Tutorial: Selecting an Optimal Recording Format

This tutorial provides recommendations on how to select the optimal recording format and attributes in Total Recorder.

The quality of digital sound is determined by discrete parameters and by compression parameters. Discrete parameters are the sample rate, bit capacity, and the number of channels. Compression parameters are the compression format being used and the bit rate.

General recommendations for selecting a recording format

Sometimes the recording format is determined based on how the target file will be used.

Also, it is not feasible to choose a target sound format with a higher quality than the source sound format. This only leads to an increase in the size of the target file.

In other cases, the optimal recording format depends upon the type of recording (e.g. speech, music, etc.) and your own tastes and preferences. Below are some recommendations that you can consider when selecting an optimal format for your recording.

Considerations for how a target file will be used

To create an audio CD, usually you need CD-quality wav files (i.e. PCM 44.1KHz 16bit Stereo).

If your CD player supports any compressed formats (WMA, MP3 etc.) or you plan to transfer recorded files to a mobile phone or any other portable device, you need to note that these devices often have special requirements for the format of recorded audio. In this case, you need to refer to their documentation.

Considerations based on the audio source

  • Selecting the recording format for Sound board recording

If you record from some external device (including microphone, CD or LP player) you need to use Sound Board recording mode.

When recording from digital input lines, you need to match the recording format (sample rate) with the format of the sound source. Sometimes you will need an exact match of the source sample rate and recording sample rate. To determine the required sample frequency, use menu Options -> Settings, General, Audio device information.

It is not recommended that you use stereo recording format when recording from mono sources (e.g. microphone).

If you record output from other software (such as Windows Media Player, RealPlayer, or Internet player), you need to take into account the format of the source audio when selecting the optimal recording format. This requires that you know whether the sound is being reproduced in a compressed or an uncompressed format.

  • Selecting the audio format for Software recording mode

If you plan to record sound that is already compressed (for example, many Internet transmissions are broadcast in MP3 or WMA format) then it is recommended that you record in the same format as the sound is being reproduced with. If you are concerned about file size, you can select a sound format with a lower quality.

If you plan to record uncompressed sound (e.g. from an audio CD) then you can either record without conversion (if you want to preserve maximum quality) or compress it (if you want to reduce the file size). You can use either lossy or lossless compression.

If the format of the sound source has a higher quality than CD (e.g. DVD audio), you may want to record it in one of the high quality formats which are available only in the Professional, Video Pro and Developer Editions.

In Software mode, we recommend that you record without conversion, and then save the recording using various target formats. After listening to the recorded file, you can determine the desired sound format. After that, record from the same source and convert it to the determined sound format on the fly.

  • "On the fly" conversion

This term means that audio is being immediately reconverted into another format as it is being recorded in Software mode. To turn "on the fly" conversion on/off: click the Recording source and parameters button, click the Change... button in the Recording format group and select or clear the Convert audio format to the one specified below check box:

As previously noted, no conversion can increase the actual sound quality. If a player reproduces "8000 Hz 8 bit mono" sound, and you convert it to "44100 Hz 16 bit stereo", it will not sound any better than "8000 Hz 8 bit mono". In fact, it may sound even worse than the original.

If you want to record in a specific sound format, use "on the fly" conversion and explicitly indicate the desired format. For example, if you want to put your recording onto an audio CD, you need to select "CD quality" (PCM 44100 Hz 16 bit stereo).

If you want to create a file with an acceptable sound quality but minimize its size, you can try to record without conversion, and then save the recording using a different target format. After listening to the recorded file, you can determine the desired sound format. Record from the same source and convert it to the determined sound format. It is not feasible to increase the sample frequency or to make a stereo recording out of a mono recording. Also, it is not recommended you change the sample frequency. If you need to use a lower sample frequency to create a smaller file, we recommend reducing the frequency by 2 or 4 times. This will minimize the loss of quality.

Example. A file was created from a recording without format conversion and its format is "PCM 24000 Hz 16 bit mono". The recommended "Save As..." formats for this file and optimal recording formats for similar files are:

  • PCM 24000 Hz 16 bit mono (save format unchanged, no loss of quality)
  • MP3 32 kBit/s 24000 Hz mono
  • MP3 24 kBit/s 24000 Hz mono

It is not recommended you use the following formats:

  • PCM 44000 Hz 16 bit stereo (CD quality) - frequency increase, mono to stereo
  • MP3 128 kBit/s 44100 Hz stereo - frequency increase, mono to stereo
  • MP3 32 kBit/s 22050 Hz mono - frequency changing

One very important case where you should not use "on the fly" MP3 conversion is when burning CDs. Some software (e.g. Music Match) lets you burn CDs directly from MP3 files. However, it is not feasible to convert from another format to MP3 and then put it to CD because you can lose quality. Instead, you should convert it to PCM 44100 Hz 16 bit stereo.

Considerations based on the audio characteristics of the data being recorded

Another important factor that you need to consider when selecting a recording format are the characteristics of the audio that you plan to record. In general, music requires higher quality formats than speech recordings.

For quality recording of music, you should use a sample rate of 44100 Hz or higher. For speech recordings of dictaphone quality, it is possible to use 11025 Hz. For quality speech recordings or middle-quality recordings of music, you can use a sample rate of 22050 Hz.

If you record an uncompressed sound, you also need to select a bit capacity. For quality recording, a bit capacity would be 16 bit.

It is recommended that you use 8-bit format only for speech recordings of dictaphone quality since they have a small dynamic range and a high noise level.

The number of channels (e.g. mono or stereo) depends upon the sound source.

If you plan to record in a compressed format, you can compress speech recordings to a greater degree without losing quality compared with music recordings. For music recordings, only a 5-10 degree compression is recommended. Also, it is recommended that for compression of speech, you use a compression format designed specifically for speech (e.g. DSP TrueSpeech) rather than for music (e.g. MP3).

If you record from a source that reproduces high-quality sound (e.g. with a professional sound card), you may consider using a high-quality format to preserve the quality of the original sound.

You can use the following table to select the target sound format based on characteristics of sound source.

Characteristics of source sound
Format
Attributes
Megabytes per hour (approximate)
Low-quality voice recording
DSP TrueSpeech
8.0kHz,1 bit, mono
4
Low-quality Internet music
MP3
11.5kHz, 16kBit/s mono
4.5
High-quality voice recording
Lernout & Hauspie
8.0kHz,16 bit, mono
9
High-quality voice recording
WMA voice
20kBit/s, 22.05kHz, mono
9
Middle-quality Internet music
MP3
22.05kHz, 56kBit/s, stereo
25
Near high-quality recordings
MP3
44.1kHz, 128kBit/s, stereo
56
High-quality recordings
WMA lossless
VBR Quality 100, 44 kHz, 2 channel 16 bit
150
High-quality recordings
Flac lossless
96kHz, 16 bit, stereo
240
High-quality (CD quality) recordings
PCM
44.1kHz,16 bit, stereo
600
DVD Audio, Super Audio CD recordings
PCM
96kHz, 24 bit, stereo
(available for Professional, VideoPro and Developer Edition users only)
1978

High quality formats

Starting with version 5.2, Total Recorder Professional Edition supports high-quality formats. This includes high-quality PCM formats (up to 192kHz 24bit and float mono and stereo), high-quality FLAC formats (up to 192kHz 24bit mono and stereo), high-quality Windows Media Audio Lossless stereo formats (up to 96kHz 24bit), and the stereo formats of the Windows Media Audio Professional codec.

Lossless compression

A lossless codec compresses sound data which results in a smaller sound file that has exactly the same quality as the original file. Normally, a sound file compressed by a lossless codec has a larger size than that compressed by a non-lossless codec.
Lossless codecs differ in the speed and degree of compression and in the sound formats they support. You can use Total Recorder for lossless compression by encoding to Windows Media Audio 9 Lossless format or to FLAC format.

Note that it doesn't make sense to apply lossless compression to a file which was encoded using a non-lossless codec. This will most likely result in only a larger sound file.