The Game Composer’s Blog

Game music, a matter of life and death.

Word of the day: Frankenstein (verb)

000264_11

Frankenstein: verb – To patch together the best bits of recorded live takes for use in an audio recording. Commonly used to take the best segments from multiple takes of a performer’s recording and create a more perfect overall track.

I first started using the word “Frankenstein” 9 years ago when I first started doing music for games. I commonly played several different instruments in my music. Maybe while playing trombone I would play one section great in one take but then find a note that’s a little off later on while listening in the monitors. I always do a few takes even if I think I’ve nailed it (just to be safe) so I would simply take that note from another take and crossfade into it then out of it. And, voila, I just Frankensteined!

With digital music recording, it has become extremely easy to take bits and pieces from various takes and patch them all together into one long, almost-perfect track. And when done correctly, the listener has no idea that they are listening to a patchwork of several different recordings all crafted into one whole. One of my professors in grad school conducted and produced new music recordings and was always extremely proud to say how many cuts there were in the final product and how undetectable they were. And, indeed, nobody can tell when you’ve Frankensteined.

I guess some purists might say Frankensteining is an abomination. Don’t you lose some of the heart or soul from a player’s performance when you stitch together takes like that? Maybe so. But when doing a recording session, a player will not really vary that much between takes unless it is a part of the music, like often it is in jazz. Breaking up a long jazz solo will be jarring, better to take a longer performance, warts and all. But there are a great many types of music that doesn’t apply to. If a long held note is out of tune, nobody will notice or care if you fix it by Frankensteining. In fact, they’ll probably appreciate it. Of course, the term itself does have some negative connotations…

I was going to claim on this blog that I’d invented the term. Back when I first started using it, I’d never heard it used before and I was very proud to have invented such a cool term that so perfectly describes what you’re doing. Unfortunately for my ego, though, after speaking with friends and other audio people, I’ve learned that they use it too. I guess it is a pretty obvious term to use. Oh well. Still, it’s a great word and a fantastic technique to help your recordings sound as good as possible.

February 6, 2009 Posted by thegamecomposer | Ponderings | , | No Comments Yet

TSN… #4: TSN overuse Auto-Tune.

… overuse Auto-Tune.

For those of you who don’t know, Auto-Tune is a tool people can use to correct pitch problems in recordings. It’s used very often with voices, especially in pop songs. It’s a very powerful plug-in and useful when a singer has done a terrific take but has a note here or there that just isn’t in tune enough.

The problem stems from that little song Cher sang called “Believe” in 1998 that used Auto-Tune in a new way to create an interesting (back then) new type of vocal effect. Suddenly, everyone had to have the “Cher effect”. And apparently everyone got the Cher effect plugged into their own studios, because you can hear it all over the place now. I’ve even been asked randomly several times from people how to go about getting that effect. My first question is always, “Why?”

Again, I love pitch correction software. It can be really useful. And, actually, I’m not complaining about the “Cher effect” in this blog post. People aren’t using it that much anymore and when they do it’s usually in an uninteresting song by an unimaginative group (so who cares what tools they misuse anyway). I’m complaining about the overuse of Auto-tune’s intended use: pitch correction.

Auto-Tune should not be used on every single note. Used aggressively like this, it will suck the life out of a performance. It’s a human singing and humans aren’t perfect. What makes the difference between a good singer or a merely tolerable one is often the subtleties, and Auto-Tune can strip away far too many subtleties leaving a bland, computerized experience. It even sounds computerized. When engineers just throw on the Auto-Tune switch, it seems sometimes they neglect to listen to the results. It can sound very artificial.

It’s amazing how prevalent Auto-Tune has become. I laugh every time I hear it, especially in a country song. Country isn’t about perfectly in-tune singing. It’s about heart and strife and livin’ and sporks. Oh, maybe not sporks. But Auto-Tune overuse just obliterates all that emotion and leaves a sterile homogeneous borefest of a song. Combine it with aggressive compression and Marvin the Paranoid Android might as well be singing.

Anyway, this isn’t purely a game music specific topic, but it is a decree that I hope all game composers will heed in their music when the time comes to use that certain plug-in.

*note:
It’s debated whether Cher’s engineers used Auto-Tune or some combination of a vocoder and other effects. It sure sounds like Auto-Tune to me. Also, there are many other pitch correction software plug-ins out there and I used Auto-Tune in this post because it’s probably the most well known, but this post of course applies to all of them.

July 27, 2008 Posted by thegamecomposer | Thou Shalt Not | , , | No Comments Yet