Published July 05, 2022
Do you think your music could sound better?
In this episode:
Andy Hagerman's career in music and education
Recording myths: Do recordings sound different in different DAWs?
The digital summing engine
Analogue and digital recording levels: saturation and clipping
The importance of clocking for audio interfaces
Recording levels and bit-depth: 16bit or 24bit?
The magic of 32bit floating point. Where it's beneficial and where it's pointless
Things you won't find in the manual: The inner workings of Pro Tools.
Latency - how to overcome the pesky echo in your headphones
How to get an authentic-sounding MIDI performance
Andy's tips on delivering great performances
Contact Andy for composition and music services: firstname.lastname@example.org
Andy Hagerman on YouTube: Pro Tools Answers
It would mean the world to me if you'd consider giving this podcast a 5-star review. Thank you!
How to Subscribe, rate and review this podcast (in less than 40 sec)
Listen and subscribe to the Production Talk Podcast within your favourite podcast app:
Contact the podcast host Yarn at mixartist.com.au
#AVID, #ProTools, #AndyHagerman, #DAW, #ProToolsEducation, #ProToolsAnswers, #HomeRecording, #Microphones, #MixedByYarnTheMixArtist, #MusicProduction, #MixArtistDotComDotAu, #Mixing, #ProductionTalkPodcast, #Podcast, #PodcastProduction, #ProductionTalk, #Recording, #SelfProducingMusician, #SoundEngineering,
Disclaimer: The Production Talk Podcast is independent of (and not related to) my teaching responsibilities at SAE.
Transcript (auto-generated by a robot - please forgive the occasional error):
Jan 'Yarn' Muths: Welcome to the Production Talk podcast with me, Yarn, of mixartists.com.au. In this podcast series, we celebrate the modern way of producing music. We want to talk about all things related to songwriting, recording at home and music production. So, if you produce your music at home, this is the place to be.
Please subscribe and recommend this podcast to all your friends.
This is the Production Talk Podcast episode 50.
Jan 'Yarn' Muths: Welcome back to another episode of the Production Talk podcast. At the beginning of this episode, as always, I would like to acknowledge the traditional owners and custodians of the country that this following interview is recorded on, the proud Arakwal people of the Bundjalong nation and pay my respects to elders past and present and emerging.
Jan 'Yarn' Muths: And on the same note, I would to like to point out that NEIDOC week is coming up, it's a week where we traditionally celebrate the achievements of our indigenous population. And, uh, I would like to invite everybody to tune into, uh, especially our music from indigenous artists over the next couple of weeks.
Jan 'Yarn' Muths: But let's move on to today's episode with me is Mr. Andy Hagerman of ever welcome to the podcast,
Andy Hagerman: Hello? Hello.
Jan 'Yarn' Muths: It is great to have you on board. Um, last week I had an amazing interview with a very creative person with, uh, Pauly B. And, um, we ended up towards the end of the episode talking about some technical stuff and I was just dying to jump in and throw some technical stuff, um, uh, towards the listener.
Jan 'Yarn' Muths: But then I noticed that it really wasn't about me, but more about poly. And I also know that I'm really bad at explaining things in simple terms. I really like to drift into really complex things very quickly. So, Andy, I know that you are a genius at, uh, phrasing difficult content in, in words so that even my grandma would understand and
Andy Hagerman: I'm pretty sure you're exaggerating and completely wrong, but let's go with, let's go. Let's go. Let's go with your endorsement, even though everybody will soon realize that you're completely wrong.
Jan 'Yarn' Muths: okay. Fantastic. Look, um, Andy, first and foremost, uh, you are in Tokyo. Japan, is that right?
Andy Hagerman: I am in Tokyo, Japan, which has the distinction this week of being four or five inches away from the surface of the sun. It is so hot here. It's unbelievable. My, my electric bill is gonna be as, as, as big as some GDPs of nations. It is just those stinking hot. Yes, I'm I'm I'm in Japan and, uh, yeah, I've lived here for 17 years.
Jan 'Yarn' Muths: For 17 years. Well, it is winter in Australia and it's way too, too cold for my liking. It's only 17 degrees. It's almost freezing for my liking. So send us some warmth our
Andy Hagerman: it's hitting 40 in some
Jan 'Yarn' Muths: 40. Uh, I love
Andy Hagerman: nuts.
Jan 'Yarn' Muths: Look, you know, I'm a sucker for the Australian summer. It's never too hot for me. 40 degrees is right on my alley, but, um,
Andy Hagerman: you what? Get in a plane. I will too. I'll go to your house. You go to mine, let's stay there for the next month and a half.
Jan 'Yarn' Muths: How swap? Yeah. Good idea, Andy. Um, you are, um, you are an American citizen, aren't you not, or the Roman correctly? Yes.
Andy Hagerman: I'm a re I'm like what I call a recovering American.
Jan 'Yarn' Muths: A recovering American well it's today, today is the 4th of July. So happy independence day to you.
Andy Hagerman: is the 4th of July, you know, I've, I've missed out. I mean, I, I, I live in Japan, obviously. I've lived here forever. I I've just lost track with all the, you know, the, the American holidays. I, I have people go, oh, happy Thanksgiving. Like the wet now. so, yeah, fourth July. I just noticed I, as you set it, I'm just looking up there and it is July 4th and there we go.
Jan 'Yarn' Muths: Cool.
Andy Hagerman: and I'm working.
Jan 'Yarn' Muths: Look, Andy, would you mind to, to share, um, some background about yourself, your, your professional career, what you are currently doing at avid and you know, where you're coming from
Andy Hagerman: Sure. So, um, so I'm a man of a certain age and, and that's important to me professionally because, you know, I was, I was coming up to where, where Jan is before people had computers right before, um, before Daws were even around. And, and when you could still buy tape. Right? So, so my, the beginning of my life was as a performing musician and as a composer and still is.
Andy Hagerman: And, um, and when the technological revolution started coming to, into Vogue in, in 1983 onward, I was in college. So I had the opportunity to kind of pick up the technology as it was being developed. Right. So there was no books, there was no classes forget about it, but there was, you know, in, in my college at Northwestern, there was just a, a real, enthusiastic, small bunch of geeks who were, who just kind of ate up the next thing, you know, de DAS ADAS, you know, spit.
Andy Hagerman: If whenever it came out, we were like all over, all, all over it. And, um, middy, you know, the whole nine yards. And, and so I, I kind of self-taught, um, taught, um, self taught myself. Is that the right way to say it? I guess not. Um, I was self taught. There we go. I was self taught in, in the ways of, of how. The music technology could let me do the things that I couldn't do well, but better.
Andy Hagerman: Like, so I'm, I'm, I'm a good composer. I, I think, but I'm a terrible keyboard player and, and MI just swooped in and saved me be for the, for the only reason that you could play it slow when you're recording and then speed it up later. That's all I needed. That's all I needed. I have terrible handwriting and, and so sips and finale and all those other applications it came in.
Andy Hagerman: So I've always kind of been the guy who I don't really care about technology except for how it helps me be creative.
Jan 'Yarn' Muths: Yeah.
Andy Hagerman: Right. So, so, you know, the, the, the, the DSP chips on, on an HDX card and, and, you know, what, what, what I've got on the Apollo twin and the, the, all that stuff could not possibly care less up to the point where it helps me to, to re you know, to, you know, get, get what I have in my mind out of the speakers.
Andy Hagerman: And that's what I still do. And, and right now, so you asked me professionally, um, I I'm now in charge of, and have been for a long time in charge of the audio curriculum for avid. Um, so that they're the, they're the, the company that makes pro tools. And, and so all the pro tools, books, all the sub books, the venue books to a certain degree, I, I coordinate that, um, curriculum and certification and all that stuff.
Andy Hagerman: And I'm also writing some of those books. I'm I just got done writing the new, uh, pros 1 0 1 book, and I'm working on the protos one 10 book, which explains why I have no hair, which you can't see because it's an audio podcast. But trust me, I'm I'm as bald as a Cub ball.
Jan 'Yarn' Muths: Yeah. Well, your impressive beat makes up for everything.
Andy Hagerman: do. It's like all my hair fell off my head and just kind of landed on my chin.
Jan 'Yarn' Muths: classic classic. And do you find enough time to be a musician these days or is it all just work
Andy Hagerman: boy, I'll tell you, you know, it's tough because when I, when I was, um, you know, before Corona, I was traveling so much that you can't really join a band. Right. Um, and, and because writing a book, anybody who's ever, anybody who hasn't written a book thinks, oh, I know all this stuff. And to take about a weekend, anybody who has written a book knows that it is tremendously time intensive and you have to shut out the world and you it's, it's just, it's just a hard slog.
Andy Hagerman: And, and so, so taking on these book projects has been great, but I, I really, the, the most creative stuff I've done is playing bass and, and, and creating stuff for the exercises in the book. And I did, I, I, I actually did one big bad chart that I'm super proud of. Um, but it was, I, the inspiration of it was what's to do an exercise with the book.
Andy Hagerman: It just happened to be good, but. When I get done with this book. And normally I usually like to keep a good project going, you know, kind of in the background. So anybody who needs a pro tools or a musician guy, that's Andy man. JCOM dot any JP contact me. I've got reasonable rates.
Jan 'Yarn' Muths: Look, um, put that in the, into the chat then from there, I can copy it into the show notes so everybody can just finish this episode, head over to the show notes and click a link to find
Andy Hagerman: shameless. Self-promotion
Jan 'Yarn' Muths: yeah, look, I would like to, um, first go through a couple of standard. Um, well, myths and misunderstandings, maybe just a couple of subjects that are come up against again and again and again.
Jan 'Yarn' Muths: And I thought, you know, being the official Protus educator of the world, you're probably the best person to, to answer this for good for us. So, um, a question that I've come across, uh, a couple of times is that, um, when people record at home and then switch between, uh, DWS from one to the other. So you might wanna start, uh, recording your guitar into D w a and then switch to, to D w B.
Jan 'Yarn' Muths: Is the sound file recorded any different? Is there any difference in tonal quality?
Andy Hagerman: the sound files perspective. Absolutely not. It can't change. Right. Um,
Jan 'Yarn' Muths: expand on this. Why is that?
Andy Hagerman: because it's be because if, if it were to change, so, so, so you record a sound file. Um, you record a sound file into, let's say logic let's let's let's, let's put a name to it. Okay. You record it into logic. It sounds great.
Andy Hagerman: Right? And, and that sound file is a sequence of ones and zeros. Now, anybody who's ever messed around with, with the, the deep data of files knows that changing a one to a zero or a zero to a one more often than not results in a invalid. Right. So, so, so usually any, any subtle change will make, won't make it sound different.
Andy Hagerman: It'll just make it corrupt. Right. Um, so, so that file is, is a sequence of ones and zeros. You move that file onto another computer. Does it, does it change size? No. Does it change the orders of the ones and zeros? No. So the file itself is completely unaltered. It has to be, if it was any other way, then, then digital technology beyond audio wouldn't work.
Andy Hagerman: This, this stream wouldn't work, you know, word documents, wouldn't work whole things. Um, now, if it sounds different to you, which is what people are saying is like, it sounds different. To me, that's a function of a lot of different things that can be a function of the mix engine of the D a w that can be the function of the, and, and this is something that a lot of people overlook the components in your interface, specifically the clock.
Andy Hagerman: Um, the, the, the clock is tiny, tiny little, you know, ugly little chip that's on, on the board, controlling the timing of the samples. Right. And it sounds like that's a nothing job, but it is possibly the most important job of any interface and, and people will go, well, my stereo image just sounds like it's, it's not as wide.
Andy Hagerman: Well, that's, that is the clock. That's the clock having jitter. And so can a file sound different on different systems? Oh, absolutely. But is the file changing? No, it's not.
Jan 'Yarn' Muths: Okay. So if I left, let's say the microphone unchanged, my preempt unchanged, the audio interface is the same, but I just switched from the first D w to another one. The actual fire recorded wouldn't differ at all.
Andy Hagerman: Wouldn't differ at all. The, at that point, if you're dealing with the same system and the only thing you've changed, um, it is the mix engine of the DAW, which I'm guessing is the question. You'll be asking me next.
Jan 'Yarn' Muths: will get to that a little bit later for me, but, um, so the mix engine is basically affecting a sound only after it's recorded. So it might affect what we hear, but not actually how the file no store to the hard drive is, uh, is, is sounding. Yep. Get
Andy Hagerman: Yeah, no, that's and that's a good point. And this is something that a lot of people overlook when they're recording is when you, when you record, it's basically going from the interface to your hard drive, right? It's not going really through your DAW. It's your Daws kind of managing the job of, of the interface, but it's not, you know, it, it's not affecting anything.
Andy Hagerman: So, so if you bring down the level or you play around with the level on a track, that's recording, or if you put a plugin on there, it doesn't affect that at all. So really it's the interface that, that controls that, um, that file going to the hard drive when you, when you listen to things, you're listening back through the DAW and that's, that's where the changes, um, you know, between Daws can come into play.
Jan 'Yarn' Muths: good. All right. The reason why I asked this question is that I once attended a master class where somebody who has sh named told me that, you know, he did exactly that switch between two DWS and immediately felt like one sounded significantly better than the other. And I was scratching my head.
Jan 'Yarn' Muths: Well, it.
Andy Hagerman: he's, he's completely right because of the, but not because the file changed it's because the engine that's playing it back changed, like it think of it this way. You've got a brand new record. Right. You put it on turntable a
Jan 'Yarn' Muths: Mm,
Andy Hagerman: on and then, and then you just, and then the whole system stays the same.
Andy Hagerman: Right. But you replace the turntable with another turntable. Could it sound different? Absolutely. Absolutely. It
Jan 'Yarn' Muths: Makes perfect sense. But the record is the
Andy Hagerman: it's but the record's the.
Jan 'Yarn' Muths: Okay. Look, what made me a bit suspicious is that the next thing that's, uh, person who has shown name said is that, you know, he then decided to, you know, drive the preempts nicely into clipping for the beautiful and analog warmth that he got.
Jan 'Yarn' Muths: What's your take on that?
Andy Hagerman: Well, first of all, if you're dealing with, uh, uh, a seminar from somebody who shall not be named, I don't think that Voldemort should be giving seminars. Okay. I'm just saying, right. I'm gonna name 'em Voldemort. We all know who, who shall not be named. It's named Voldemort. Um, so VDI I'm throwing you under the bus.
Andy Hagerman: Um, so there's a difference between saturation and clipping. Okay. And, and so do you mind if I go down this rabbit hole,
Jan 'Yarn' Muths: Um, yes. Yes. For, for a little while,
Andy Hagerman: Okay, I'm gonna go down this rabbit hole. Let's see how let's help briefly. So when you're dealing with tape, when you're recording the tape, um, tape is, is magnetic particles that are, are being realigned, right? So, so you, you know, it, it has a noise floor that de that, that digital doesn't have to that, to that degree.
Andy Hagerman: So you wanna make sure your signals are loud enough. So it kind of hits this magnetic tape sweet spot where it's fairly transparent. It doesn't sound any different, really it's, it's, it's as accurate as tape can be. Right. And then what you do is you push it just a little bit. And what happens is the magnetic tape, like a sponge cannot absorb any more signal.
Andy Hagerman: It does. What a sponge does when it can't absorb more water, it is. Saturated,
Jan 'Yarn' Muths: Yeah.
Andy Hagerman: It's not distorted. It's just, it's just going to ignore it. Can't store that additional stuff right now, digital. And, and by the way, when we have, uh, saturation, you know, that can give us that warmth, that it does things psycho acoustically that we super super love.
Andy Hagerman: Right. And that's something that people, people will point to. And they'll say analog's better because of that over anything else. Right. Um, and then you go to digital. Now, digital is painfully honest, right now. It doesn't have a lot of noise floor has, has a very, very small one bit noise floor, which if you're recording at 24 bit is way below what you can hear.
Andy Hagerman: And, and it, it tends to be very accurate and get more accurate. The more bits you use right up to, up to a certain degree. I mean, you don't have to record as hot as possible without clipping all the time, but it's, it's very clear. It's very neutral. It sounds exactly like what, or as close to exactly as we can with the technology we have now, it sounds like what you put into it until you give it too much data.
Andy Hagerman: And when you give it too much data, it doesn't saturate. It rejects, right? It's like, okay, I can go up to minus zero DB. If you give me too much, I'm gonna stay at minus zero DB and go F yourself. And what you're gonna get is it's going to, it's going to do what we call clipping and clipping basically is a straight line from one.
Andy Hagerman: Sample to another, at the highest amplitude and what it does, what that does is it, it says, speaker push out and stay pushed out, which is not good for your speakers. And it doesn't, it doesn't have that sat. It's not gonna give you a saturation kind of feel. Now a lot of people use clipping and saturation interchangeable.
Andy Hagerman: Um, and so maybe that's what this vort person did. Um, you know, maybe what he did is he saturated it and he liked the sound of it. Um, but generally speaking, when you actually digitally clip, it's not a sound that people generally go for in terms of, of warmth and fullness.
Jan 'Yarn' Muths: Good. Thank you very much for that. Um, I'm definitely on board with, uh, staying away from digital clipping. There's very few situations in which I find that could be useful. Um,
Andy Hagerman: no, I mean, there's, there's effects like, you know, you know, Trent Resner makes very good use of it intentionally,
Jan 'Yarn' Muths: yeah. Yes. Intentionally yes, yes. That that's a good point, you know, but, um, yeah, you're right. He probably doesn't, you know, tried once and that's it. He probably. Tries several times and pushes a bit more, a bit less until it sits just right. So it's very intentional, as you said. Yes. Good. All right. The next question that I have for you, um, let me just assume that we have a very consistent player.
Jan 'Yarn' Muths: Let's say a really good keyboard player who plays pads and they come in at pretty much the exact same volume. Um, and now instead of exchanging a D w we stay with the same one, but we exchange audio interfaces. So that's the only variable here now. Um, will the recording level in our D w be always the same,
Andy Hagerman: the recording level. Isn't in your DAW. The recording level is never in your DAW. The recording level is always in your interface. So, so this is, this is the thing, is that, um, when, when you've got so, so I'm gonna use pro tools as an example, I'm not trying to sell it. I'm just gonna use it as an example.
Andy Hagerman: But if you, if you hit the record button on, on a track, um, the fade, the, the little fader thing, uh, turns red, that indicates that that's a recording track. That fader, however, is not changing the level of the signal that is going to disk that is changing the level of the signal. That's going out, the speakers that you're listening to.
Andy Hagerman: So if you see your meters, which are, are giving you an indication of, of what's coming in, if they're clipping and you bring down that fader on the recording track, you've done absolutely nothing of value, right? You've made, you've made it softer coming out of your speakers, but it's still clipping going into the hard drive, which is what you don't want.
Andy Hagerman: So you, you said keyboard player, so right. It's a line level signal,
Jan 'Yarn' Muths: So exactly. I was just about say, you know, let's say we have a line level signal going into a different interfaces, all at line level and.
Andy Hagerman: Well then turned down the keyboard.
Jan 'Yarn' Muths: Yeah, good point. But, uh, let let's say if, you know, the keyboard player was giving us a healthy level. Um, and I just exchanged, let's say from an evident interface to an Apogee, to, I don't know, Yamaha, or what have you would the level that we see visually in our DW or meter and O DW would, would that always be the same?
Andy Hagerman: Probably not.
Jan 'Yarn' Muths: Mm. Can, can you expand on this? Because you know, I've, I've come across people teaching religiously that, you know, certain analog level equals a certain digital level. But my understanding is also that that's not a given
Andy Hagerman: Well, you know, keep in mind that, that your interfaces are always interpreting a, a signal coming in an electrically variable signal. So basically, you know, the signal coming on the line, a level input is neither analog or digital it's, it's, it's a stream of electrons. Right. And what happens. Your audio interface takes a look at that and is going to then convert that to ones and zeros.
Andy Hagerman: Now, um, that's subject to a lot of different things, right? Mo and again, I hate to be sound like a broken record, but you know, your clock, which controls the accuracy of the timing of samples being output is also controlling the accuracy of samples going in. Right. So when you record on another interface, um, could the, uh, could the, you know, could that jitter affect the, the stereo image of your sound?
Andy Hagerman: Yes, it could. And can a narrower stereo image sound louder? Yes, it can. Right. Uh, we're talking about very small amounts of, of volume change, but the total change can be sign.
Jan 'Yarn' Muths: Mm. Okay.
Andy Hagerman: Which is, which is why, you know, and, and forgive me for, for preaching a little bit here.
Jan 'Yarn' Muths: go for it.
Andy Hagerman: um, if there's one place you don't want ACRI, it's the audio interface, because you can have a great microphone.
Andy Hagerman: You can have, you know, all the, you can have all the best stuff. You can have great speakers and all that stuff, but if the stuff getting into your system, isn't the best you can possibly get. That's that's a weak link that you can't afford to have.
Jan 'Yarn' Muths: Okay, let me just, um, expand a bit on this question. So some people say that, you know, uh, negative 18 DBFS is the reference level. There's now this term of, of, you know, a reference level is, is, uh, something that I come across occasionally. And I guess that relates to line level, um, being traditionally plus four DBU, we're getting technically here.
Andy Hagerman: Yeah. Do it.
Jan 'Yarn' Muths: I, I found that, you know, when you let's say, um, change from, let's say an interface, which is traditionally, you know, the H DS, I believe are calibrated to negative 18 DBFS if you feed it plus for DBU, and that could be something else on another interface. So I now found some AEs. Are then shifted by two DBS to negative 16.
Jan 'Yarn' Muths: Is this something that you've come across?
Andy Hagerman: So, so that MI I think we're, we're shifting specifications. So when we talk about 1980 minus 18, sometimes minus 20, sometimes minus 16, it, it varies by region. And what you're talking about are your loudness targets, right? So you're, you're your, the way that you monitor that's, that's really what those affect more than anything else.
Andy Hagerman: When you record, what you wanna do is you wanna make sure that you've got nice hot levels with enough headroom that you're not clipping and you're done. You're done. There's no, there's no target. There's no target recording level, like there was with, with analog. And I'll tell you why there was with analog in just a second, but there isn't in digital. Simply because there's no, there's no specific non-linearity here. Right? So when we talk about the loudness levels, you know, minus 18, you know, you're, you're talking about speaker calibration, right? Not recording, you're talking about speaker calibration. And so those speaker calibrations, those are how you should set your system up so that you know, that the levels that you're hearing are going to translate.
Andy Hagerman: But when you're recording, if I record, if I record a track, that's, that's 10 DB. You know that that's peaking it around 10 minus, sorry. Minus 10 DB. Am I completely fine? Completely. Am I gonna bring the level down? Absolutely not. Right. Because you bring the level down. It's it's going to bring up whatever ambient noise floor, right.
Andy Hagerman: In terms of a percentage.
Jan 'Yarn' Muths: yep.
Andy Hagerman: Now you don't now somebody might say, oh great. I, I get what Andy's saying. I'm gonna record always as hot as possible without clipping mathematic. By the way, back in the old days of, of only 16 bit audio, we used to have to do that because the dynamic range of 16 bit audio was less than the dynamic range of the human being.
Andy Hagerman: However, if you record it, I'm sorry, I'm getting you really technical. If
Jan 'Yarn' Muths: go for it. Yeah.
Andy Hagerman: it 24 bit, your dynamic range from top to bottom is 144 DB. Your dynamic range is a human being is only 120, so you've got plenty of headroom to play with, right? So, so you could, you could completely have a track that is, is, is clip, not clipping is peaking at minus 10, minus six, minus five, minus whatever.
Andy Hagerman: And you're not gonna lose any, any significant quality, but I don't want you to go around thinking, oh, now I have to change my level down to minus 18. That is not how recording workflows happen
Jan 'Yarn' Muths: yep.
Andy Hagerman: However, uh, that used to be, um, that used to have more of a place when it came to analog, because when you're dealing with faders and we're dealing with an analog console fade, each fader is, is essentially an amplifier, right.
Andy Hagerman: And amplifiers, analog amplifiers have what's called, uh, non-linearities. They, they will distort. Right. And all amplifiers have a degree of non-linearity. And so what people used to do is they used to try to keep their recording levels to the point where their faders didn't have to move too much. And so what that would do is that we keep the, the non-linearity from the faders, from being too much of an issue.
Andy Hagerman: Right. That's it, it's kind of, it's a version of game staging. Right.
Jan 'Yarn' Muths: Yes.
Andy Hagerman: And, and so people used to do that because of the non-linearities that are introduced when you move a fade around now in pro tools or in any, any DWM. When you move the faders inside approaches, you are not introducing non-linearity in the same way because it's not an analog system.
Jan 'Yarn' Muths: Well said well said, wow. Um, look, um, I I've done some, uh, episodes on similar subjects before. I might just throw a couple of links into the show notes as well. If, uh, anybody wants to know more about those things. Um, there's one thing that I want to go in or, you know, ask a couple more questions about, you just mentioned 16 and 24 bit, and we all know this when we create a new project, uh, hopefully we make a decision right at the beginning, uh, to either record 16 or 24.
Jan 'Yarn' Muths: So, um, am I correct in saying that, uh, 16 bit is not your choice? You choose 24. Yeah.
Andy Hagerman: you want, you want the most data, sorry. You want the most relevant data that you can possibly get. Um, now, um, my dynamic range and yours is, is about 120 DB at our prime. Right. And, and, and I'm not at my prime and, and you still are. So, so you're probably, you're probably able to tolerate a lot a, a, a higher dynamic range than I can as an old guy.
Andy Hagerman: That's fine. Right. Um, I, neither one of us can tolerate much more than 120 DB that becomes the threshold of pain right now. When you're recording at, uh, 16 bit, then your di your first of all, let me back up. Do what, what's a good dynamic range for recording the tape. Do you think.
Jan 'Yarn' Muths: I think tape had about 60 DBS if I remember correctly.
Andy Hagerman: 60 or 70. And by the way, that would change during the day. That's why we had those D goers. Right? Remember we, we used to have to maintain those things at least in the morning. Right. Um, but, but on a great day, it would have maybe 60 or 70 DB of dynamic range. Um, and so that's roughly half of your dynamic range.
Andy Hagerman: So the noise floor was absolutely audible. Right. And we had to do a lot of things and that's why, that's why we, we, you know, we made sure that we recorded fairly loud, generally speaking, so that we would basically mask the noise of the tape
Jan 'Yarn' Muths: Mm
Andy Hagerman: now that that largely doesn't exist on a digital system. Um, and if you're getting a lot of hits and noise, it's probably your wires are, are next to AC or something like that.
Jan 'Yarn' Muths: right. Yep. That's right. That's exactly what we discussed last week with Pauly B. And he basically mentioned that digital doesn't have a noise flow anymore. Although it's it is, it is there, but it's so quiet that if gain stage correctly, it's irrelevant. Is that a fair statement?
Andy Hagerman: not exactly
Jan 'Yarn' Muths: Okay, please. Correct me.
Andy Hagerman: something that's misapplied, but, but let's, let's stay with the noise floor. So the, the, the, the built in noise floor of digital audio is the final is the least significant bit, right? So, uh, it's the, it's the 16th bit of 16 bit audio. It's the 24th bit of, of 24 bit audio.
Andy Hagerman: Now here's the thing this, that 16th bit of 16 bit audio is minus 96 DB, which is, which is a lot better than tape was right. And, and it's not a hiss. It's just a little tiny bit of distortion. It's rounding error, essentially when you're D when you're recording at 24 bit, that's at minus 144 DB. So it's, uh, if you're recording at a decent level, you'll never hear.
Andy Hagerman: It's just that, that, that, that becomes a non-issue in terms of, of, of the D a w um, so I would, now, now here's the other thing, here's, here's the, where you might go next and maybe I'm preempting your next question. Where does 32 bit floating point audio fit into this? Right. And here's the thing to remember is that 32 bit audio, I'm gonna say this right.
Andy Hagerman: 32 bit audio is 24 bit audio with a rapper.
Jan 'Yarn' Muths: Mm-hmm
Andy Hagerman: So, so, so, so 32 bit audio is a 24 bit. What's called a Mansa, right? That's the, that's the, the raw data. And that's really, what's being captured by the audio interface. And what's wrapped around. It is a thing called an X exponent, which allows you to shift the, the, the audio up or down, depending upon whether or not it it's, it's gonna be clipping.
Andy Hagerman: Um, 32 bit audio is tremendously useful in D a w mixing. And that's why it's built into most Daws pro tools being one of them.
Jan 'Yarn' Muths: that's right. Um, is it correct to say that, uh, today, um, most orient interfaces actually have 24 bit converters, 32 bit flung converters. Are sort of on the horizons. Some manufacturers have them, but they're not a standard yet. So
Andy Hagerman: They're Def deaf.
Jan 'Yarn' Muths: recording. Yeah. Sorry, please go for it.
Andy Hagerman: go ahead, please,
Jan 'Yarn' Muths: So recording, let's say 32 bit audio files with a 24 bit converter.
Jan 'Yarn' Muths: We could argue that there is not much to be gained here. Is, is that correct?
Andy Hagerman: No, we could state that. There's nothing to be gained there.
Jan 'Yarn' Muths: yeah. Cool. All right. Okay. Mm-hmm
Andy Hagerman: cuz you, could, you still get all the benefit of floating point inside the, the, the DAW. So you could, you could record and create a 32 bit floating point word, but you're, you're not gaining anything and you're arguably wasting space. Right?
Andy Hagerman: Cause the file's gonna be bigger. Um, yeah. And I, and I would say that, um, Keeping in mind that 32 bit audio has a 24 bit, man. Um, I would say that any interface that purports to have 32 bit floating point recording is not giving you the value you think it's giving you because the interface has to create that 24 bit word first.
Andy Hagerman: And, and that's, that's the data we really care about. Right? So with, with the technology we have right now, and I want, you know, could things change. Sure. Certainly, hopefully they will. They certainly will. Right. Um, but with the technology that we have today, um, recording at 24 bit, first of all, is completely adequate.
Andy Hagerman: It's more dynamic range than we need. And then you take advantage of what floating point gives you where you really need it. And that's in the DAW.
Jan 'Yarn' Muths: Okay. Good. All right. So let me just sum this up and, and please interrupt if I get this wrong, but you know, most of our converters input and output converters are 24 bid, but what happens in between, let's say in our DW, when we move a fader or a pan port, or at a compressor that's done at a higher bid resolution being 32 bit float, or occasionally 64 bit float, depending on, on, on the tool, but it's always only ever in between those two converters.
Jan 'Yarn' Muths: So the converters are, I'm not sure if I could call it a bottleneck at 140 40 B, but you know, there's, there's more dynamic range to be had inside the D a w am I on the right track here,
Andy Hagerman: And, and, and at this point I have to talk about ProTool specifically, cuz it's the da not I'm trying to, I'm not trying to sell it I'm but this is the D a I no is, um, inside each individual track. Um, the signal flow in each individual track is 32 bit floating point. And that's 32 bit floating point between each one of the, uh, inserts, your plug-ins, all that stuff, the fades a whole nine yards, right?
Andy Hagerman: That whole signal track or signal flow in an audio track, ox track, whatever, um, is 32 bit floating point, which means it's going to be extremely difficult to clip. And if you do clip like you clip on the input of a, of a, of a plugin or whatever, you can recover from that easily with the, with the gain input or the gain from the, the plugin.
Andy Hagerman: And there's no re there's no permanent damage done, which is not what it would be like with, with fixed point or with, with tape, frankly.
Jan 'Yarn' Muths: okay. So what you're saying is that once we, in floating point, we actually have numbers above zero
Andy Hagerman: Mm. Yeah. I, I, I don't think I've ever said it that way, but basically what the exponent does is it steps in when needed to raise the Mansa and, and, and, and extend our headroom when needed. Right. And, and so, so I would say that, yes, it, it, it, it allow, it shifts that dynamic range up when, when it's needed, usually it's not needed at all.
Andy Hagerman: Right. Um, but when it is needed, then, then it steps in which can give you a lot more clarity than people, uh, would normally have. Um, now 32 bit floating point. Avid determined. And I agree, um, was not sufficient when you're, when you're combining large numbers of tracks. So each track again has 32 bit floating point in its own signal flow on each track.
Andy Hagerman: When the track's output goes to the mix engine, that mix engine is 64 bit floating point, which is a ridiculous amount of, of, of, of headroom. Right? And, and again, you can go extremely hard with a large number of tracks. You just bring that master fader down. Boom, you're done. And, and, and there's no coloring.
Andy Hagerman: There's no, you know, people talk about not touching the master fader that not touching the master fater is an analog console rule. It is not a digital D a w rule.
Jan 'Yarn' Muths: couldn't agree more. Yes.
Andy Hagerman: And, and I'll say this in, in closing, not all Daws are built the same when it comes to the mix engine and that's why people choose one D a w over the other.
Andy Hagerman: So I'm talking about how pro tools does it, not all Daws have the same, uh, robustness.
Jan 'Yarn' Muths: Yes. Got it. Got it. Okay. So, uh, you just mentioned the summing engine. So if, if I let's say now use a center route snare to Rever, is that affected as well? If it, you know, then also sums with the Toms or other signals is
Andy Hagerman: Say that again. I'm not sure I understand the question.
Jan 'Yarn' Muths: So does that only affect, you know, my main output or the output that feeds my speakers or all the buses as well?
Andy Hagerman: Oh, everything, everything, anything that summed together is summed together, uh, in a, in, in, in pro tools. Again, I'm talking about, um, in, in a 64 bit world where it's very IM it it's, it's, it's very impossible. It's, it's almost impossible to, to get into unrecoverable trouble,
Jan 'Yarn' Muths: got it.
Andy Hagerman: but yeah, that, that, that covers the whole thing.
Jan 'Yarn' Muths: Wow. Okay. Love it. This is a really good discussion to have, um, let's move on from, you know, the theory to some more practical examples. Um, a lot of musicians who know, who start recording themselves often run into one problem called latency that you play your instrument, and then you put your headphones on and you hear yourself later.
Jan 'Yarn' Muths: It sounds like an echo. Can you explain any strategies to, to deal with that and, you know, overcome that because that can be a bit of a mood killer when performing
Andy Hagerman: Oh, it's definitely a mood killer. Oh, it's it's it's, you know, as, as a musician myself, it's, it's, it's, it's, it's ion impossible to play when, when latency becomes that much of an issue. So, um, it's generally speaking a hardware solution. Right. And there's a number of different ways you can do that. So, so, uh there's you know, so I I'm, I'm looking on my desk here and I can see my Apollo twin, which is a fine little interface.
Andy Hagerman: Right. And one of the things you can do with this fine Lu in her face, As you can take the input that you're recording and monitor it from the interface itself, not through the DAW. So you record into, into the DAW. I almost said pro tools, you record in the DAW, the track that you're recording to is muted so that you don't hear that.
Andy Hagerman: And then what you're doing is you're monitoring through the interface itself. Now some interfaces, and this is a relatively new addition, but some interfaces will then allow you to have a little bit of, of effect. So you wanna put some reverb on there and still keep low latency. You can, but you're almost at that point, you're starting to defeat the feature because once you start putting plugin plugins on there, you're introducing latency. As well inside the interface, cuz there's no, there's no free lunch when you get to it now, um, that works great with smaller interfaces with, with smaller recording, you know, for you and I doing these this podcast, it's completely fine. When you're recording a symphony orchestra, that's that breaks down.
Andy Hagerman: You can't have a symphony orchestra recording to 52 Apollo twins. No, it's not gonna happen.
Jan 'Yarn' Muths: Yeah.
Andy Hagerman: Um, and, and that's why people will, at some point go to larger hardware, DSP solutions, um, where, where there's so much DSP horsepower under the hood in terms of, of, of in, and I'm trying not to, to sell avid again, but your HDX cards and those things, those are specifically built with one, one primary goal over anything else.
Andy Hagerman: And that is to get that signal in as quickly as possible and get it right back out to the players. And it really doesn't go to the hard drive. It just, it it's, it lives on that DSP chip on a card, but that, that might as well be just another interface. It's, it's so close to, to the input and the
Jan 'Yarn' Muths: Mm. Okay. So that's a solution that would that's suitable for large scale studios. Um, but
Andy Hagerman: Sure. Yeah. But for smaller scale studios, use your use your, your buy and interface that allows you to monitor through it, uh, which almost all interfaces do. Um, and, and, and then if you feel strongly about, you know, having reverb on your vocal track when you're recording, which is reasonable, um, you know, there's ways to, to work around that as well,
Jan 'Yarn' Muths: mm, mm.
Andy Hagerman: almost no free lunch when it comes to smaller situations because you're at the mercy of the USB bus, you're at the, the mercy of, of the, of the computer CPU, the Ram, all that stuff.
Andy Hagerman: Uh, and, and that's not, there's no DAW that really fixes that without leveraging the hardware. Now there's an, you know, so there's the proto side. There's also a universal audio side, which is their Luna, D a w do you know what it does? It uses the audio interface. Just the same way that pro tools uses uses the, the HDX cards.
Andy Hagerman: So it's, it's not a software solution, it's a hardware solution.
Jan 'Yarn' Muths: Yes, that's right. That's right. So there's limits to what computers can do. Um, what, what effect, uh, has the, the Daws buffer size on, on latency? Do we need to consider
Andy Hagerman: Again, you're, you're asking, you're asking questions that differ by DAW. Um, and, and for the most part, a D a W's hardware buffer is the storage place for audio that's coming in, being processed, and then having to go out. Now, the, the higher your buffer is the more tasks can be, can be, can be, uh, accomplished without overrunning your buffer and causing problems.
Andy Hagerman: However, the bigger your buffer, the more time you're giving to your CPU to get those jobs. Right. And so the larger latency between an input time and an output time. So the problem has been that, and the recommendation has always been, is like when you're recording, keep your buffer small so that, so that audio gets into the buffer and gets out of it quickly.
Andy Hagerman: Oh, that's great. Except for you can't do much with it because you know, small buffer can't run too many plugins. So when you're done recording, the, the common wisdom is to then set your buffer to be as large as possible, which is gonna give you the maximum Nu amount of time that you're giving your system to do a lot of different, you know, DSP processes.
Jan 'Yarn' Muths: Got it. Mm-hmm
Andy Hagerman: some Daws, one DAW, it begins with a P ends with row tools, um, has two buffers in it. And this is something a lot of people don't know. There is a buffer for record arm tracks, which you can set low. And get low latency, but there is a hidden buffer. That's always active on any track. That's not recording back and it's always set high.
Andy Hagerman: So you can actually, if you get the right DAW, you can actually get the best of both worlds. If there are two hardware buffers, if there's only one hardware buffer, then, then Jan, you're just right back to where you have to make that decision. When you're recording, set it as low as possible when you're playing back, set it higher and, and you just keep on doing that ping pong and you'll never get the best of both worlds, but there