I just got back from FedEx / Kinko’s where I picked up the final proofs for the trio CD and shipped it all off to National Tape and Disc: CD-R master, CD of PhotoShop art files, proofs, intellectual property rights form (declaring we’re not using NTDC for bootlegging), and the first payment, all traveling together in a bubble mailer, nice and cozy.

It should arrive in Nashville in three days, and come back in another three and a half weeks.

Woo-hoo!

Yesterday I got the final approval from Jerry and Craig, with one revision to the booklet. I did that revision and took it to Kinko’s to order a new proof, and then went down to Wilburland to get some Taiyo Yuden CD-Rs and show Matt our work. After listening to bits of two tracks, he said it sounds good, better than it ought to for a home recording. Gratifying to get approval from a real engineer.

I hadn’t listened to the CD since last Wednesday when I sent it off to the guys for their approval. Too much “in it,” I was afraid that if I kept listening to it I’d keep hearing things I would want to try to change.

When I got home and burned the tracks onto the new CD-R, though, I had to listen to it to make sure it burned properly. About half-way through I started getting nervous and thinking there might still be things I needed to fix. Even though I know that I’m a high-anxiety type and overly sensitive, it was hard not to take this nervousness seriously.

I listened again a few hours later, and decided it’s fine.

It’s fine!

Not perfect, but if we waited for perfection we’d never get anything done. Especially if, like me, your standard of perfection gets more and more absolute the more you work at something. I think imperfections, imprecisions, inaccuracies, and so on, are a given in this earthly life. It’s a fallen world, and we’re fallen people. It’s still important to work on narrowing margins of error, but I think they will always exist this side of heaven.

The trio CD continues to near completion… still.

We are done recording, and I’ve got decent mockups of the graphic design, so all that remains is finishing.

Thursday evening we got together to listen to the CD. We agreed that the panning and balance among instruments seemed fine, and the volume from one piece to another seemed consistent. But in general, the guitar is just a little too bassy or muddy on some tunes, and the fiddle just a little too sharp.

I had already experimented a little with the EQ plugin to do a bass cut on the guitar and a high cut on the fiddle. It’s hard, with little computer speakers, to tell how much difference it’s making. Apparently not enough yet.

Yesterday I revisited the first four tracks and tried again. It’s certainly a learning experience.

One very useful thing I discovered is that you can have the plugin open during playback, adjusting and instantly hearing the change, even turning it on and off to compare to the original.

I’ve also been experimenting with what to adjust, too. I started with presets for bass cut and high cut, and have been modifiying the gain and bandwidth knobs — I only have a vague idea of how they work, but by tweaking one or the other or both I seem to be able to get the sound I want — at least on the computer speakers. We’ll see what happens with the CD player.

A very surprising discovery was that on one piece where the guitar starts with fingerpicking, the guitar track has a lot of bleed from the fiddle and dulcimer — perhaps that mic’s gain had been boosted to get a strong enough signal compared to flatpicking. So I added the fiddle EQ adjustment to that track as well, and I also understand now why everything sounded balanced even though the guitar track’s meter runs higher than the other two.

Two other discoveries, not EQ-related:

1. “Spootiskerry” is not a traditional tune. It was written by the late Ian Burns, copyright 1980. Thanks to Susan Songer (of The Portland Collection), I was able to get in touch with his daughter to get permission to include the tune. Whew!

2. For medium runs (500-1000 CDs), it’s only pennies more expensive to have them professionally replicated (manufactured), versus professional duplication (CD-Rs; actually more expensive than replication for this many CDs), having imprinted blank CD-Rs made and the graphics professionally printed, or doing everything ourselves.

Replication costs more up front, and you have to commit to a pretty high number of CDs, but the quality and appearance is better, it’s less time and work on our part, they’ll be shrinkwrapped, and even if we only sell 100, we’ll at least break even.

So — we’ve decided to have the CDs done by the same company (National Tape and Disc) that pressed my two solo albums.

Last night The Hanshaw Trio recorded our guitarist Craig’s tune “Cherry on Top.”

This is the tune with a syncopated A part — a lot of the accents fall one eighth note before the downbeat, so that it’s easy to lose the feel for those downbeats, tempting to shift them ahead a bit and lose the syncopation. Fiddler Jerry and I both struggled with this, but in different places. Then the transition to the unsyncopated B part always throws me off a little, and it’s hard to get the expression and timing of my part just right. That and my hand hurt last night.

Still, it’s been a fruitful and rewarding experience to put a trio version of this tune together.

A few times at past sessions, while waiting for something to save, Craig would start playing the tune and Jerry would improvise some long sustained notes over it. I may have tried a few things myself, although I think I hummed them rather than trying to play them.

With some difficulties — because of that syncopation and because I’m not really a guitar player — I managed to transcribe something like the tune, and draft some preliminary fiddle and dulcimer parts.

We talked through some arrangement ideas, put them on paper, and played with them last night, adjusting and refining and adding as we went.

Guitar starts alone. Fiddle comes in with a long sustained note and a little run. At the B part, dulcimer replaces fiddle. Second time through, it’s everybody. Third time on the A part, the dulcimer part gets fuller, mixing melody and response parts. Then we cut to an alternative B part — at first it was going to be just guitar, but we added some dulcimer bass stuff — then a single strum, a pause, and back with a bang to the regular B part with everybody, closing with a slight fiddle slide into the final note.

It’s one thing to decide how to do something, and another to actually do it.

After trying a few times to run the whole thing, we decided to do three takes without that pesky fiddle run. Then on a new track, we worked on putting those runs back in with the right timing. I think I did something wrong on each take, but maybe with editing we can put something smooth together.

Later.

The Hanshaw Trio met to record this past Tuesday; we worked on Hills of Lorne and Midnight Maze, then had a little time to run through some Christmas music.

Hills of Lorne is a slow piece. We start with just fiddle, then add guitar on melody and dulcimer on harmony for the second time through, then all three playing melody the third time.

For the second time through, I decided to use hand-damping.

On a dulcimer, every note keeps ringing after you strike it. Different builders and designs have different sustain lengths, but sustain is part of that characteristic dulcimer sound. Sometimes, however, a player might want to reduce or stop the sustain. Options include mechanical dampers mounted on the sides and operated by a foot pedal, threading yarn or other material through the strings, or hand-damping.

There are different ways to hand-damp. Some hammers, especially in other countries, are designed to let the player easily turn their palms to the strings without the hammer getting in the way; it’s a bit trickier with typical American hammers.

I based my method on some instruments in the Indonesian gamelan, which require the player to strike the current note with one hand while the other hand damps the previous note. It’s an interesting challenge to think simultaneously about the current note, the next note, and the previous note.

For the third time through, on one take I played normally, so that there is a contrast not only of melody versus harmony, but sustained versus damped. After that take I thought I ought to damp this part, too, which I did for the other takes. Now, though, I like the contrast better.

We’ll also add mandolin to this time through, and perhaps something else — maybe some kind of drones or descant.

I’ve spent a lot of time already on engineering Hills of Lorne. We had four takes, and I’m using three of them. Take 1 for the fiddle solo, takes 0 and 2 for the rest.

The third time through is the trickiest, because there’s only that one take with un-damped dulcimer, and the fiddle on that take is not as nice as it is on the other. But I’m almost satisfied with the way I’m editing the takes together.

I haven’t even listened to the Midnight Maze takes yet.

Yesterday I started mixing and editing “Down the Brae / Ballydesmond Polka #2 / Ballydesmond Polka #3,” the most recent track for The Hanshaw Trio‘s home recording, and today I just finished mixing it down.

Mixing involves things like panning, effects, and volume. I set the panning so that the fiddle is 33% to the right, the dulcimer 33% left, and the guitar is centered. Then I add just a hair of reverb to the fiddle.

The interesting part is drawing the volume envelopes. If I had a mixing board, I would move the faders manually; newer boards can record this movement so that mixing can be automated. With the software we’re using, Cakewalk’s Guitar Tracks, I could do the same thing, using the mouse to move the faders and having the program record that movement. However, I find it easier to use their envelope method instead.

You start by creating an envelope, which shows up as a straight line on the track, with a dot (node) at each end. You can then add more nodes and define the movement from one to the next (jump, linear, fast curve, slow curve).

My first task is to mix large chunks, like sections where dulcimer has the melody or sections where the guitar is fingerpicking. Then I may have to make smaller, shorter adjustments like when a particular guitar strum goes over 0 dB (which creates distorted noise), or where the fiddler stepped away from his mic a bit and therefore needs a boost. I think some folks use a tool called compression to deal with the guitar spike problem, but it’s easy enough just to add a dipping node at each spike, so I haven’t explored the compression option.

This particular medley involved editing the best “Down the Brae” take with the best “Ballydesmonds” take. We started the “Ballydesmonds” takes with the last two measures of “Down the Brae”; the overlap gives me more elbow room to find the best editing point. I ended up switching dulcimer tracks at the point where I hit a bass note before starting a set of arpeggios leading into the “Ballydesmonds.” I switched the fiddle and guitar over a bit later after their last “Down the Brae” notes had faded out.

Today’s work involved finishing mixing the individual tracks, and then mixing down the three tracks into a single stereo track, which will later be converted into a WAV file ready to burn onto a CD.

I like this medley. It has a lot of energy and momentum. And I’m especially impressed with the guitar work on this one. Kudos to Craig.

Lunchtime!