Pro Audio Files

EQ Ear Training Premium Courses

Melodyne 4: Beginners Guide

Transcript
Warren: Hi, it’s Warren Huart here, hope you’re doing marvelously well. I’m sitting here with Stefan Lindlahr. Is that right?

Stefan: Nearly, yeah.

Warren: Nearly!

Stefan: Just say Stefan. [laughs]

Warren: Stefan from Celemony, and we’re going to be talking about Melodyne. How are you?

Stefan: Very well. Still a little bit jet lagged, we’re sitting in LA, and we just flew in.

Warren: Which is fantastic. And you’ve come in for NAMM?

Stefan: Exactly, yeah.

Warren: Which is where we met a year ago? Probably a year ago today. And you had just released, what is the latest version of Melodyne called?

Stefan: It’s version 4.1 we just released a few days before Christmas at the end of 2016.

Warren: And when I had seen you at NAMM, it was 4?

Stefan: It was 4.0. It was a major update, which was new at last year’s NAMM when we spoke.

Warren: And those of you that were watching me talk to Stefan about that will know that I was kind of freaking out, because we had taken, what was the Indian instrument?

Stefan: I think it was something like a tambura.

Warren: Yes, which was great, because I’d just finished Trevin Hall’s album, which was covered with Indian instruments, so I was all intrigued by it, and we were taking second and third harmonics, and creating other melodies out of it, and you and I had this whole long conversation, probably driven by my obsession, but talked about Bowie’s Berlin albums with Eno, and just what we could’ve imagined somebody like Bowie would’ve done with that kind of technology.

So what we want to do in this series of lessons, and this series of tutorials, is like, going on the basics of what Melodyne can do, and I get asked all the time, so let’s sort of jump in. Let’s jump in with some basic tuning tutorials, because look, one of the conversations you and I had on Skype before we set this up, and there’s going to be a big debate below with all the questions and comments, and it’s fine, debate it, is why do you tune, and when should you, and all of this kind of stuff, because I’ve been blessed, and I’m sure many people watching have been blessed to work with incredible singers, and those that know me know I’m obsessed with Freddie Mercury, who’s the greatest of the greats, and these are guys that were never tuned. However, they did punch in and fix.

They were coming in, and they were re-singing parts until they nailed it, and then moving on. One of the blessings that we had when you and I were kids coming up doing this is budgets were big. You could go in, a six week record was a standard album. I don’t know about people watching now, but I’m making records in days, sometimes weeks, but often in days, that do sound great, so these kind of tools aren’t obligatory things that we go in there and tune every single note. Used tastefully, they’re actually a wonderful way to take an incredible vocal and adjust maybe just a note.

Maybe when that singer went just a little flat, it was great one time, but the second time, it was a little too flat, and it’s not always about just snapping it back to the grid, but using it as a tool — as a tool to fit into the modern world where we live where we have a limited amount of time to get great, great results.

So I wanted to talk about that, because I didn’t want to have this — we’re not saying to go in and tune every single note on everything perfectly. I don’t believe in that. My favorite singers…

Stefan: It would be the biggest mistake you could do with any tool is to go and put the one tool on everything. That’s really — would be a very bad idea. It’s actually the opposite. You should listen to the performance you have, listen to all of the nuances, and Melodyne is helping you a lot in analyzing the music, and analyzing the performance, and then you find bits and pieces here and there, which you might want to change, not only because something was actually wrong in the terms of a mistake. I mean, if you do a mistake, you just sing it again. You do a lot of takes, you comp it, and then you get a really nice comped take with no mistakes, but still, there’s something in the expression here or there, and the timing, and nuances of the intonation, and that’s where Melodyne comes in great. Not to put 100% tuning on everything. That would be really a mistake.

Warren: Yes, absolutely. As we were talking about earlier off camera, taste is the biggest thing as a producer/engineer, or anyone working in music in general, taste is your biggest friend. Having good taste and knowing when to do something.

Stefan: And you know what, I personally think where Melodyne helps you is in training your ear first, like, looking on a performance in Melodyne, understanding more about what’s going on there, and then you can make your decision which things or which aspects you want to change, and which parts you want to leave untouched, and all that. This is actually the basic idea of Melodyne. It’s not about correcting mistakes. It starts earlier. It’s about looking into music.

Warren: Right.

Stefan: Like what we have here for example on the screen, this is a bunch of notes. It looks like tiny small waveforms. I can play it back, it’s a guitar. Normally, in a DAW, you’d see one long waveform, which would actually give you not too much information. You can see, “Oh, it’s loud there, and it’s louder there,” and that’s about all.

You don’t see which note is played, or which chord is played, and what Melodyne does, is it analyzes the music when you load it into it, and then it just translates it into notes like a MIDI editor, actually. It’s a bit like the note editor here, you have the notes, and you can see, for example, this note here, he picked the E, and so you can even see the song is in F Major, so that’s what Melodyne finds out automatically, and then you look at it, and it gives you the freedom to change things which you couldn’t do if you just had one big waveform only.

For example, if you — just let me play this back without the cycle.

[acoustic guitar]

Let me bring in some more instruments probably. Like let’s listen to the vocal also. The bass in here. You see, I just added a few instruments, they all show up there if I want to, and one of the first things I would do, like this guitar picking, it’s really nice, we recorded it to start with the song, I just have to turn off the bass again and the piano.

So the red ones are the picked guitar, and the other one is the vocal. I can just switch the roads, so you can see now, this is the vocal, and there’s a specific part where the picking, if you listen to it on its own like this, it’s a nice variation or so, but it doesn’t work with the vocals that we recorded later, so what we then simply do, we just — this note for example could go there. This note probably could go there.

[guitar]

So I just took out the variation he actually played, because I don’t want it to interfere with the singer if you listen like this.

[guitar and vocals]

So what I just did, first we did the recording of the guitar, then we added the vocal, and then later, we decided, “Oh, it would have been nice if the guitar picking was in this example had less variations.” So I just change it.

Something I did here, which we used to do, since the case on MIDI, but with Melodyne, you can do it on audio, and you can also do the opposite, if he played the same pattern all the time, and you kind of get bored a little bit, when you add the other instruments, you see, “Oh, maybe a nice variation would be nice,” just drop a note, and there’s your variation there.

So that’s one thing you change the composition, and you don’t change mistakes. I’m not talking about mistakes or intonation issues, I’m just…

Warren: Using it as a creative tool.

Stefan: Yeah, well…

Warren: Yeah, no, it’s a creative tool.

Stefan: Creative is a big word for what I just did. I moved a few notes around. That’s not…

Warren: Yeah, but from a producer’s point of view, a lot of what I do is getting to the point, and you know, I’ve worked with artists, and myself even, I play something, and I’ll embellish it. Obviously, that’s a subtle embellishment, but sometimes, I might embellish it so much that my ear is taken away by the guitar part, when I should be listening to the vocal.

And as a producer, you’re sitting there working with an artist with a guitar player, and going, “Okay, that’s a really nice version, but my ear went there, and I didn’t hear what the singer’s saying.” So you simplify, and then you go, “Ah, but look in this two bar sequence here, or this two beat sequence here,” it goes, [sings notes]. There’s no vocal, just put that little flourish in that section between. So it is a creative tool.

Stefan: It’s beautiful that you can do that after you’ve done the original recording. And that’s pretty unique, you couldn’t do this any other way. Then of course, you use it for kind of correction everything. So at the beginning of this part, this note, and I think the lower string down here, he played the right notes, or he played the notes he wanted to play, but they don’t sound — just listen to it.

[guitar]

It’s a little… He just didn’t hit it right. So what I would do then, I would listen to it, and see if I could find some better sounding versions of these notes, like here, so this note, and this note, they sound better, so I just copy them, I just now hit Command+C, go back to the two notes I didn’t like, and just put paste. So I copied the better sounding notes over there, and I’m done.

[guitar and vocals]

I just repaired the sound. You could do this in your DAW, you do it all the time. You copy and paste, but what I copied here is just one note, or in this example, two notes, out of many notes going into each other. So you can’t do this actually on many spots with products like this, you can’t do it on your DAW.

Warren: Well no, because you’re taking a chunk of three notes combined together, and you’re going to approximate where the ending of all those different notes is here, where here, you’re taking those two notes and replacing them with two completely different notes.

Stefan: Or even to go to more creative, you could for example select this note, if you particularly like it, hit copy, and paste it in here, and there it is. So you just added another finger on your guitar. Something you couldn’t do without Melodyne.

Warren: I could do that, I just need to do my…

Stefan: Actually, if you recorded into Melodyne, I could change just one finger on the fret. I can make a lower third — the minor third to a major third, or the other way around. So that’s where the beauty comes in terms of composing. You really can change the chords, and if you say, work with samples, which is a pretty different thing, but if we were recording guitar here in the studio, or we’re sitting here, and we don’t like the recorded way, then just play a different chord until you like it. But you find a sample which you like, and this sample is an F minor, and we want to use it in our song…

Warren: You’re making me think of a really interesting question. So when you’re taking a chord like an F major, to sound perfect on a guitar, which is tuned in fourths, and the A has to sound very slightly flat, that gives us a really beautiful sounding F major chord. So, my question would be, where I’m going, can you go in there and detune that A ever so lightly. Is there an incremental way of doing it?

Stefan: Sure. What you can do is you can tune every note. So I’ll select this one note, and then I can move it up and down as you already saw, you can do it really subtle, and you can also do the left here, you can see the parameters here. This was actually tuned too high, so I can just tune it down slightly. You can see how the note moves, I’m just moving it over to show it, and actually, I would — the first thing I would’ve done here to correct this slightly tuning issues that there are in the guitar, like this one was a little bit high, so I would just do it like this. And if I wanted to have equal tuning, I would do it slightly flat.

On the other hand, if you don’t know exactly where it should go, because equal tuning is kind of a speciality, there’s a much easier way, just select the notes that make up the F, just disregard the D that’s playing in the bass, or select them all together, it will give you a different result, whatever you’re after, then we have a command here, it tells you to apply dynamic just tuning, so what it does now, it tunes all those notes I selected so they fall into a just tune. A just tune. Equal tuning, sorry.

ADVERTISEMENT

[acoustic]

Directly starts to sound as it should be.

Warren: Yeah, although the A starts to scream out to me now, so let’s take that A and detune it ever so slightly.

Stefan: In which direction you want to have it?

Warren: Flatten it.

Stefan: Flatten it, so we’d just move it slightly down. I’m just double-clicking.

Warren: Yeah, let’s have a listen to that.

[acoustic]

It’s interesting. So now we’re in a territory which is great, because now the chord soloed will probably sound really, really sweet, but it no longer works against the vocal. This is where taste comes in. It’s not a case of just tuning things. I find a good analogy is like drums. You record drums, and bass, and guitars live on the floor, you go back, you edit the drums. Suddenly, the bass player and the guitar player have no relationship with the drums anymore. So what do you do?

You edit within itself. You go, “Oh, I like this fill,” you put that fill in. You don’t re-edit the drums super tight, because you’re going to be recutting the bass and guitars. So from a tuning point of view there, it’s like, yeah, now that chord is probably soloed beautiful. That F Major probably sounds amazing with that slightly flattened A, but it has no relationship now with the vocal or any other instrument, because they’re all in that tuning around each other. Great musicians are able to play off each other.

She’s singing to the guitar, he is playing to a piano, which may or may not be perfectly in tune.

Stefan: So the thing you probably want to do if you recorded an instrument that wasn’t tuned pretty well — not a good idea to do all that often, because a good singer will work with this, and then he or she will adapt to it, then later, you tune the guitar, and the vocal gets…

Warren: Exactly.

Stefan: So if you record an instrument that wasn’t tuned perfectly, you should correct it, and then do your overdubs, but if it happens and you know how it is, and maybe you’re working with samples that are kind of strangely tuned, and people give it to you for mixing, and you realize the tuning wasn’t good, and the people that sent it to you didn’t realize, so you can do it later, and what I just did on this guitar, I would actually do for all the tracks simultaneously on this part, which is actually pretty easy, because the bass for this part, I would put on the screen along with the guitar. The vocal also. Then also the piano. Then I would select all notes that are playing in this part, or in the whole everything, and then I would apply the dynamic tuning for all of them.

So now, they are tuned in respect to each other, which is supposed to sound better.

[mix]

Stefan: It’s actually already sounding better, but I would then, in this example, the vocal has a kind of insecure attention to it in this specific note here. See, this is a — I just zoomed in here, you can see…

[mix]

So this “beyond” is actually two notes, she’s singing F and A, but she’s not a pitching fork or what you say, it’s like, [sings notes]. She’s going there, this is the expression, which is the feeling I may like or may not, and this is when one thing now, which is very important to understand, you see the two notes, and to the left, you can read the pitch, this is F, this is A, but watch this line in front of it. This curve moving up and down. This is the actual pitch, and you can see how she’s going from up to there, and during the note of this — I think that’s the second part of the word, she’s going too high.

So the note in general, the whole note is measured by Melodyne like all the movements from here until here, they have an average of so and so much hertz, and this would be in respect, the master tuning of the song, and this will result in an A, which in this example, is still a little bit flat, we just moved it there, because we wanted to have this dynamic tuning, so just remove this example, and it should be a little bit flat. So it’s probably, it’s the average of this curve is okay, but you have to listen to it, and then you’ll realize she’s kind of getting too high for a short moment.

[vocals]

It sounds insecure. What you do there, you have different strategies now.

Warren: I like that, it sounds insecure. I like that.

Stefan: Yeah, that’s — we are actually — we’re talking not about mistakes here. It’s like the attitude. Is it secure, is it soulful, is it strict, is it cold, or whatever. When you work with Melodyne, you start to care for those kind of attributes in a performance. It’s not only about frequencies.

Warren: Oh no, trust me, I spend — I record and comp vocals all day before I get anywhere near any potential tuning. That’s the last thing I do in the process. I will — I like to loop sections, and move through playlists, and build a comp, and then the last possible thing I do is look for things to tune.

Stefan: Well, I will show you comping later, and this may give you some new insight on comping, which we just released on this NAMM, or shortly before NAMM, this 4.1 new comping stuff, which is really interesting.

But let’s get back to this part. Actually, probably, I would have a better take there, and I would comp some other thing. But say I don’t have a better take, or that’s something somebody sent to me, maybe I didn’t record it myself, so I have to live with it, so what would I do?

I have different strategies. You see I have different tools for different things, and you see I have one tool for changing the pitch modulation. So the pitch modulation would be like the vibrato for example. You could have a strong vibrato, your curve will go up and down, and then you can say, “Well, that’s too much of it.”

So you just turn it down a little bit. So you don’t draw in a new vibrato, technically, you just take what’s there, and just lower it a little bit, or you can over exaggerate it with this tool, I do here now, my problem here now is she was going up there too much.

I’m just only doing it. I now make it a straight line. This might now sound artificial.

[vocals and guitar]

I don’t like it. Personally, it’s too clean. I would probably choose a setting like this or so, or I have a different strategy, just let me redo this to the original state, I would probably still cut this note, so I have a different note for cutting things, and watch when I cut the note changes, but it doesn’t change acoustically. That’s an important thing to understand. It’s really a basic first thing when you’re about tuning, if you do additional separations, that’s what we call this cutting up, remember I told you Melodyne is recalculating the pitch on the basis of the average of the whole movement.

So now, I have a new note with a different movement, so the average pitch gets reinterpreted, and now this part, which I just shove up, is measured as 33Hz too high. That’s how you can see how much too high this was, and so I bring this back to — if you don’t like this annoying kind of synthesize or pitch thing like this, I’ll turn this off. Now watch this thing to the left where it says it’s -13. So it should probably go — normally you would say to 0 cent offset, but in our case, because we wanted to work with just tuning, that would actually be around -7. So that’s actually where this note should go, and then I would turn this up there a little bit.

Maybe this I would leave, because I like how she’s going back there. So let’s listen to it.

[vocals and guitar]

Still this is pretty much too high. And maybe I don’t like the time she uses, so then I could with a different tool make this become steeper. Or I would go in, then now it’s getting a little bit sophisticated, I’d use the timing tool, then only this part, I’d make become a little bit faster.

[vocals and guitar]

Like the, “beyond,” she’s getting more expression to this, because I just changed the timing. So actually, with these tools you do things you normally would tell the singer, “Do it more precisely,” or, “Do it more laid back, do it softer,” do it whatever. So we have tools for it.

But to use these tools properly, you have to learn how to read this curve thing. That’s what I was wanting to show him with this example. Reading this curve properly and translating what you hear into the equivalent of what you see in this curve, so this is the key of using it properly. If you watch people use Melodyne on a very professional basis, it’s amazing to see how many things they cut up, just to work quick and easy, and we take this into account, because we have — just let me select these two notes, and we have a command here which separates automatically this stuff where the movement is going, and then you could with this slider here, you could correct these slices like this. Make it come really steady thing, and just do it in the cycle again, just to here. So this was the original.

[vocals and guitar]

And now I… So again, this is now — I didn’t choose 100% now. That’s a matter of taste, again, how much control you want to have.

So this was a quick run through through the basic tools, the basic understanding, read this pitch curve, decide where you want to edit. In my example, I did a lot of things, because I now edit all parts of this note. That’s probably because this take wasn’t so good. I might have had a better take, then I wouldn’t have needed this kind of…

Warren: No, I understand totally, but we’re sort of assuming this is all we have to work with, so this is either our best comp, or like you said, somebody has sent you a vocal. I think honestly, as a mixer these days, I mean, I’m still mainly a producer, engineer, mixer in that order, but as a mixer these days, I find that I’m having to — people are expecting it to come back better. That’s the point of hiring a mixer, so a lot of mixers like myself are forced into either comping a vocal sometimes. I’d rather not, because that’s hours of work, but at the very least, we’re getting in there and doing pitch correction when necessary, even sometimes adding harmonies and stuff like that, if they want to take it over the edge.

Because a lot of people recording don’t have necessarily this knowledge. They know how to get amazing ideas down, and they’re looking for you to take it over the finish line. And maybe that’s a bit more than the job description as a mixer to do this stuff, but ultimately, if you’re getting hired as a mixer, and you want to give somebody an amazing result, then get stuck in and do these subtle things.

Maybe it’s just twice in a song, but it’s the difference between somebody listening to it and going, “Wow, he’s improved my song.”

Stefan: That’s actually really truly — improvement often lies in just this one spot here, this one spot there. That’s — some people may over hear it, and they can — they would say, “Well, that’s a good vocal.” It’s a very good vocal, it’s nice, but you and the other skills of a good Melodyne editor, you would say, “It’s a good vocal, but I could make an even better vocal, maybe possibly even a perfect vocal if I go in for this little thing there, and a little thing there.”

I’ve been talking to many producers, and I ask them, because we want to learn from our power users, I’ll actually use it, and I’ll ask them questions like, “Okay, how many takes do you need for your comping, when do you use Melodyne, how many spots do you edit with Melodyne,” and the pro guys, they always tell me, “I have such good singers, I’m lucky, I’m blessed to be working with great singers, they have great material, and they only go in there for one or two, three, four spots, but that’s the final thing that really makes the difference, man.”

Then you’re actually pretty quick once you have the skills to identify the spots where you need to go to, and coming back to the first thing we started when we talked about, “Don’t put everything into it.” That would be crazy. Just listen to it, find the spots, do subtle things.

Warren: Thanks ever so much, that’s great. The basic tools in here, the next video, we’ll get into talking — which will be coming up soon, we’ll be talking about doing a double. Thank you, I appreciate it.

Stefan: You’re welcome, it’s fun.

Warren: Have a marvelous time recording and mixing. Please leave a bunch of questions and comments below. Let’s talk about your comping process, how you comp, and when you choose to use tuning. You know, do you tune at the very end, do you move — sometimes I find I’m comping and tuning at the same time, if I’m moving really, really fast, but when I have the time, and I have plenty of takes, I build something as beautifully as I can, then just cherry pick one or two notes. That’s when I have a lot of time, and a lot of takes with a singer and stuff. But, as we were talking about, a lot of times I’m mixing, and I’ve just got one take, so I’m just doing the minimal amount of work.

But thanks ever so much for sharing that. Have a marvelous time recording and mixing, leave a bunch of questions and comments below.

Expand

Warren Huart

Warren Huart

Warren Huart is an English record producer/musician/composer and recording engineer based in Los Angeles, California. Learn more at producelikeapro.com.

Free Video on Mixing Low End

Download a FREE 40-minute tutorial from Matthew Weiss on mixing low end.

Powered by ConvertKit
/> /> /> /> /> /> /> /> /> />