Wednesday, March 25, 2015

Popular Music Songwriting = Lyrics + Approximation of Melody. Everything Else Is Arrangement.

I find infringement / plagiarism lawsuits involving songs in the popular music format (verse-chorus-verse, et al.) to be fascinating -- and unnecessary. If I can relay what I've read more-or-less accurately without looking it up, the case involving "Blurred Lines" is currently in favor of the Marvin Gaye estate. Apparently, the jury found that the rhythm section, bass and percussion, of Robin Thicke and Pharrell Williams' production of "Blurred Lines" had a very similar vibe as Gaye's "Got to Give It Up." Musicians who have analyzed and compared both songs have found that the bass line, while having similar rhythms, are actually using different modes -- basically scales, without getting too much into the nitty-gritty of music theory.

The jury, it seems, has created a precedent in which genre conventions -- e.g., funky basslines and more cowbell -- can be grounds for copyright infringement. I am certainly glad that the birth of rock 'n roll, in the 1950s, wasn't so litigious. If you haven't heard enough '50s music, there were basically three ways to write a rock 'n roll song:  (1) 12 bar blues with a boogie woogie backbeat, (2) a waltzy doo-wop with a I vi IV V chord progression (e.g., C Am F G, or G Em C D, etc.), or (3) a properly composed pop song.  But really, early rock 'n roll was about the 12 bar blues and/or the doo-wop progression. My point is that almost every '50s song sound like a plethora of other '50s songs.

I think an overlooked problematic area in "Blurred Lines" was that, according to what I've read in the song's composition process, it was basically cobbled together in the studio. It seems like Thicke and Williams, and perhaps T.I. too, were jamming with keyboards and drum machines and computer loops and vocal riffing. They created the backbone structure of the song to have the same groove as similar songs in the ballpark of a genre -- funk / pop / or whatever. They then brought in session players for the rhythm section, if they didn't want to track it themselves, and the entire envisioned vibe came to life.

They might or might not have done this, but the writers of the song should have written the song itself on paper, and not just lyrics. Ideally, one of them -- or their assistant(s) -- should have taken the, I'm guessing, Pro Tools session and lyric ideas, and transcribed a lead sheet:  Tempo approximation at the start, chords on top, treble clef melody in any key with meter in the middle, and lyrics on the bottom layer -- all the way through the page.

That should, in my learned-by-trial-and-error, yet somewhat idealistic view of music, show whether or not a song infringes on another song. In fact, the song itself is really only the lyrics and an approximation of the melody. In the previous paragraph, I listed the basic elements of a lead sheet:


... and four (and a half) of the six are fluid parts of the song, and can be changed in process of arrangement, especially when an artist is actually performing a rendition, or cover, of another songwriter's song.

Tempo: Speed it up, slow it down; it really doesn't matter. The song will be the same, no matter what pace.

Chords: Every major chord/scale has a relative minor chord/scale, and vice versa. Every riff implies a strummed chord, and vice versa. A diminished chord is basically a dominant seventh chord without a root note, and the reverse can also be true, depending on the key. It is a basic skill for any musician to substitute chords, while maintaining the integrity of a song.

Speaking of tempo and chords, compare the Rolling Stones' "Satisfaction" with Devo's cover. It's the same song at heart, no matter the delivery.

Meter: It might feel a bit weird at first, but a backbeat rocker can easily be transformed into a jangly or romantic waltz, and vice versa.

Key: Any key can be transposed into another key. Next ...

Melody: Due to artistic choices or perhaps physical limitations, a cover song's melody can still be in the ballpark of the original intent and still be a cover. Compare the major key electric "Number of the Beast" by Iron Maiden with the minor key, darker, grittier acoustic cover by Zwan. Substituted chords also affect the melody.

Lyrics: A musician covering a song might change a word here or there (e.g., Frank Sinatra covering the Beatles' "Something" or Johnny Cash covering Nine Inch Nails' "Hurt"), if most of the words are the same, it's a cover. If extremely similar lyrics are not admitted as such, then it's plagiarism.

In my opinion, genre conventions shouldn't be grounds for intellectual property infringement. The only focus on these types of lawsuits should be in the areas of melody -- and I think I've given a large ballpark on what can be considered a cover, so that might be applied to similar melodies -- and lyrics. And even in regard to melodies, there are only so many combinations that are accessible to any given culture.

So I really don't know, do I?

Friday, March 20, 2015

How to Record A Power Trio, Circa 1975, with a 2015 Super-Low Budget

On a personal level, new music recordings are just not exciting. This sentiment might even transcend all forms of popular music. But let's talk about rock music, in its purest form: A band consisting of guitar, bass, and drums, with a singer who either plays one of the instruments or does not -- the power trio.

For established bands, either popular since last century or currently receiving record company advances and marketing, recording is no problem. They just need to spend X amount of dollars at a well-equipped, acoustically-treated recording studio with isolation booths and lots of space. They can record on actual tape, with a legendary recording console, etc.

For fool-hardy, no-name, upstart bands -- the modern-day, low-budget recording process is awful. They can either noodle with their computer(s) at home, or go to a $20/hour recording studio, but the process will likely be similar: Record one instrument at a time, recorded with a click track, then quantized and Autotuned to sterility. (Been there, done that; not my favorite thing in the world.)

After decades of trial and error, I think I have a low-budget process (a couple thousand, tops?) that takes the idealism of 1970s-style recording with the relatively "ethical" parts of the Pro Tools era. The conclusion of the recording process, a rough mix, will unfortunately sound quite sterile but it will feel human.

For this situation, I assume that the band is a power trio (guitar, bass, and drums), and the lead singer is either the guitarist or the bassist (or the drummer). I also assume that the recording space is one room, with no isolation booths whatsoever, and with no acoustic treatment.

The first thing to do is to record the band, live. You'll need a decent computer with multi-track software, with enough RAM and hard drive space to handle a maximum of 16 tracks simultaneously. That will be your "tape machine." You'll also need a hardware interface with inputs that can handle all instruments: One for guitar, one for bass, one for kick, one for snare, one for floor tom, one for mid tom, one for high tom, one for hi-hats, one for ride, one for one crash, one for another crash, etc. Since this recording happens in one room, with no isolation booths, you'll have to direct input from either the guitar/bass to the interface, or from the amp to the interface, no mics. Since the room wouldn't be ideal as a drum room, you won't traditionally mic the drums with stereo overheads. Every piece of the kit should be mic'd because, in the end, you will have to replace the drum recording with samples, unfortunately. There is software that can analyze the velocity of each hit and the output digital "notes" will have relatively similar variation.

The alternative to mic'ing all drumkit pieces is to use an electronic drumkit, and plug that directly into the computer or interface. The end result -- the use of samples -- will be the same. Then you really won't need 16 inputs, if you're using an electronic drum kit -- two should suffice for the guitar and bass.

Record the band live, with no click track. The push and pull of the rhythm is basically what makes the performance feel human. You're going to have to record multiple takes, until the drum performance feels right. (You'll want to keep the quantizing and Autotuning to a minimum, if any.) With the best take, punch in to correct any bum notes for the bass and guitar. Let the guitarist record overdubs, extra rhythm guitar, lead guitar, guitar solos, as necessary. You'll have to go somewhere else, with great mics and a good-sounding room, to record the vocals. Or lean a mattress against the corner of the room and use an SM58 -- that might sound okay.

Since no mics were used in the recording process, other than as ad-hoc triggers for an acoustic drum set, to be replaced by drum samples, the recording will sound quite sterile -- a little too perfect. Using studio trickery to, ironically, dirty-up the sound will happen during the mixing process. (Don't go crazy with the overdubs and additional instruments. Get someone else to mix it. Or if you're mixing it, don't muddy up the mix. And don't fight the loudness war when mastering the mix.)

Performance-wise, assuming no quantization or elastic timing fixes are used, the musical feel will be human. Underneath all the overdubs, you will still have three musicians playing off one another's vibes, with a solid but not perfect performance, and that is something that cannot be replicated by any machine.


Well, one can always program a drum machine to play at 100 beats per minute, then 99.7 bpm, then 101.1 bpm, etc., but that really isn't the same.

These days, recorded music is cheap. MP3s are low quality. It's free to stream if you can find a particular song. It's cheap to buy, but who buys? Long ago, concert tickets were cheap because they promoted relatively pricey albums (remember $17 CDs?). Now, cheap-to-access music is probably the promotional material for expensive concerts (for big bands) or band T-shirts at small shows (for smaller acts).

The moral of the story for upstart bands: Be good, don't use a lot of trickery, but don't break the bank.

I'll probably repost this rant at Chord du Jour, with links to purchase some equipment. That site should have the money-hungry affiliate links. This blog, not so much.