West-Coast Stubbornness

Well, fantastic. I put the album online last Friday and a few people have heard it. Feedback from my friends has generally been good; so far no one has felt the need to tell me they didn't like it. There doesn't seem to be a clear-cut best song, as the nine people who've mentioned theirs to me have picked six different songs. Which is too bad in that I don't feel confident in heavily promoting a single from it, but good in the sense that it means I've done solid, consistent work. As I get better on future releases I can just try to up this average quality.

Speaking of which, I'm writing for the next album, which I fantasize I can be finished with by the end of February. Being able to produce an album's worth of material from scratch in three months would be a great turnaround time. That way I could do three a year, and if I can complete thirty-plus songs that I'm proud of, you'd think there'd be a dozen in there that anyone would be proud of, that would stand up to a proper commercial release (with some professional remixing).

Even if this is just an informal release, I'd like to try to promote it a bit, at least to get a sense of how good it really is and how far I still have to go. My promotional goal for this one is really just to get on a few people's musical radar who don't already know me personally. And to make some industry contacts, at least find some people I can annoy again when I'm done the next one. I also want to contact some of the artists that inspired me while making this last one, with no specific goal except that if one of them happened to like my stuff it would be gratifying. So I suppose promoting this album is about letting some people know I'm throwing my hat in the ring, and testing the strength of my material. We'll see what comes of it!

Traditional/Not Traditional

I have finished my first draft masters for the album, and told the (dozen or so) people that want to listen to it that it will be done by Friday. I have yet to really test the songs on other sound systems, but listening to them on headphones is quite gratifying. I've never spent any time learning the mastering process so the last couple days have been great for me. I don't usually think of mastering as a dramatic process but I think I've dramatically improved on the sound of the mixes; this was the missing link in my record-making chain.

One neat thing I've done is used EQ matching to try and reshape the overall tonal balance of my songs to a professionally recorded one. I grabbed four songs from the sample CDs I'd uploaded previously and tried matching the lead album track EQ to them. They were "The Suburbs" by Arcade Fire, "Nowhere Lullaby" by Built to Spill, "Losing My Religion" by REM, and "Black Day in December" by Said the Whale.

The Built to Spill track is quieter on the high end, and applied to my mix it sounded a bit dull and muffled. "Losing My Religion", with the bright string instruments, had the opposite problem. The Arcade Fire track had more low bass than the other reference tracks, which didn't make an audible difference to my mix, but it also had a weird curve in the high frequencies. Instead of following the 6 dB per octave curve (like the other three do, at different levels) they start to roll off around 10kHz and continue downward parabolically. When applied to my mix I found it made the mids sound a bit boomy. I found that the Said the Whale track to be clear and bright, but not overly so. I loved how my mixes sounded when matched to this EQ, especially with a bit of tape saturation applied as well. 

So I tried it across the whole album. Since the instrumentation varies considerably, this meant a lot of tweaking, but right now I have an album that sounds much more coherent than anything I've done before. I'll try it on some other systems and probably have to work on the bass, but for now I'm excited.

35 People are Here

I got the mixes to where I thought they were solid on my own, but today I let my immediate family listen to them. It was also the first time I've heard the mixes on a commercial-grade stereo, which was very helpful. Mixing as I do on tiny 3" monitors or Sony MDR-7509HD headphones, judging the bass level and evenness is tough. Fortunately the bass sounded fine on a few of the songs, so I'm hoping I can just replicate those settings on the problematic ones. More a/b testing is required.

On first listen they didn't have many specific concerns about the mixes or the track sequencing, except to redo one vocal line and turn the trombone down, so I think I'm doing okay. I'm getting better at setting the vocal level but melody instruments are still tough for me (check how loud the violin is on this EP I mixed); I think it's that I hate to bury a part I've worked hard on. But as I tweak the mixes I find I'm thinking about them less as "band" recordings and more as voice and guitar recordings that happen to have other instruments to fill out the sound. Because I don't have a band, and when I perform these things I'll likely be doing them solo, so the mixes should be set up so that listeners can at least conceive of them that way.

I haven't actually given much thought to adapting them for performance yet. Instead I'm conceptualizing material for my next album, which may even be another notch softer than this one, and this one's already pretty AAA-friendly. I'd like to try recording the next one very simply, without a click, and with minimal instrumentation. I've learned all kinds of things on this last project all ready and I'm sure when I put it out I'll get all kinds of useful feedback as well. So the next one can only be better!

The Story of a Sound

I've never liked like the idea of using reference tracks. I listen to a fair amount of music and figure that I know what it sounds like, and know when it sounds good, and I don't need to check my answers in the back of the book. But I'm forcing myself to do it with this album, so I grabbed a few CDs that I had lying around and imported them.

Looking at them in waveform made me feel a whole lot better. These CDs have been mastered completely differently, but that doesn't stop people from loving them the same. So there are no right answers, although Bob Katz argues there should at least be standards and I think his argument is a good one.

I used the Izotope Ozone 5 Mastering plugin to measure a few parameters that I could compare across my sample CDs. I compared the loudest moments (M), the average loudness (I), the loudness range (LU), and the meter peaks for the left and right channels. The unit is LUFS, for what it's worth.

The loudest tracks are the newest. The Old Canes' "Flower Faces", from 2009's Feral Harmonic, is going for lo-fi and plans to misuse your stereo equipment to achieve a very dirty sound, which is great and suits the music just fine. The contemporary Arcade Fire, Coldplay, and Said the Whale tracks I picked were also very loud, but fairly clean. The Deer Tick album is an exception. It's more in line with the 90's albums (Built to Spill, Fiona Apple, REM) in terms of loudness, as is the Jam's "That's Entertainment" which was digitally remastered in 1997. The Beatles' "Here Comes the Sun", remastered ten years earlier, is by far the quietest.

The loudness range isn't very comparable since it's dependent on song structure and dynamics, but they can be guessed pretty easily by looking at the waveforms. The Arcade Fire, Old Canes, and Said the Whale songs almost have a binary dynamic range; they come on and there is music (loud) or there is not music (off).   The Built to Spill song I happened to choose has the clearest dynamics, but the rest of the album behaves much like contemporary late 2000 releases, at least on my tiny sample size.

The highest peaks show us that only the Old Canes and the Said the Whale tracks exceed 0 by any appreciable margin, while the Arcade Fire and Fiona Apple tracks achieve a L/R sum of zero at their loudest moments. The Beatles remaster leaves a lot more room than any other track.

All in all, I think even with my small sample I've learned that CD mastering is different than it was ten or twenty years ago, and I know what loudness to shoot for when I try to master mine, but that I shouldn't equate "modern" with "correct".

Top Album Is Delicious

Mixing is getting done, I suppose. I've been doing a morning listen, noting everything I want changed, and taking the day to do those fixes. Of course one change may necessitate another, or reveal problems that were hidden before. I feel like I'm trying to move a house of cards across a room. Each round of fixes is getting shorter and shorter though, so it won't be long now.

I've been doing some art direction things to give my ears periodic breaks during mixing. I made a bandcamp page for myself and am experimenting with different colours and fonts. They allow track-specific art so I'm preparing a series of photos, each captioned with a judiciously chosen couplet. For the header I'm using the "Marquisette" font right now, overlaid on a picture of my wall and ceiling meeting. The imperfect line of paint is supposed to contrast with the clean lines of the art deco font, demonstrating that the execution of an idea always falls short of imagination. Or something. I really hope the White Stripes haven't copyrighted the colours red, white, and black.

You think you're right/NO! /You are wrong!/WRONG!

Principal recording has been done for awhile. The tracks are nearly mixed, for pete's sake, but I still get sucked in to vocal reshoots. Change a word here, change an inflection there, see if I can hit that note this time, so on. Some vocalists don't even like to re-record anything less than a verse, but I don't mind redoing a syllable if my ears tell me it's better that way.

I'm all about lyrics, and sometimes changing a word can change the song's whole message. I mean, if Hendrix meant us to excuse him while he kissed this guy, we'd probably have Purple Haze poppers by now. Oh wait, apparently we do. In any case, the two verses that I've been most hesitant about happen right in the middle of my (ever-changing) prospective track order: the last verse of track four and the intro to track five. This isn't a concept album and the songs are independent entities, but it's still a critical moment in the album and I'm vacillating between saying a couple of different things. I have to decide what kind of stories I want to tell, what conclusions I want to draw.

I don't want people to listen to the songs and think that I haven't fleshed my main character out enough. Because that's me, and I get enough of that in real life anyway. I look at my songs as a place to work out and  concretize my emotional reactions, and in fine-tuning I'm molding words and music into something that resembles my internal feeling and memory as closely as I can. I direct my writing by using memory-images: situations and moments in my past that are tied to the specific emotion that I am trying to channel in the song. I'm not Capote and I don't have 94% recall, so the imagery and dialogue I lift from my real-life situations (sometimes I write scraps down in a notebook) is usually filled in somewhat with imagination and invention, with the only rule being that it doesn't stay unless it 'feels' right to me. I think it's possible to tell an emotionally true story which differs on some key factual points from real life, as long as the diversions are in line with what I felt at the time, representative of the forces I perceived were at work. That's artistic license.

But that's just me describing the method. Wait 'til you get the madness!

Returning As A Hologram

I have a working computer again. Hooray! And all it took was a new motherboard. It took some time to figure out, but I'm glad I worked it out myself. Speaking of which, here are some (no doubt elementary) ways I have improved my workflow while working on this album:

• I've started aggressively multi-tracking parts. I realized that as individual instrument dynamics change from section to section in a song, a new fader balance is usually needed. So now if I decide I want the acoustic guitar 3db quieter on all of the verses, it's easy.
• I've also started using multi-tracking to apply special effects to only certain moments of a part. For vocals, I'll set up an identical track where I'll just put the words I want delay or reverb applied to. Then only one instance of the plugin has to be run as opposed to a new instance for every clip.
• I've started sending these multi-tracked parts to a single bus. Having individual busses for each major instrument part is great, because you have a fader for the whole instrument and if you want to limit the whole bass part, you can use a single instance of a plugin on the bus. Additionally, it also makes it super easy to export mix stems, if you solo the appropriate bus beforehand.
• I've started panning reverb. Never did that before, for some reason. But it goes a long way towards creating a believable stereo image.
• I've started using the nudge function. I used to always line up waveforms by hand but now I've customized the nudge lengths to times that work for me. I have them mapped to my number pad so that 1 means nudge 1 left, 3 means nudge 1 right, 4 means nudge 2 left, 6 means nudge 2 right, etc. It takes me a while to realize when I need hotkeys for things, but I'll get there.