Find out how to get the highest-quality audio into your computer music system.
The all-digital, computer-based studio may have finally come of age, but there will never be plug-ins that can substitute for mics, mic preamps and monitoring systems. Though some audio interfaces include mic preamps, most offer either line-level analogue or some form of digital input, which leaves the user with the problem of how to get the best-quality signal into the system. This article looks at some of the options, with a view to pointing out their weaknesses as well as their strengths.
Budget soundcards intended primarily for the games market often include both mic and line inputs, but while the line input may offer adequate quality, the mic input is usually designed to accept a low-cost electret microphone, often bundled with the card. These work rather better than they should for the price, but still don't approach the quality or sensitivity needed for serious vocal or instrumental recording. Even if the preamp is sensitive enough, it is extremely unlikely to provide the phantom-powered balanced input required by most capacitor microphones.
The only practical option with such cards is to feed them from a mixer, a separate mic preamp or a voice channel. Because the converters used in budget cards are not always of the highest quality (or, as is more often the case, are compromised by poor circuit-board layout), the best results are usually achieved by using the card's digital input. This means choosing a mixer, preamp or voice channel with a digital output. Such a device is likely to cost rather more than your budget soundcard, but at least it will still be useful when you eventually come to upgrade your soundcard or switch to a hardware audio interface.
Hardware external interfaces, such as the type that connect to the computer via a cable and PCI card or a FireWire cable, effectively remove the converters from the noisy environment of the computer case, so they often have a better noise spec than all-in-one cards. Don't let this put you off a card interface if that's the way you'd prefer to work, as some perform extremely well, but, as a generalisation, external hardware is better, and if you need multiple channels of I/O, it gives the manufacturer somewhere to hang the sockets!
If the interface has only line inputs, then you'll have the same problems as you would with a soundcard if you want to record using microphones, but some of the current crop of interfaces, such as the Digidesign Digi 001 and the MOTU 828, have built-in mic preamps, though usually only on two channels. Naturally, mic preamps are not all created equal and there have been reports that some of the mic amps found on audio interfaces aren't as sensitive as those on a good mixer. A typical mixer has around 60dB of mic gain with around a further 10dB of gain available on the channel faders. A quoted mic gain in excess of around 58dB should be fine for most applications, but anything less may give you trouble if you're recording quiet or distant sources, especially if you use dynamic rather than capacitor mics. If you end up with such an interface, you may still want to buy a better or more sensitive mic preamp, voice channel or mixer.
The traditional way of recording is via an analogue mixer, and if you need to record several mic channels in one go, it could still be the cheapest option. Some of the smaller mixers around, such as the Mackie VLZpro series, have exceptionally good mic preamps, so provided that you can get enough separate outputs (via direct outs, insert sends, buss outs, pre-fade aux outs and so on) this could be a good solution. As most computer-based studios need to mix MIDI instruments and digital audio at some stage, the chances are you'll have, or soon need, a mixer anyway. But because most analogue mixers don't have digital outs (there are one or two now that do), your audio quality will also be influenced by the quality of your soundcard/interface converters.
Digital mixers invariably come with S/PDIF stereo outs as standard and they may also include expansion slots for ADAT and TDIF multi-channel digital interface cards. By pairing one of the latter with a soundcard that has a compatible multi-channel digital input, you can pipe audio into the computer without having to worry about the soundcard converter quality.
I've found the quality of mic preamps on digital mixers to be extremely variable and some also seem to be a bit short on mic gain, so it's important you get to try the mixer you're thinking of buying with a microphone and a realistic sound source in a quiet environment before you buy. If you don't have any need for digital I/O, then a budget digital mixer may not be the best choice for you, as the signal path through a well-designed analogue mixer is often better when you measure from mic input to analogue output.
Electric guitars may be DI'd into computer systems, but it's not really a great idea to plug them directly into soundcard line inputs, as the impedance is wrong. Guitars like to 'see' an impedance of around 1MΩ or higher. Either use a suitable active DI box with a high-impedance input or the instrument input of a voice channel. There are numerous software guitar preamps that can be used to shape your tone once the audio is inside the computer, but if you want to hear the effect of these as you play, without a distracting delay, you'll need a system with very low latency.
An alternative is to use a hardware guitar preamp. Numerous models are available covering all proclivities from tube to solid-state to digital physical modelling, but you need to choose a unit that also incorporates speaker simulation, otherwise the guitar sound probably won't turn out the way you want it to. One tip if you are using a hardware preamp with integral effects is to switch off its delay and reverb and instead add these effects using plug-ins. Not only does this greatly simplify editing (you won't be chopping off reverb tails or delays), it also means you can record onto a mono track to save space. You can still process the sound using stereo plug-ins, which can be automated if you wish, and there's no reason not to combine a hardware preamp with software modelling if you need to treat the recorded sound further.
Mic preamps are now available in single, dual and multi-channel configurations and they are ideal for feeding into soundcards or audio interfaces that have good-quality analogue inputs. You have a choice of solid-state or tube models, if you prefer the sound of one over the other, and any respectable model will give you all the gain and sensitivity you need, along with a low noise floor. You could also use a separate mic preamp with a digital mixer that only has unremarkable mic preamps of its own by feeding the preamp into the mixer's line inputs.
Multi-channel mic preamps (usually designed in blocks of eight) can be financially viable for recording multiple instruments, but if you only tend to record one or two parts at a time, then consider using one of the available voice channels instead, as these have distinct advantages.
Voice channels are mic preamps that also include compression, some form of EQ, and often additional features such as de-essing, expansion, gating and limiting. They may also include a monitor section (possibly with a headphone amp) to enable the performer to monitor the source signal mixed with the output from a computer soundcard, so as to avoid the distraction of latency when trying to sing or play.
Other than this zero latency monitoring workaround, which you can also rig up using just about any mic preamp/voice channel and a mixer, it seems that most software packages already include counterparts to the processing sections of a voice channel, but having a hardware compressor offers a very real advantage. One of the benefits of analogue compression is that you can work closer to the maximum signal level of your system's converters with less risk of clipping, and even lower level signals will be boosted in level, so they get recorded using more bits. More bits means less distortion and a better signal-to-noise ratio, so compressing your signal in the analogue domain prior to conversion tends to produce better results than applying digital compression after conversion. A good analogue EQ is also hard to beat, so if you can get a really classy sound at source using the controls on a voice channel, you're more likely to end up with the sound you want when you come to mix.
A good voice channel will include mic, line and high-impedance instrument inputs, and a number of models either include a digital output as standard or offer it as a plug-in option. On balance then, a voice channel with a digital output provides the most elegant means of getting a quality signal into your system (provided that your soundcard has a digital input), but you can only use one or two channels with a stereo converter — a word-clock distribution system would be needed to run multiple units (assuming you have multiple digital inputs to connect them to, as is the case with some Pro Tools interfaces). Perhaps more of a consideration is that good voice channels can be costly per channel compared to a mixer or multi-channel mic preamp, so it may be less attractive to go this route if you want to run a dozen mics or more at once. However, if you do tend to overdub a single part at a time, a high-quality voice channel is easily the best solution.
A signal spoiled at source can never be satisfactorily recovered, so choosing the right mic and preamp system for recording probably makes more difference than just about anything else in the recording chain. As most musicians who work in their own studios tend to record relatively few parts at the same time, using high-quality voice channels is probably the best solution, and if these can be digitally coupled to the input of the computer system, so much the better. Some voice channels also offer optional converters which support 24-bit conversion up to 96kHz if you feel you need to take that step.
For multi-mic recording, a good analogue mixer or a multi-channel mic preamp is probably the right option and I wouldn't suggest using a low-cost digital console for tracking unless you also need one for mixing. Even then, you might still feel the need to buy a good voice channel for recording important parts. Whichever option you go for, by taking care with the sound source, you're much more likely to end up with a recorded sound quality that you're happy with.
Although the main thrust of this article is related to getting high-quality live audio into your sequencer, rhythm loops and other samples can be imported from CD or CD-ROM and then placed into your audio tracks. The exact procedure varies depending on the sequencer you use, and whether you use a Mac or a PC, but material from audio CDs usually has to be ripped first using an audio extraction utility. On my Mac system, I find Roxio's Toast Audio Extractor is about the easiest to use, though the number of alternatives grows daily. Mac users will probably want to save their files as AIFF or SDII, whereas PC users are probably safest choosing WAV.
Once the audio track has been ripped, it can be imported into the audio library of the song you're working on and from there transferred to the arrangement just like any other piece of recorded audio. If you only need part of a track, you can use the waveform editor of your sequencer to trim it to length. The exact procedure is different for each sequencer, but the general outline is roughly similar — in Logic, audio files that have been converted to WAV, AIFF or SDII, or any CD-ROM samples already in that format, may be dragged directly into the Audio window, and SDII files may even be dragged directly into the Arrange window.
Some Mac sequencers can only use drag and drop with SDII files, whereas AIFF files need to be imported using the Add File menu. Also note that SDII files come in both interleaved and non-interleaved stereo formats. In most instances, an extra procedure has to be followed in order to split interleaved files where it is required to use the left and right components independently — for example, if you have received a CD comprising a vocal on one track and a guide backing part on the other.