|
Table of contents
In principle, a recording device must be able to capture the frequency response (<10,000 Hz) and dynamic range (40bB) of speech. However, there are other important considerations while choosing a field recording device. Below is an overview of some of the field recorders available today.
Analog recorders
The analog tape is able to provide the necessary frequency response and dynamic range to capture a fair amount of detail in the speech signal. In fact, if a premium-grade microphone and pre-amplifier have been used, a professional cassette recorder, such as Marantz PMD 222 [PHOTO], or a reel-to-reel recorder, such as the NAGRA IV [PHOTO] are capable of capturing fairly high-quality acoustic signals. Cassette recorders are usually quite durable and inexpensive. They also operate on regular alkaline batteries and use standard audio cassettes, which are inexpensive and widely available. However, analog recorders typically have noisy transport mechanisms and offer no time code, which makes logging and analysis more difficult. In addition, the analog tape is a rather fragile medium and should be avoided by researchers interested in long-term preservation of their recordings. Perhaps, the most serious disadvantage of analog recordings is that they have to be digitized before any computer-assisted analysis can be performed. Even though, the audio digitization process can offer an extremely high quality of transfer, it is often done poorly, which results in a considerable degradation of the original material.
Digital recorders
Digital recorders are becoming more and more common. They come in a variety of flavors, such as DAT, minidisk, solid state, CD/DVD-R, and hard disk. What distinguishes them from one another are the recording medium and the recording format. For the purposes of acoustic analysis, one should use digital recorders that capture sound in an uncompressed, PCM (pulse code modulation) format.
The most typical digitization settings (so-called CD quality) of a 44,100 Hz sample rate and a 16-bit bit-depth render a recording that has a broad frequency response (about 21,000 Hz) and an impressive dynamic range of 96 dB. Many modern ADCs are capable of sampling at the rate of 96 kHz and a 24-bit resolution, which produces a highly accurate signal with negligible quantization noise. After the signal has been quantized the values of each sample have to be stored on the storage medium that the recording device happens to be using.
DAT recorders
DAT recorders use magnetic tape as a storage medium. Typical DAT tapes allow the storage of 120 minutes of uncompressed, high-quality mono recording at 48,000 Hz/16-bit. TASCAM DA-P1 (PHOTO>>) offers a rugged construction, and a pair of good pre-amplifiers. It also comes with a built-in limiter and microphone/line inputs. It features the S/PIDF digital I/O interface. The migration of digital audio from DAT to a personal computer hard drive for analysis is a lossless process, provided the correct digital transfer interface is used. The DAT tape is as fragile as a regular analog cassette tape, which, again, raises the issue of long-term preservation. Most digital-born recordings stored on DAT tapes will eventually have to be migrated to either spinning disk or optical storage.
MiniDisc recorders
Minidisk recorders are quite popular among field workers, particularly due to their small size and ease of use. However, most minidisk recorders do not produce the quality required for detailed acoustic analysis. First of all, most of the portable minidisk recorders lack a quality pre-amplifier, which makes it difficult to interface them with premium-grade microphones. Minidisk field recordings are usually shipped with small, inexpensive lavalier condenser microphones whose frequency response and dynamic range are typically rather poor. In addition, there is virtually no control over the incoming signal. The A/D converter is inferior to that found on larger DAT or hard disk recorders.
Many portable minidisk recorders lack the S/PDIF I/O interface, which makes it necessary to use a standalone minidisk deck for digital audio transfer. Finally, to achieve a small disc diameter, MD uses data reduction based on psychoacoustic principles. Prior to storage, the audio data rate (bit rate) of 1.41 Mbps (uncompressed PCM) is compressed using a perceptual coder (proprietary algorithm developed by SONY) to lower the bit rate to 292 Kbps (1/5 of the original). ATRAC preserves the sample rate (44,100 Hz), but decreases the word length (resolution), which results in a less faithful representation of the analog original and quantization noise, which is then masked by the algorithm, yet, technically it is still there, though we can't hear it. A lower word length, naturally, results in decreased dynamic range, which, of course, the algorithm cannot compensate for. As a result, we have a recording that has the full 20-20000 Hz frequency response, but has a decreased resolution, increased quantization noise, and decreased dynamic range.
Solid state recorders
Marantz, Nagra, and other manufacturers have developed professional quality solid state (typically using PCMCIA or Compact Flash memory cards) filed recorders. They have impressive features, one of which is the ability to record uncompressed PCM 16-bit audio at sample rates of 44,100 Hz and higher. Initially, the prohibitively expensive memory cards made these recorders difficult to afford. However, the cost of memory cards is constantly going down, and one can expect the popularity of solid state recorders to climb substantially.
One of the most attractive features of solid state recorders is the speed and ease with which recorders are transferred to audio workstations. Since the recorders encode audio in standard audio file formats (e.g., MS wav), they can be easily and quickly copied to any computer equipped with a memory card reader.
Hard disk recorders
Hard disk recording is usually associated with expensive recording studio multi-track recording devices connected to SCSI hard disk farms and racks full of processing hardware. However, recently, small USB devices have appeared on the market. USB Pre is an impressive piece of hardware (PHOTO>>). It features two independent studio quality microphone pre-amps, phantom power, 24-bit A/D converters, 106 dB dynamic range and a variety of inputs and outputs (tape, line, instrument, microphone, and S/PDIF). The device is powered through a computer USB port and can be used with most modern PC and Macintosh machines, including a variety of laptops. Recordings obtained with USB Pre are superb. Whenever the recording situation allows the use of a laptop, USB Pre (or a similar device) should be used to acquire speech signals. At the time of writing this article, USB Pre is one of the best USB digital recording devices, though other manufacturers, such as Digidesign (the M-box), and M-audio have released their own recorders. Still, SoundDevices, the maker of USBPre, has a proven track record in field recording equipment, and the USBPre is likely to remain a favorite tool of phoneticians for quite some time.
|