The 702, 702T, 722, and 744T recorders can record to several different file formats to extend record time and reduce file size. A review of the options are discussed in this article.
Uncompressed WAV files, especially high bit rate, high sampling rate files, can be large and unweildy. While computer storage systems have kept pace with the ever-growing number and size of sound files, file size reduction is still a necessity. In addition to recording uncompressed WAV files, the 702, 702T, 722, and 744T recorders can record to two "lossy" file formats, MP2 and MP3, and one data compressed format, FLAC.
Data Rate Calculation - WAV Files
PCM audio is represented as a data rate. For example, the common production-sound recording rate of 24-bit, 48 kHz can be expressed as a data rate of 1125 kb/s per channel. The calculation is as follows:
- 24 bits X 48000 samples per second = 1,152,000 bits per second
- 1,152,000 bits per second / 1024 bits per kilobit = 1125 kb/s per second per channel.
That 1,125 kb/s per channel is the uncompressed data rate. When we look at the storage required to hold audio, we multiply the data rate by the amount of time needed to record. For instance:
- 1125 kb/s X two channels = 2250 kb/s for two channels
- 2250 kb/s X 3600 seconds (60 minutes) = 8,2000,000 kb
- 8,100,000 kb = 988 MB (mega bytes)
Therefore, 988 MB is required to store one hour of two-channel, 24-bit 48 kHz uncompressed audio.
Data Reduction Versus Data Compression
If data storage were infinitely fast and limitless in capacity, there would be no need to lower the data rate, and consequently reduce the storage space needed for a given audio program. Since there are always limits, ways to reduce data are important to evaluate. Two common methods are employed to reduce the amount of storage space required for audio, lossy data reduction and lossless data compression.
MP2 and MP3 - Lossy Audio Data Reduction
Audio data reduction is a process where a perceptual encoder analyzes audio data (during recording) and removes bits from the audio data based on a specified target quality level or data rate (for instance, 128 kb/s). The bits removed are deleted and cannot be restored during the decoding (playback) process. This is sometimes referred to as a lossy codec. MP3, AAC, WMA, and ATRAC are the most common audio data reducing codecs available, although there are numerous others. The upside of these codecs is that very high rates of data reduction can be obtained. Additionally, with higher data rates (>160 kb/s) audio quality is quite good when done by a high quality codec. The downside is that the original audio data is altered, sometimes substantially. If MP3 recording is done in the field, there is no way to restore the audio data that was discarded during the encoding process, leaving the original material compromised.
Another factor in lossy codecs like MP3 is that high bit and sampling rates are not supported. MP3, for instance, has a maximum sampling rate of 48 kHz. Higher sampling rates input to an MP3 encoder are bandwidth-limited by the encoder. There is no doubt that there is a place for lossy encoding. The question for the production sound engineer is whether that place is on the original material.
Data compression is a common process used to reduce file size of digital information without altering the content of the original. Computer file archives such as .ZIP an .SIT compress data using a variety of strategies. The amount of data compression varies according to the type of information that is being compressed. Some file types themselves are compressed, therefore data compression can do little to compress them. Again, with lossless data compression, the decoded data is identical to the original. That makes lossless compression suitable for original recordings as well as for archives.
FLAC - Lossless Data Compression
FLAC is a convenient way to extend recording time with no effect on audio quality. It is suitable for recording and archiving. FLAC is in the class of lossless compression algorithms similar to Apple Lossless and WMA Lossless audio codecs. These differ from other lossless algorithms, such as ZIP because they are specifically designed to compress audio data. FLAC, in particular, offers data compression rates that range anywhere between 30% and 60%, depending on the audio program.
FLAC encoding is suitable for use on original program material and for archiving purposes. Because there is no lossy perceptual encoding, no audio material is destroyed, leaving a perfect copy of the PCM audio file. To learn more about the specific FLAC file format capabilities of the 702, 702T, 722, and 744T recorders see Recording to FLAC File Format with 7-Series Recorders.