lose a little bit of low, but relatively the amount of harmonics that are coming back in there, it doesn't feel like it's getting choked. So, that's just helping me control that low end whilst getting the loudness, which is a nice sort of tradeoff. Now, we've got that all dialed in. Let's just have a listen. Okay. [Music] Okay. Okay. [Music] There you have it. How to work with a limited mix or how at least I worked on this limited mix. Some of the considerations you can take in the tone and manipulating the tone and harmonics and dynamic depth are all there. Okay, it's your playground. If if anything you take away from this is when you get a limited mix, throw it in. Get the clip gain, drop it down 6dB, and then the world's your oyster. You just have to think about the music, and then you're good. Anyway, until next time, take care.
All right, I'm going to show you this mastering chain. It's drum and bass, but it is juicy as hell. So, you're not going to want to miss out on this. It's not going to want to—not going to want to miss out on it. Before we dive into it, the Studio Flow Masterass comes out in 2 weeks. And if you want early access to it, you know what you need to do? Sign up to the newsletter. panoramamastering.com/newsletter. Even if you don't want to join the masterass, that's perfectly fine because it is still the best place for working audio professionals to get the drip. And I love doing it because it's so much fun to explore these details free of the constraints of, let's just say, other platforms. Fair enough. Now, let's jump in to this chain. First up is standard clip. Now, you know, I use standard clip a lot, and I'm going to show you what it's doing, but also I'm
Going to answer. So many people have been asking, where do you put it? Why do you put it at the start? Why do you put it after things? When, like, what's your protocol? I'm going to detail that.
So, first, let's just look at this mix. This is before the clipper. Lots of offshoot peaks between the kick and the snare just poking out. Very sharp, very sharp peaks. And you know, like they're they're they're absolutely crazy short peaks. Like if you look at this one, it's like four or five samples just just floating up in the air. Okay, just maybe about a dozen, but still a dozen in 48,000. Not a lot. What I did is I hard clipped it. Perfect. It brought me six like I was able to gain up. So I'm clipping at 10 dB. So I've got 10 dibels extra of headroom to push it up. So I put my crest factor to -12 after doing that. So that's the first thing.
Now protocols with clippers. I'm always placing clippers second in my chain. I always leave a blank slot before them. Now there's a few reasons. I frontload my masters with a clipper because for dynamic processors like compressors, multiband processors, parallel like um D1 parallel processing, for any processor that may rely on peak detection circuits for compression, I want the peaks that go through that detection circuit to be consistent. Okay. And you tell me which one is more consistent; this controlling a compressor or this controlling a compressor. Okay, so that's why I clip up front.
Next step in the protocol, why is it second in the chain? The reason why this is second in the chain, I've spoken about it before, is if we place any EQ filters or filters that shift the phase of the waveform, the clipped portion stays at the top. If you clip after the phase rotation, if you turn those around, and I'll put an oscilloscope here and show you. If I turn those around and place the clipper first and then any EQ processing, oopsies, I'll have to answer that later. And then place any EQ processing afterwards, what happens is the peak actually shifts. Okay? So, you generate a new peak with the phase shift. The clip portion goes onto the side. So, that's why it's always second. Okay, number one. So, I've got consistent compression that it feeds. That's inconsistent for the compressor. That's consistent. And the second one, I'll just show you this now. I'll get uh where we going to put this. Let's do it here. Let's put a signal generator. So, here we go. We've got the setup. We've got a 50 Hz sine wave. We're going to clip it. Okay. And then after the clip, we're going to add a filter. Uh not that high. Going to bring that down. Okay. Now, let's bypass that. It's generated a new peak because it shifts the phase of the waveform. All right. See that? That's hard clip not going into an EQ. That's hard clip going into an EQ. So, that's why I don't do any shelving filters or pass filters after standard clip or after a clipping process. So, I'll get rid of that now so we can get back to looking at the rest of the chain. But that's how I frontload my mastering process um with clipping. So, we'll just get rid of that. It's good. They're all inactive. Beautiful.
Then after that, I've got the god particle. Now, the reason I got the god particle is I actually really like the sound of this. And it's not necessarily mastering per se, but I just like the sound of it. So, I've put that on. So, now we're over to ozone. I've got a few modules here. Um, the first one is clarity. Now, the reason why I put this in is because the upper mids of the mix were very delicate. And the way I balanced them was was good. Like, don't get me wrong, I liked it, but I felt like EQing any individual element or doing any broadstroke EQ, I don't know, it just felt very aggressive, like artificial. So, I used the clarity module just to experiment and see where it would take me. And it took me to a really nice place. So, I'll put this in solo and we'll turn it start with it off and then we'll turn it on and you'll hear the upper mids just a bit of space pull out in a really nice [Music] way. But I know you really show [Music] me. Beautiful. I really like that. A little hint when using clarity. Uh I never really go above 10k because it always boosts for some reason in the algorithm. So, I always have that top always pulled in.
Uh, the next one was stabilizer. And again, this is just a little bit of shape to it. Why reggae? I actually struggled with stabilizer. Um, I struggle with stabilizer a lot. I don't always use it. I don't always not use it. I tend to set it and then go through all these styles to see which one feels good. It's more a mixing thing. Actually, these two processes, clarity and stabilizer, I implemented while I was mixing the record just to get things feeling right. And I think that's really important when it comes to mastering is that we've got the technical side, but we've also got the feeling side. And when we're mixing, I personally like to lean on a little bit of my mastering work to fill out some of the space because if what I'm hearing excites me, it keeps me captivated in the mix and keeps me moving. So clarity and stabilizer was set up pretty early on in the mix. Um, vintage limiter, that's pretty normal. The maximizer, I did a few little things with this, which is a little bit different. Um, we can see these down sections here, these softer sections. So I've enabled the upward compressor there. Then that's how you get to the final mastering chain. That's what it looks like. The clipper into the gob particle into ozone 11. The final one here, maximize, is just a true peak limiter. And then you're at it. Thanks for watching the mastering chain. I'm just going to vibe to this now. Back in the first chorus. Let's go. [Music] I need something more trying to [Music]
All right, that wraps up my journey for last year. Now this year, why haven't I done any mastering updates? You're lucky. You get to stay to the end of the video to find out why. It's because there's been a few things I've been doing which have been including plugins which aren't released. So, I can't show them because they're embargoed. I can't share unreleased plugins which are still in development. They're not mine. They're other companies I'm working with and um they're great plugins and I'm going to share them as they get released. But I am going to be doing more mastering updates in the later half of this year once all those plugins are released. All right, until next time. Actually, join the fiveday mastering sprint. Take care.