A Recording, Deconstructed

I take a behind-the-scenes look at a recent recording project I am working on as producer and engineer, explaining the technical details of the session, along with the pre-production planning. This episode has samples from the individual tracks, plus earlier versions and a close-to-finished version. The artist and songwriter is Corrie Green, and this is one song from a 12-song project we are working on.
I welcome your comments. You can reach me at dwfearn@dwfearn.com
If this proves to be useful to you, please let me know and I will do more like this in the future.
Also, be sure to subscribe, rate, and comment with any of the 30+ podcast providers who carry My Take on Music Recording. Thanks.

email: dwfearn@dwfearn.com
www.youtube.com/c/DWFearn
https://dwfearn.com/

Transcript

59 Recording Deconstruction                                                      29 December 2021

I’m Doug Fearn and this is My Take On Music Recording

I started this podcast in March 2020, during the Covid lockdown. I had been thinking about a podcast for several years, and as the pandemic reduced most in-person recording, I concluded that this was the time.

My plan was to do one episode a week, for as long as the pandemic lasted, or until I ran out of topics, whichever came first.

As we trended towards some degree of normalcy, I found my available time diminishing. So, I went to every other week episodes.

Even that proved challenging as I got busier, not only with recording but also with new product development. But I intend to continue – until I run out of things to say. If you have any suggestions for topics, let me know.

Here’s a story of a recent recording project.

Like many of us, I often have non-industry people telling me I should “listen to this friend of mine, who is really talented.” I listen, and find something encouraging to say.

In early 2020, a friend in Virginia sent me a link to a YouTube video by a friend of hers. I took that seriously, since I knew the musical tastes and discrimination of my friend. She a serious lover of music, but not involved in music making in any way.

I listened to one song by Corrie Green. And then I listened to it again. I must have listened to it over and over for about 20 times over the next few days. I was taken with her song, her voice, and her confidence.

I asked my friend to make the introduction, and immediately Corrie and I had a rapport and enthusiasm for doing something with her music.

It was nearly a year before it was safe for her to come here from her home in the Shenandoah Valley in western Virginia. We recorded demos of about 20 of her songs over a couple of days, and I spent some time introducing her to the art of making music in a recording studio.

Corrie learns really quickly, and grasped the concepts immediately.

We decided to seriously record some of her songs to see what we could do.

But who do we get to play on the session? Her music is difficult to categorize, but it is definitely influenced by the Appalachian style where she grew up. I always loved that music, but never before had a chance to record it.

But as we talked about it, and I listened to more of her songs, I realized that her music went far beyond that genre.

Still, we needed players that understood her style. I had to come up with the right players for the session.

I called a few friends that I thought would know the right people. One of them, Adam Monaco, is local. I’ve recorded him in the past, playing mandolin, and heard him performing live and at virtual concerts. He is a talented songwriter and has a great voice, in a Bluegrass style. When I asked him about a guitarist and upright bass player he could recommend, he volunteered himself for guitar. And he recommended his friend, Peter Oswald, who plays cello with Adam in performance. Peter, I later found out, has connections to the Shenandoah Valley.

In August of 2021, we found a date that everyone could make.

That session had acoustic guitar, cello, mostly played like an upright bass, but sometimes bowed, Corrie singing and playing banjo, and Cate, who is Adam’s wife, doing harmony vocal parts. I had worked with Cate before and she was my first choice for that role from the beginning.

The next two sessions brought in John O’Reilly Jr on drums. You may have heard the conversation I had with John in episode 53. I’ve worked with John a lot and he was perfect for this project.

Long before we were in the studio, Corrie and I recorded demos, and discussed the songs at length. We refined what we wanted, and, fortunately, we were always in agreement.

For each session, we picked two songs that we wanted to get basic tracks done, plus a third, to have ready if there was time.

One great thing about working with good players is that recording these tracks goes fast. In all three sessions, we were able to get three excellent basic tracks done, and some overdubs, while we had the players there.

Also, I thought the instrumentation should be acoustic, and sparse and tasteful, with the emphasis on the lyrics.

I knew I wanted to record this in DSD. The pristine acoustic nature of the music could best be captured in DSD. But, of course, DSD comes with lots of limitations, as discussed in Episode 52.

Once the basic tracks and a finished vocal were recorded, the session was converted to a 24-bit, 96kHz sample rate PCM, for any overdubs, and for mixing.

I wanted to use ribbon mics for everything, and, except for acoustic guitar, those choices were just right. Adam has an old, custom Gibson guitar that sounds beautiful. But it was too boomy with a ribbon mic, and not even extreme eq rolling off the bottom was giving me the sound I wanted.

So, for the guitar, I used a pair of Flea C12s, set in Blumlein stereo. That’s two bi-directional mics at right angles to each other. That’s always been my favorite stereo technique, as described in Episode 56. And the C12s did just what I wanted for the guitar.

Corrie plays banjo on most songs. She writes on banjo and performs with it. The banjo has a great sound, but it is loud! You have to amplify any string instrument to equal the volume produced by a banjo. And the sharp attack of the strings is intense.

And Corrie projects when she sings. I suppose writing and performing on banjo necessitates that.

An obvious problem was isolating her vocal and banjo from the other instruments, because I planned to separately replace both of them later. I wanted her to perform the songs the way she is used to, in order to capture the spirit of the songs. But recording vocal and banjo at the same time compromised the sound of both, plus I wanted Corrie to be able to fully concentrate on one overdub at a time.

The solution was to put her in my iso room, which is also the shop where I develop new products. It’s not a big room, but sounds decent, especially after some acoustic treatment done over the years. We weren’t going to use anything recorded in there, but I wanted it to sound as good as possible, just in case. You never know.

The guitar and cello were in the studio. I wasn’t too concerned about any bleed, since we would not be replacing those parts. The isolation was pretty good anyway.

The three songs we recorded at that first session came out great. I was convinced that I had the right people for this project.

When we got to the next sessions, with drums, I was concerned about how we were going to make it work. In previous sessions with John, about half were with him in shop, and half in the studio. In talking with John prior to the session, we felt pretty confident about having him in studio. John’s ability to play quietly yet with a big sound always amazes me. But we also had a backup plan to move his kit to the shop if necessary. To allow for that eventuality, I put Corrie and Cate in the garage. I have mic lines running there, and it’s a big space, and highly reverberant. On occasion, I have used isolated vocals recorded in the garage, but that was not likely for this project.

It worked fine with the drums in the studio. Sure, if you solo the guitar or cello mics, you can hear the drums. But that does not bother me, and, in fact, adds some “room mics” to the drum sound.

Here’s the drum track by itself, followed by the guitar track, and then the cello track.

SOLOED CLIPS OF DRUMS, GUITAR, AND CELLO

Although Cate’s harmony vocal part would be recorded later, we thought it was good to have her at the session, singing with Corrie. Cate is excellent at duplicating the style, phrasing, inflection, and mood of whoever she is singing with. I felt it was important for Cate to build a relationship with Corrie.

The lead vocal and banjo are hardly isolated in Cate’s mic, but that was OK since we would replace all those parts. In this song, however, there would be no harmony part.

I was pretty certain that I had found the perfect mic for Corrie’s vocal: an AEA R44. During our demo sessions, I tried a few others, including a Flea 48, Flea 49, and a Horch RM2J. They all sounded good, but the 44 was exceptional. Corrie heard the difference right away, and we never considered anything but the 44 after that experiment.

For her banjo, I tried a Coles 4038, a Beyer-Dynamic M160, and a Sennhesier MKH8050, but what sounded best was a Royer R121. The 121 also gave the best isolation from her vocal.

For Cate’s harmony part, I tried another R44, but the isolation was practically non-existent, even with the nulls from the ribbon mic aimed at Corie and her banjo. The room is just too small to expect much isolation.

I ended up using an AEA KU4, which is a cardioid ribbon mic, and that sounded very much like the 44, but with better isolation. I wanted Corrie and Cate close so they could watch each other and get the phrasing right.

I tried another R44 on the cello, and never needed to experiment. It was just the sound I wanted. It was pretty close – less than a foot from the instrument, and the 44 has a ton of proximity effect. That added some heft to the cello sound.

Even so, it needed some bass rolloff, so for both the cello and Corrie’s vocal mic, I used D.W. Fearn VT-4 equalizers with 4dB cut at 40Hz. I also gave the cello a slight boost at 5k. I used a similar roll-off on the stereo guitar track, from a VT-5. That was the only eq I used on any of the tracks.

For John’s drums, I have always used an AEA R88 stereo ribbon mic as the sole pickup. I put it about 5 feet out in front of the drum kit, with the center of the mic about even in height with the snare drum. Usually, I move the R88 up or down as needed to get the proper balance with the kick drum. But this time, I set it up and it sounded great from the start, so it never moved.

All the mics went through D.W. Fearn VT-2 or VT-24 mic preamps, and then directly to Merging Technologies Hapi converters, unless they had to go through a VT-4 on the way. On Corrie’s vocal, I also had a VT-7 compressor after the eq, with about 3db of gain reduction on peaks.

This was recorded with Merging’s Pyramix software.

The audio was monitored through PMC Two Two Six speakers, fed through a monitor controller I built for myself many years ago. Tube, of course. In fact, the entire recording chain is vacuum tube, except for the solid-state analog input stages of the converters.

There was a lot of prep ahead of the session. Corrie and I decided on the songs, the tempos, and the instrumentation. She wrote out chord sheets for all the songs. Well in advance of the session, everyone received the chord sheets and the demos. I discussed the plan for the session with the players ahead of time, and answered any questions. Plus, I asked them for any suggestions they had. I explained the studio setup and the challenges, and why I planned to isolate things the way I did.

My terrific second engineer, Ian Alexander, arrived around 9:30 that morning and we set up mics and headphones in general locations, to be properly placed once everyone was in position. Ian has been on just about every major session I have done since the mid-1980s. He is very good at mic placement. He is also a talented singer, which helps him in dealing with vocalists. Ian has a way of making everyone at the session feel relaxed, sometimes with his corny jokes and other times with the easy way he relates to musicians, making them comfortable, and helping them to do their best. I also had Ian run the session a lot of the time after everything was sounding right, which allowed me to concentrate on listening.

The first thing we did was to get everybody in the studio for a rundown. Everyone sat in a circle. They seemed to like sitting on the floor for the most part, with the music spread out in front of them. I made sure everyone had a pencil to mark their sheets.

And then Corrie just sang the song and played banjo. As everyone started to feel their parts forming, they joined it. They obviously already had ideas, from listening to the demos.

This was the recording the players received to introduce them to the song.

ANGELS TEETH DEMO

There was a lot of experimenting, not all of it successful, and many suggestions. Everyone had their own talents to bring, and the song slowly evolved as everything came together. During this process, I sat with them, encouraging things that I thought added to the song, and suggesting things that might work better. But my job as producer at this stage was pretty minimal. When I get great players in the room, I just let them work it out, with some gentle nudges.

I also worked with John to get the sound I had in mind for the drums. This was mostly just fine-tuning, since John understood what the song needed right away.

This run-through-process built enthusiasm for the song. When things sound great, it’s easy to be excited. By the time we had the song in a form that everyone was happy with, we moved to our locations.

From experience, I knew where most of the levels should be set on the preamps, and as Ian moved mics into their final position, we were immediately ready to record.

I always record the first run-through, even though there may be mistakes, or new ideas. Sometimes we even modify the structure of the song at this stage.

It’s not likely that that recording will be used for anything, but it is helpful to get everyone into the control room for a listen.

Listening to the playback has a dual purpose: if it is sounding good, and it did, everyone is invigorated. Plus, it allows them to hear their parts in context, which often leads to minor adjustments in what they play.

By the way, the headphone feed was identical to what I was monitoring, even though I can modify it if necessary. It was mono, and the same for everyone. I did have a separate mix available if anyone needed it, but that was not necessary. Players could adjust their own headphone volume, but even that was mostly set for them ahead of time and left alone. Experienced players are not picky about their headphones, in my experience, especially if you give them a good mix to begin with.

It wasn’t necessary on this project, but sometimes I hear that a player is having trouble with time. I know this is usually a headphone mix problem, and I will make an adjustment. It is amazing how that can pull a track together, and no one is aware of it. They just start playing better.

This song, “Angels Teeth,” is somewhat different from the rest of the songs we recorded, but it has become my favorite, and Corrie’s too. I think most of the players on the session consider it their favorite as well.

We did four takes, by which time everyone was comfortable with their parts. It was this take that became the final version. After a playback to make sure we didn’t miss anything that would be a problem, we moved on to the next song.

After the session, I made a quick mix to send to everyone. 

I don’t like to edit together multiple takes, but it would have been easy to do since the tempo was solid throughout. I never use a click track. John’s playing is always consistent, and he checks tempo with a metronome before every take.

As I listened to the first mix, the intro bothered me. But the dynamics of the song were great! I liked the contrast between the loud and soft sections.

Here’s an early reference mix.

ANGELS TEETH EARLY MIX

In a mix, my usual tendency is to make the things I like louder and things I don’t like as much, softer. It’s really more complex than that, but you get the idea. Bring out the good stuff!

The intro still bothered me, so I started paring it down, track by track, until it started with just Corrie’s vocal. Then I reduced the level of the guitar and cello, and muted the drum track. At the end of the first verse, the drums and banjo come in and the guitar and bass come up to normal level.

As performed, the second verse was quieter, a decision we made during the run-through. This was nice, but I thought it needed more contrast, so I took out everything except Corrie’s vocal and the cello.

You can hear a hint of John’s cymbal work that bled into the cello mic if you listen carefully. That was OK with me.

The acoustic guitar solo was overdubbed, and Adam really wanted to record that right away, an impulse I fully understand. In an ideal situation, I would have done that. But a producer has to keep the big picture in mind, and taking the time to work out a solo was not the best use of the limited time we had with everyone in the studio. I made the painful decision to move on to the next song.

There was time at the end of the session for Adam to record his solo, but he was not happy with it. Corrie and I think it is great, but we plan to give Adam another shot at it the next time he is in.

The next verse is even louder, but I did not have to do a thing. From here to the end, the group played the dynamics exactly as you hear them.

The end is quiet again. I like the symmetry of the song. And the drama of dynamics.

I had my friend Mark Minassian come in to put a subtle electric guitar part, at some key points in the song. They add some texture without being obvious.

This version is pretty close to how the song will ultimately sound.

ANGELS TEETH CURRENT MIX

Corrie’s songs all tell a story and I wanted to be sure that the story was easy to hear. She and I listened to Appalachian, Country, and Pop music, up through the 1960s, and it was apparent that they mixed the vocals way out in front back then. A lot of those songs were stories, too, and you never wondered what the words were. We agreed that her vocal should be out front, perhaps not to that extreme, but more than most current records.

If I have ever been criticized for my mixes, it is because I am too subtle. During the period of my career when I was doing mostly record label work, I had to work to change my approach so that the mix was hot and heavy, and things that needed to stand out would jump out at the listener. That was OK when I was recording pop or R&B records, and I eventually learned to be more heavy-handed in my mixes.

But I am not convinced that was the right thing to do. Going to vinyl, for radio station play, is one thing. But people have the ability to hear a far less restricted version of songs these days, and I wanted to return to my instincts for this project.

When I first listened to a reference mix in the car, or on a phone speaker, I was appalled at how practically all you could hear was the vocal. I made several more mixes with the vocal progressively lower in the mix. That sounded better in the car, but I did not like the way it sounded in the control room, nor on headphones or earbuds.

Consequently, I pushed the vocal back up to about three-quarters of its apparent volume compared to my preferred mix. That seemed like a decent compromise.

But I want the recording to challenge the listener to want to hear it better, so they can experience the full dynamic range and the subtle parts, like the cello in the quiet verse. Corrie agrees with me, and I think this is close to what the final vocal level will be like.

I am very sensitive to irritants in music, and I think there is a lot of music with huge amounts of distortion. With modern equipment, we can theoretically have vanishing low amounts of distortion. But with a lot of gear being over-driven, purposely or not, plus a huge amount of compression and digital limiting, the final result is very fatiguing for me to listen to.

Most listeners are not aware of the cause of their irritation, but they don’t need to know why it sounds bad for them to react negatively to the sound. Throughout my career, I have tried to reduce those fatiguing artifacts as much as possible, both in my recordings and in my equipment designs.

Contrary to contemporary practice, I do not use a lot of eq to make parts fit together, nor a lot of compression to slam everything to a zero level. Those techniques make a song sound reasonably good in a wide variety of listening situations, but it doesn’t really sound all that great to me under any conditions. Those are some of the factors that led me to the mix you hear.

For reverb, I used a combination of Quantec and Bricasti hardware units. The Quantec has some very nice ambience programs that add some room sound to my otherwise fairly dead studio. And then the Bricasti M7 contributes some bigger room sound. I am not certain if I will use these settings, but on this mix, the M7 is set to an acoustic echo chamber program, with some parameter tweaks, with an adjustment to the pre-delay and reducing the reverb time down to 1 second.

Since all the drums are on one stereo track, there is reverb on everything in the drum kit, including the kick drum. But I like the reverb on kick in this song, even if that is not often done.

The reverb send levels were adjusted on each track, but they all fed to the same reverb hardware, so there is a consistency to the overall sound of the reverb in the mix. I like that better than using individual reverb plugins on each track.

I do not use any compression on any tracks in this mix, except for the vocal, which I did while cutting the final vocal. I did not add any eq beyond what I used to record the tracks.

The only other eq is on the mix buss, with a 2dB rolloff at 30Hz and a 2dB boost at 10k, from a VT-5 equalizer.

Before the eq is a VT-7 compressor, set to about 3-5dB maximum gain reduction on peaks.

I also used a Flux Elixir plugin at the end of the mix buss chain. It is set so that it does nothing most of the time, but catches those occasional peaks that would otherwise reduce the overall level. That’s the only plugin used, by the way.

I aim for a -16LUFS final loudness, which required almost no change in any of the processing. I think a lot of well-recorded music, particularly acoustic music like this, naturally sounds best at that loudness level. In fact, I have noted that most acoustic music is naturally at around -16, without any processing.

That is not as loud as most pop records these days, which could be as much as 8dB louder. But I don’t worry about that. If people like the song, they can turn it up. Streaming services and radio play will even it out compared to whatever is played before and after. I think it sounds better with minimal processing, although it might not translate as well in all listening environments. I am counting on Corrie’s fans to appreciate the openness and dynamics of this mix.

What would I do differently if I had this recording to do over again? Not a whole lot. In general, I like the sounds, the parts, and the overall structure and the feel of the song. But I would probably have moved the R88 on the drums up a little bit on the kit for this song. The kick drum gets a bit too loud in places. It worked just right in the other two songs we did at this session, and I now regret that I did not catch that while recording this song.

I might also experiment with filling out the mid-range chordal structure a bit more with some additional instruments, mixed low into the texture. That is still an option, but we are running up against a deadline. I’m not convinced that adding any parts is necessary, and it could be a waste of time and money bringing in additional players.

I might back off on the reverb a bit in the final mix. I have never listened to one of my mixes years later and said, “I wish I had put more reverb on that.” A little bit of reverb goes a long way, and this song has about as much reverb as I would ever use.

It wasn’t our plan from the start, but as we completed, or nearly completed, all the songs, it occurred to me that these could be ordered in a way that the album makes sense as series of stories. This isn’t done much anymore, with the emphasis on individual songs. But I think we lose something when we think of each song in isolation, so I want to go back to an album concept. I suspect most people will not listen this way, but they will have the option. I think they will find that the collection is more than the sum of the individual songs.

Our twelve songs will make a nice CD, and although few listeners have the ability or the inclination to play a CD anymore, people who attend performances often want to take something home. We plan to have a run of CDs made for that purpose, plus for radio station play. We want to make a vinyl version, too, with fewer songs however. I suspect a lot of those tangible objects will never be played, but the CD or vinyl album fulfills a purpose, even if never opened.

We plan to have the completed songs ready for the world in Spring 2022.

That’s the story of what I have been up to recently, rather than making podcast episodes. If you found this story useful, let me know. I’ll do more like this. You can comment on the recording, too, if you hear something you think could have be done better.

You can reach me at dwfearn@dwfearn.com  The link is also in the description. You can also tell me what you would like to hear about.

And please subscribe, rate, and comment on any of the 30-plus podcast providers that carry this.

This is My Take On Music Recording. I’m Doug Fearn. See you next time.