We start with basic selections of hardware and software. The software part is easy: I will dictate it to you. The hardware part requires some discussion, but in order to discuss the hardware we need to first talk about the software we'll be using.
To record basic tracks, we'll be using Ecasound, a popular command-line sound recorder and processor. Ecasound supports everything under the sun, but we'll be using it only to record and to play back tracks. More specifically, we'll be using it to play back tracks already recorded while recording new tracks, and we'll be recording the tracks one at a time.
There are at least three choices for mastering: Audacity, Ardour, and Ecasound. We'll use Audacity because it's entry-level. Its user interface is easy to understand, yet it's very powerful. When you've been doing this for awhile, you might find you'd prefer to work with Ardour or Ecasound instead.
While Audacity supports recording, its a fairly resource-intensive piece of software, and I haven't had a lot of luck with recording with Audacity. There are hundreds of
thousands of users of Audacity whose experience contradicts my own, though. If you elect to use Audacity to record as well as master, you can eliminate Ecasound from the list of needed software.
Hardware is where you may want to spend some money. Manipulating sound is CPU-intensive work. The multi-threaded nature of these applications means that a dual-processor computer will make the work go faster, though a faster front bus and more memory will serve you even better than a faster processor. Sound files are also very large, so you'll need lots of memory to make sure there's plenty of buffer space and lots of hard drive space to store it.
Sound is generally defined as vibrations in matter. When you strum a guitar string, the string swings one way, applying pressure onto the surrounding molecules creating a high pressure zone. Then it swings back, creating a low pressure zone. The number of swings it makes each second determines pitch. That's where the word "frequency" comes from. The frequency of a sound is the number of high and low pressure waves the sound consists of, measured in one second.
A microphone converts the high and low pressure waves that reach it and varies its voltage output accordingly. This voltage output is referred to as "analog" sound because it is analogous to the original sound wave.
Digital sound is one step further in the mix. The analog to digital converter sits on the other end of the circuit from the microphone and takes periodic snapshots of the voltage levels coming from the microphone. These snapshots are referred to as samples.
The best way to conceptualize digital audio is to think of it working the same way film does. In a movie, you see a series of still pictures in rapid sequence to give the impression of movement. The fewer frames shown per second, the rougher the movement appears. In digital audio, a sample is the equivalent of a movie frame. A sample represents a single snapshot of sound and can be represented in a number of different ways. Each frame is a given size. When we refer to 16-bit digital audio, we mean that each frame represents a single 16-bit integer (a integer value ranging from -32768 through 32767) that represents the amount of voltage found in analog sound for the same noise. The number of frames per second, or frequency, is measured in hertz. CD quality digital audio is 16-bit 44100Hz, or 44.1kHz.
Sixteen bits isn't a lot of space to represent a sound, and the small size becomes a problem if you want to modify digital sounds. Applying an effect to the sound can be restated as "doing complex math on a series of samples." Every time you apply effects on 16-bit integers, you lose 3dB of sound in the affected area due to the low resolution of the sample. Complex math with 16-bit integers requires a lot of rounding decisions to be made along the way, and that's one place you will lose sound. The other place you will lose sound is at the upper end when samples start to clip, or exceed the maximum voltage the 16-bit integer is capable of representing.
Sound editing programs that want to retain the fidelity of digital sound use 32-bit floating point numbers internally. Floating point is a way of representing decimal numbers in a computer that only understands integers. In the 32 bits occupied by a floating point number, some of those bits represent a number on the right side of the decimal place, and some represent a number on the left. Put them together, and you wind up with a number like 3.1415. The advantage of using floating point numbers is that you don't have to make as many rounding decisions while performing complex math on a sample, and any rounding decisions you do make will have much lesser impact than they would with 16-bit integer samples. You can consider effects processing in a 32-bit float system as lossless and you won't go far wrong. Deep down inside, there are actual losses due to the imperfect nature of floating point math, but you'd need a bat's ears to discern the losses.
Sound hardware is the hard part. To build a recording studio of maximum quality, you can buy an external analog/digital converter, but the most inexpensive solution is to just use the sound card you already have. It's an acceptable solution for these reasons:
- It meets the minimum requirements of 16-bit 44.1KHz input on at least one track (and usually two are available, if you want to use them).
- The software we've selected uses either 24-bit int or 32-bit float internally, so there won't be any loss of sound normally associated with 16-bit samples during the mastering stage.
- You'll be recording one track at a time. If you need to record more than one track at a time, you'll need to spend some money on a different sound card.
You need to either place a decent quality microphone near your amplifier or instrument, or be able to plug your instrument directly into the sound card. Since microphone placement is an art into and of itself, I just plug my guitar into a Boss GT-3 digital effects processor and use the line-out jack to plug directly into the stereo line-in jack of my sound card. I have used inexpensive analog mixers that deal with converting instrument signals to line signals. In the long run, though, you'll probably want a fancy sound card with a digital mixer.
Fancy sound card options
The other options you have are to get a USB box, such as the Emagic 2|6, or to get a PCI card, such as the RME Hammerfall line or the popular Delta line. Musician's Friend usually has a solid and varied selection of computer recording gear. All of these PCI cards have ADAT plugs, take in 24-bit 48kHz sound (minimum, some of them will take 96kHz), and all three brands are supported by ALSA, the sound system for Linux. In general I suggest you steer clear of USB devices because of the latency you'll experience during live recording and playback, but they do have benefits -- they get your A/D converter out of your case and offer limited multi-track support. By purchasing a good quality USB digital mixer with ADAT connectors, which use fiber optic cables, you won't have to deal with interference in the connection. USB sound devices are pushing towards inexpensive and can be a good low budget solution. The Hammerfalls and Deltas both offer extensive multi-track support, and the Hammerfalls especially expand up to lots of tracks, ranging from 32 to 64. My preference is the Hammerfall line, because of the ADAT connectors on it and the reputation behind it. I've never used the Delta line. For pricing, the Deltas are marketed as budget high-end sound cards, and the Hammerfalls are marketed as "better" high-end sound cards.
For a home recording studio, you're probably going to be just fine with the full-duplex sound card that came with your computer, along with at least a 2.4 series Linux kernel and full-duplex sound card drivers. If your drivers aren't full-duplex, then you'll need ALSA. If you have a 2.6 series kernel or a distribution like Mandrakelinux or SUSE, then chances are you're already using ALSA. If you don't have ALSA by now, you should get it. It doesn't cost you anything and ALSA is on its way to setting the standards of excellence for computing sound systems, and this will indirectly affect the quality of your own work.
Once you've installed the hardware and software you need, you're ready to begin:
- Using Ecasound, record the first desired track (you might want to use a metronome to make this a click track).
- Using Ecasound, record the next desired track while playing the previous tracks recorded.
- Continue using Ecasound to record all of your tracks.
- Import all the tracks into Audacity.
- Master the tracks with Audacity and output to a stereo WAV file.
We'll walk through the process next time.