Ask the Engineer: Should I Mix as I Go or All at Once?

Masterdisk Ask the Engineer graphic

Today’s “Ask the Engineer” question goes to mastering engineer Randy Merrill.

Randy Merrill joined Scott Hull Mastering in early 2006 as Scott’s production engineer. Shortly thereafter he started building his own mastering clientele, and today he’s a staff engineer at Masterdisk. Randy’s approach is to be as attuned to his clients’ aesthetic and practical goals as possible. He goes the extra mile to make sure the finished product reflects how you want your music to sound. Randy’s credits include Bruce Hornsby, Bill McHenry, Tom Wopat, 3 Cohens, Perez Hilton, Darcy James Argue, Paul Jacobs (Naxos) and Chantal Claret.

Q: I’m recording my album over a span of about a year. Should I mix tracks as I go along or have them all mixed at the end?

Photo of Duduka Da Fonseca and Randy MerrillA: It’s best when a project is mixed in a somewhat short time span. Things like relative level between instruments tends to suffer when songs are mixed apart from each other. Sometimes the vocal can be set “in” the mix on one song, and “on top” in another song, depending on what the engineer feels that day. Other considerations include how the drums sit in the mix, and how the bass sits. If there’s a lot of variation from track to track it can cause an album to feel disjointed.

Q: How much can mastering do to “tie” the different mixes together?

A: It can do a fair amount, but not as much as can be done in the mixing. Relative vocal levels can be approximated somewhat; same for bass. If one vocal is really “in” the track and another is upfront, it’s tough to get the two to sit similarly. Likewise with the drums: if they’re in a different place from song to song, it’s hard to get them to match.

If your project can’t be mixed in a short span for whatever reason, the previous mixes should be reviewed while the new ones are being done. This will help with the overall consistency of the album.

Photo: Randy Merrill (right) with Duduka Da Fonseca.

Ask the Engineer: Tim Boyce on Your Unusual Music

Masterdisk Ask the Engineer graphic

Tim Boyce’s specialties lie in world, dance, dub, hip-hop, ethnic, remixed, and other genres, and his production credits include French Montana, SLDGHMR, Deathrow Tull, MSTRKRFT, Nightbox, and RAC. Tim’s creative desire to breath life into a static space allows him to finely tune an album into a living movement, not just a passive listening experience. If you need cutting-edge mastering, Tim’s your engineer.

Scan from COLLECTION OF DANCES IN CHOREOGRAPHY NOTATION (1700) at the Public Domain ReviewQ: My music is pretty unusual. What kinds of things should I talk to my mastering engineer about before mastering?

A: With so many styles of music, and hybrid/fusions happening in both the live and production music scenes, it’s sometimes difficult for a mastering engineer to guess what the artist has in mind. Sometimes it’s obvious what path to take. For example, a ballad or orchestral work has a very different mastering approach than an aggressive club banger. But what about a folk song with traditional instruments (banjo, fiddle, acoustic guitar), synths, and a hip hop rhythm on upright bass? (It happens: I heard it last week, and it was awesome!)

If you’re making new cutting-edge music, or re-defining your sound by trying something new, it’s often best to let your engineer know exactly what you have in mind. Let them know you really want the bass larger than life, even though it might not be the most dominant element of the arrangement. Or that we are experimenting with filters on the banjo to make it sound filthy and really cut through the mix like a dance synth. Your mastering engineer won’t know unless you tell them.

Personally, I love working on music that pushes the boundaries. You should always be able to feel free to reach out directly to your engineer. We’re not mind readers, but we’re all very nice and we want you to be thrilled with how your music sounds. So reach out, and lets talk.

Fill out my online form.

Ask the Engineer with Scott Hull: What Happened To My Vocal?


Today we’ve got a question from a reader about how her track was affected by mastering (at another facility). Scott Hull answers.

Q: I had a song mastered and the vocals seemed to sound harsher, with a loss of ambience. Can this occur from the mastering? — Sherri

A: Sherri, thanks for your email.

The effects of mastering can be very profound, both positively and sometimes negatively. An “average” engineer might have been taught or learned to brighten the midrange and high-end even if the music doesn’t need it. As a veteran of thousands of mastering sessions, I can tell you that the hardest thing to learn was when not to “master”.

I can help you get the sound you are looking for. I will give you a free song / mix evaluation ($99 value) — for free! Just for sending us your great question.

You don’t have to compromise. The right engineer can make a world of difference.

All the best,
Scott Hull

Fill out my online form.

Ask the Engineer: Randy Merrill on “How Loud Is Too Loud?”


Today’s “Ask the Engineer” question goes to mastering engineer Randy Merrill.

Randy Merrill joined Scott Hull Mastering in early 2006 as Scott’s production engineer. Shortly thereafter he started building his own mastering clientele, and today he’s a staff engineer at Masterdisk. Randy’s approach is to be as attuned to his clients’ aesthetic and practical goals as possible. He goes the extra mile to make sure the finished product reflects how you want your music to sound. Randy’s credits include Bruce Hornsby, Bill McHenry, Tom Wopat, 3 Cohens, Perez Hilton, Darcy James Argue, Paul Jacobs (Naxos) and Chantal Claret.

Photo of Randy Merrill and Tom Wopat
Tom Wopat and Randy Merrill
Q: How loud is too loud?

A: We’ve been hearing about the “loudness wars” for a long time now, but one point people don’t make is that “loud” can be done badly, and it can be done well. I’m not somebody who’s going to say that everything today is too loud. There are some great sounding albums that are loud, dense, exciting, and punchy.

The first thing to keep in mind, when you’re talking about loudness and mastering, is that the client ultimately sets the target. When I’m mastering, I’m serving the client’s vision for their project. That said, just because a client wants a loud album doesn’t mean it can’t still sound great. It can.

The first thing you need if you’re going to make something loud AND sound good is the skill to do so. That, plus effort, experience, and the ability to scrutinize your work in a finely tuned environment.

There are entirely different ways to get things loud. There are good ways and not so good ways, and are often dependent on the qualities of the actual mixes. Every approach has its caveats. I may do it differently from project to project depending on a lot of factors, like how the mixes come in; what the music is like; what kind of intensity the music needs.

But to really answer your question, it’s too loud when the music is fatiguing and unpleasant to listen to. In this case, it’s either too loud or it was done badly.

If a client wants a very uncompressed, dynamic recording then that’s the direction we go, of course. I have clients in classical, jazz, and even singer-songwriter stuff where that’s the right approach. But if my client wants their project louder, then its a matter of finding a balance. On different resolution playback systems loudness can sound either good or bad. If it sounds good on an iPod dock it doesn’t mean it’s going to sound good on an expensive hi-fi system. And the opposite can be the case too: it may sound great on the hi-fi but it doesn’t sound as impressive or alive through the dock. So you have to find a balance; the music has to sound good however it’s played.

So even if you want a loud record, you can still have a great sounding record. You and your mastering engineer will find the right balance.

Ask the Engineer with Matt Agoglia: The Art of Sequencing

Matt Agoglia is a mastering engineer with a deep love for the album format. Talk to him about one of the recent albums he’s mastered — he always seems to have one that he’s particularly excited about — and you’ll pick up on just how keen his attention is to the nuances of the long-player. Matt is passionate about the album as an art form, and it shows in his work: listen to his work with legendary singer-songwriter Emmylou Harris, or the pop-punk group Sleeper Agent, or avant-pop composer Mikel Rouse and hear for yourself.

Since this is the first “Ask the Engineer” that Matt’s doing, I thought we’d have to start with that very ephemeral but all-important aspect of the album art form: sequencing.

photo of Ventures vinylFirst of all, when we’re talking about sequencing, we’re not really talking about song order, right?

That’s right; I don’t really deal with song order. That’s usually dictated. Often the spacing isn’t though.

Do you ever get asked to contribute to the song order?

Clients sometimes ask for my input. The important thing is to have a storyline, a thread. To have highs and lows. With the song order you play around with those feelings. Make a journey out of it because you know, some people still listen to records all the way through!

What’s important about the gaps between the songs?

In the mastering stage you’re taking your mixes and turning them into a listening experience. How each song sits and breathes can enhance a listener’s enjoyment of the record. And that’s what mastering is about. Enhancement.

So how do you do that?

A couple different ways. You get projects where the artist has a specific idea about how each song should go into the other. Crossfades and things like that. Sometimes it’s very elaborate and they’ll send me an mp3 mockup and I can recreate it.

Other times, if we’re not doing elaborate crossfades, I ask myself how much time my ear needs to settle before the next song comes in. A lot of people like these really fast-paced records and I don’t think that’s ALWAYS appropriate. You can rob a listener of a better experience of your album by smashing it all together. It’s nice to have a mix of short and long spaces. Take the listener on a journey, and make it a pleasant one. Maybe it’s quick in the beginning then maybe you need an intermission. It depends on the music, of course.

So it’s by feel?

Yes. And you can really enhance a listener’s perspective on the songs. Music still has energy after the last note has died off. Sometimes the emotional response needs to subside, you need to have that time between songs. That’s what sequencing can add to a project.

Is there any particular technique involved?

A lot of times when people are sequencing, they listen only to the last few seconds of a song before trying to determine the gap before the next one. It’s really not ideal; you want to feel the energy of the full song in order to determine how much of a rest you’ll need before moving on. So I’d say that ideally, you’re going to get the best sequencing by listening through the whole album while you’re making your decisions. You’ll spend more time doing it this way, but it’s worth it.

Ask the Engineer with Scott Hull: How Much Music Fits on an LP Side PART 2

Last time we cut mono test tones and found out how big a difference a little bit of level can make on the duration of an LP side.

So let’s now look at a typical rock program. For our example this music is compressed — something like a classic Rolling Stones record — but not peak limited like a typical radio rock hit of today.

The grooves for this cut wiggle back and forth and up and down. That is how a stereo disk works. The more stereo the mix is (for example, guitars and drums panned to the sides) the deeper the groove is. A deeper cut is also wider — check your geometry lessons from middle school. 

image of vinyl grooves, magnifiedDoes anyone know what the word “analog” means? The signal and the groove is an analog of the original audio, i.e. the groove and signal are analogous. (I should not try to use such big words.) But louder alone does not determine how much space the grooves take up on the disk. The character of the program, how loud, how soft, how much bass and how much stereo all contribute to the picture. Bass has the biggest overall impact on duration.  

So when measuring the grooves to see if they will fit on a side, the cutting engineer has to consider the bass, the peak level, the average level and the duration of the music. Every change you make to the music is a compromise, so deciding how much bass, how much level and how much compression has to be decided by ear and with experience.  

Why do records that come from one studio sound better? One simple reason is the engineer, and how much they care.  

John McLaughlin Now Here ThisThe sad truth is, a typical engineer hears the music and says, “Oh it’s kind of bright, lets roll of the top; they probably won’t notice… and wow, they put a lot of bass in the mix too so we had better roll that off too! And wait a second, the floor tom is panned all the way to the side and that means we have to put in a low frequency EQ called an Elliptical to partially mono the bass.” You can certainly expect this record to sound weak, limp and dull. I don’t even like my breakfast cereal that way.  

What if that same music could be cut and fit, in full frequency range, with the bass intact and the floor tom where it belongs — but just lower the level 1db. Only a patient, determined and experienced disk cutting engineer, who is compensated for his or her time, will push that cut, take the right chances and make an amazing record. But what about the compromises?  

I cut a recent John McLaughlin album, “Now Here This,” for the Abstract Logix label. It would have sounded awful if the bottom end had been rolled off. It was in-your-face with bass and that was how the artist wanted it. So to fit the music on the side, the level had to be reduced. Not a lot, just a db – but in this case, with a quiet pressing, there was no creative damage done to the music.  This is the caring part. 

I recently cut a very demanding Glen Frey record, “After Hours,” with my young ace cutter Alex DeTurk assisting. We cut many refs and compared the playback to the tape master — that’s right I said “analog tape master.” It IS very cool to cut an analog record from an analog tape through an analog console! The original Elliot Scheiner-mixed 2 track analog was beautiful. I was tasked with making the record sound exactly the same as the analog tape. They wanted full range, no filters, almost no de-essing, and NO digital or analog processing of any kind. Well, let me say it wasn’t easy. But I’m super proud that Michael Fremmer’s review claimed it to be a 10 of 10 for sound. (link )  It only took experience, determination, and patience. Just like anything worth doing well.

Fill out my online form.

Ask the Engineer with Scott Hull: How Much Music Fits on an LP Side?


Ask the Engineer is a series on the Masterdisk blog where our engineers answer questions about music production.

SCOTT HULL is a 28-year veteran mastering engineer and the owner of Masterdisk studios in NYC. Scott started his career in 1983 and has mastered hit records and classic albums in every genre, as well as many Grammy-winning titles. He is widely regarded as an expert in vinyl mastering. Recent projects include albums for Donald Fagen, Sting, Dave Matthews, Glenn Frey and KISS, as well as multiple albums for the Sh-K-Boom, Tzadik and Luaka Bop labels. Over the course of his career Scott has mastered albums for Bob Dylan, Bruce Springsteen, Miles Davis, Lou Reed, Ravi Shankar, Herbie Hancock, Tom Zé, John Mayer Joe Bonamassa, and many more.

Photo of the Masterdisk latheQ: How many minutes of music will fit on the side of an LP vinyl disk?

A: It’s a simple question with a complex answer. Many websites publish charts explaining how much music fits on one side of a vinyl record. The main purpose of those guidelines is to make it easy for the cutting engineer to do his job. But do you want to have an average record or an extraordinary one? Ah, I thought so. You need to read on.

Lets just say, for argument’s sake, that we wanted to cut a vinyl side with a 1k test tone (midrange near a middle B on the piano). Pretty boring “music,” but this control measure will help me explain the process. And lets say that that tone an be cut on a particular lathe at a level of 0db and at a duration of 30 minutes. The relationship between level and duration is due to the fact that a louder signal cut into the disk takes up more room on the disk and thus the grooves have to be farther apart to avoid cutting over themselves.

Now lets take the tone generator and lower the frequency to 500hz (down one octave). Cutting this signal at the same level as the 1k tone, we will run out of disk near 24 minutes. The bass frequencies have longer wavelengths and use more space as they squiggle back and forth.

Lower it another octave to 250hz and we run out of disk at 18 minutes. Surprised? So how can we possibly cut rock and roll, with energy down to 20hz, for more than 20 minutes? There’s more to the story.

Let’s go back to 1k. Remember, it fit on the LP side for 30 minutes. If I lower the level 1 db, we can now record 33 minutes of tone on the disk. Wow, only 1 db? The reason is that it’s 1 db throughout the entire side: the average level is down all the way across the disk. This is very important.

Then let’s raise the level to +2 db from the first test. What do you expect to happen? We run out of disk at 25 minutes. That’s 5 minutes less audio recording space with just a 2 db raise in level. So level is king, bass is queen and hi-frequencies are the jack, ten and nine. Remember we are still talking about simple test tones, not music.

The point I’m trying to make is that music doesn’t obey rules of thumb. No two projects are the same. Even if the music was identical, two different producers might have different objectives. One might want the record loud, another may be more concerned with being very high quality / low distortion and might not mind a slightly lower level.

Before you decide if your music “fits” on a side please talk to your cutting engineer. The engineer has to listen to your music, and measure how his or her lathe will respond to your music. Anything will fit if you turn the level down far enough. Don’t just send your cd master to the vinly pressing plant asking for an “average” cut. Your music doesn’t have to sound average on vinyl – it should sound amazing! And you already know who to contact to make that happen. (That’s me!)

I’ll go a little deeper into the grooves next time when I talk about what happens when we aren’t cutting mono test tones. I’ll give you a hint… the grooves get deeper and that causes them to take up more room on the disk. Uh oh…

Fill out my online form.

Ask the Engineer with Scott Hull: Can We Fix it in the Mastering?


Ask the Engineer is a series on the Masterdisk blog where our engineers answer questions about music production.

SCOTT HULL is a 28-year veteran mastering engineer and the owner of Masterdisk studios in NYC. Scott started his career in 1983 and has mastered hit records and classic albums in every genre, as well as many Grammy-winning titles. He is widely regarded as an expert in vinyl mastering. Recent projects include albums for Donald Fagen, Sting, Dave Matthews, Glenn Frey and KISS, as well as multiple albums for the Sh-K-Boom, Tzadik and Luaka Bop labels. Over the course of his career Scott has mastered albums for Bob Dylan, Bruce Springsteen, Miles Davis, Lou Reed, Ravi Shankar, Herbie Hancock, Tom Zé, John Mayer Joe Bonamassa, and many more.

Photo of Scott HullQ: Can you bring out the vocal in mastering? Can you push the snare drum back in the mix? Can you soften the cymbals? (And other “Fix it in the mastering” questions.)

A: There is a lot of subtle tone change that can be applied in mastering, but there is always a “but.” I’ll start with the least successful stuff first. If your mix is harsh, too much cymbals or too much high frequency “zing” on the vocal, it is going to be hard to make that sound great in mastering.

Don’t get me wrong — there are solutions and there are ways to soften the upper-mids and top, but often these fixes cause more damage to the rest of the mix. That is why I always suggest that your final mix SHOULD be just a little dark and a little warm. This means different things to different people and in some genres it’s desirable to be dark and warm and in others it is not. But the main take-away point is that if you “aim” your mix to sound exactly like a mastered CD, I may not be able to make it sing. When I use mastering EQ to soften the high frequencies, ALL of the elements of the mix get darker, not just the one(s) that are harsh. So if the hi-hat is ear splitting, I can roll off 7k and up, but even in very small amounts that will make the vocals sound less impressive and the guitar crunch will change and so on.

Tonal balance is the key to this equation. If ALL of the elements are too bright, then the EQ will work. If it’s just some elements then the EQ only serves to un-mix your music. Many people reach for multi-band compression in this case to try to take the sting out of the offending elements. This ONLY works when the offending element is the loudest thing in that part of the frequency spectrum. The effect of the hi frequency band-specific compression will be to soften the loudest, most transient part.

De-essing is a special case and is surprisingly effective. The de-esser has been designed to identify vocal esses and not confuse them with the brightness of the cymbals or snare. Because of the extremely bright tone, trumpets – especially muted trumpet – will get eaten alive by most de-ssers, so great care has to be used around brass and vocals.

Many times, the solution for an overly bright mix is not to cut the harshness, but to find ways of making the bass, warmth and punch regions stand out more. Balance is the key.

Fill out my online form.

Ask the Engineer: Scott Hull on Mixing: Master Buss EQ and Peak Limiting, Part 2

Ask the Engineer graphic

Ask the Engineer is a series on the Masterdisk blog where our engineers answer questions about music production.

SCOTT HULL is a 28-year veteran mastering engineer and the owner of Masterdisk studios in NYC. Scott started his career in 1983 and has mastered hit records and classic albums in every genre, as well as many Grammy-winning titles. He is widely regarded as an expert in vinyl mastering. Recent projects include albums for Donald Fagen, Sting, Dave Matthews, Glenn Frey and KISS, as well as multiple albums for the Sh-K-Boom, Tzadik and Luaka Bop labels. Over the course of his career Scott has mastered albums for Bob Dylan, Bruce Springsteen, Miles Davis, Lou Reed, Ravi Shankar, Herbie Hancock, Tom Zé, John Mayer Joe Bonamassa, and many more.

Photo of Scott HullQ: When I send my mixes to my mastering engineer should I remove the buss limiting and EQ? (Read Part 1 of Scott’s answer here.)

A: Here are “Scott Hull’s Guidelines for the Use of Buss Processing.”

1. If you have mixed through the buss processing, i.e. had the compressor ON while you were mixing, then you most likely should leave the buss compressor or EQ ON when you send the mix to the mastering engineer.

2. If you find that when you take the buss compressor out the mix “falls apart” or loses it’s soul, then you should probably leave the compressor or effects ON when you create the mix for mastering.

3. If however, you didn’t mix thru the compressor – but added it after the mix to bump up the level for references — in that case you really must take the buss effects OFF the mix when you send it to mastering.

4. If someone else did your mixing and you are not sure if they mixed “through” the buss compressor or not, ask the engineer to print the mix with AND without the buss effects. Please remember too that your final 24 bit mix Does NOT have to be limited to peak at zero.

One other detail. Many times my clients come to me with a mix that has very obvious “flat topped” peak limiting. We call them “bricks” or “Tootsie Rolls” because of the way the waveform looks. At this point I’ll ask for a mix without the limiting, and they respond that they hadn’t put a limiter on the mix. And they hadn’t — but there WAS limiting, it just wasn’t done by a plugin called a limiter. Anytime you raise the level, or combine two or more signals together, or process the signal in a plug in, you run the risk of peak limiting within that component — and in many cases the software doesn’t flag the overshoot. You have to consistently be aware of your internal gain structure through your workstation and keep an ear and eye out for hidden limiting.

It’s like this: if every peak peaks at exactly the same level, then limiting happened.

One more final note: Don’t confuse track compression or limiting with buss compression or limiting. Individual tack by track gain control is not only a good thing — it’s absolutely essential. Oftentimes, when I find someone has overused a master buss compressor they did so because they had not applied enough gain control (compression or limiting) to the individual tracks.

Fill out my online form.

Ask the Engineer: Scott Hull on Mixing: Master Buss EQ and Peak Limiting, Part 1

Ask the Engineer graphic

Ask the Engineer is a series on the Masterdisk blog where our engineers answer questions about music production.

SCOTT HULL is a 28-year veteran mastering engineer and the owner of Masterdisk studios in NYC. Scott started his career in 1983 and has mastered hit records and classic albums in every genre, as well as many Grammy-winning titles. He is widely regarded as an expert in vinyl mastering. Recent projects include albums for Donald Fagen, Sting, Dave Matthews, Glenn Frey and KISS, as well as multiple albums for the Sh-K-Boom, Tzadik and Luaka Bop labels. Over the course of his career Scott has mastered albums for Bob Dylan, Bruce Springsteen, Miles Davis, Lou Reed, Ravi Shankar, Herbie Hancock, Tom Zé, John Mayer Joe Bonamassa, and many more.

Photo of Scott HullQ: When I send my mixes to my mastering engineer should I remove the buss limiting and eq?

A: Yes and no… it really depends on how you mix. You have to think about what outcome you’re going for before you can know what will produce the best results. I’ll give you some guidelines.

Mixing through a buss compressor is not necessarily a bad thing. But you have to understand that it is adjusting the levels automatically when it gets loud and in effect the buss compressor is a sort of auto-mix tool. If this is used in moderation, for a specific desired effect, that’s great. However, you have to consider that when you take the buss effects off your mix will change. Sometimes subtly, sometimes not.

Another typical use of a buss compressor or limiter is to bump up the level of the mix for the final level. Some mix engineers wrongly refer to this as mastering. This is like putting a mic on a stand and calling that recording. But that’s a separate subject! Anytime you or your mix engineer places a level bump on the buss you have to ask that a non-limited version is also created… no mater how much the mix engineer objects.

Let’s talk for a minute about the problems with mastering from a maxed-out mix.

The promise that professional mastering can help transform your mix is only possible when the mix is supplied without peak limiting. I prefer no peak limiting of any kind. The peak limiting helps you turn up the level and make the mix louder, but this process makes mastering less effective. I find that it takes more EQ to make things sparkle and it takes even more compression to make the changes needed in mastering when the mix has too much peak limiting.

Basically, limiting the mix simply limits the possibilities of your music and should only be done with great care. Often the mix environment is not revealing enough to judge the positive and negative effects of limiting. For the best sounding product, limiting – as much as is desired – MUST happen last, after EQ changes and any “color” processing like tubes and compressors are applied.

The mix engineer’s job is to supply a mix that will get approved. But the mix engineer should also explain to their clients that the “loud” setting of their project must be determined in mastering. And that mastering cannot effectively be done by the same engineer and in the same environment that the mix was done in.

Next time I’ll list my guidelines for the use of buss processing.

(Read Part Two here.)

Fill out my online form.