More hints and tips revealed as the expert teaching staff of London's Gateway School of Recording & Music Technology answer your queries.
A regular column where members of the teaching staff of the Gateway School of Recording & Music Technology answer your questions.
Q: What exactly does the term 'sampling frequency' mean, and how does it affect the sound quality?
Godric Wilkie of Gozen Studios, who teaches the sampling course at Gateway, answers:
A: The best way to answer this is to describe the process of sampling itself. All samplers are digital recording devices. They differ from their analogue counterparts, tape recorders, in the way the information (the sampled sound) is recorded.
A tape recorder makes a copy of the sound on tape in the form of a continuously varying magnetic field which represents the sound. A sampler makes a copy of the sound in its memory in the form of discrete binary digits (numbers). These numbers are a series of measurements of the height (amplitude/volume) of the sound waveform from moment to moment. The height measurement is performed by an Analogue-to-Digital Convertor (ADC) - the waveform's height being an analogue quantity, the number stored in memory being digital.
As the number only represents the height of the waveform at one instant in time, many thousands of measurements have to be taken to re-create the entire waveform. And the more measurements that are made during the life of the sound, the more accurate the recording of that sound will be.
Put simply, the sampling frequency (or sampling rate as it may be called) is the number of times the sound's waveform is measured in a given time period. This time period is usually one second, and thus the sampling rate is usually expressed in 'kHz' (1 Hz = 1 cycle per second).
A general rule is observed about sampling rates (often called the Nyquist rule), that the sampling rate should (at worst) be set to twice the maximum frequency you wish to hear played back. Thus, if you want to hear all the information in the human audio spectrum (20Hz-20kHz), the sampling frequency should be no less than 40kHz. The sampling frequency chosen for Compact Disc has, therefore, been set at 44.1kHz, giving a 10% margin for error.
Remember: the lower the sampling rate, the lower the maximum frequency you will get out of your sampler. This includes harmonics. Just because a bass drum's fundamental note is tuned to 90Hz (say) doesn't mean that you will get a decent sounding bass drum sample if the sampling rate is 180Hz. Bass drums generate harmonics (multiples of the fundamental note) which reach well into the 10kHz region, and these harmonics are essential to the authentic reproduction of the bass drum sound. So, if in doubt use the highest sampling frequency you possibly can.
Q: Do you think compressing a sound before recording it into a sampler helps to achieve a better quality sample?
Godric Wilkie of Gozen Studios, who teaches the sampling course at Gateway, replies:
A: A tricky one! Really this is a matter of 'horses-for-courses' and experimentation. Compression is a method of limiting the dynamic range of a sound, ie. the difference between the quietest and loudest parts of a sound. It works by reducing (ie. compressing) the volume of the sound above a certain threshold. It, therefore, ensures that a sound cannot exceed a pre-determined volume.
This is very useful in recording where you want to ensure a good average volume level. The unfortunate side-effect is that as the dynamic range is reduced, so is the signal-to-noise (S/N) ratio. This means that although your sound is at a fairly constant level, so is any accompanying noise!
Avoiding noise in samplers is crucial, so you might think that anything that makes your sound more noisy before you sample it is best avoided.
However, early samplers which used 8-bit Analogue-to-Digital Convertors (ADCs) had a very limited dynamic range of about 48dB. As 'real' sounds might have a dynamic range of 70-80dB it was useful to compress them, if only to avoid overloading the sampler's input and causing distortion.
Modern 16-bit samplers like the Fairlight Series III and forthcoming Casio FZ1 have a dynamic range between 90-96dB and, therefore, compression is no longer strictly necessary, but some sounds such as percussion benefit from the creative aspects of compression (the general tightening up of the sound and envelope shaping) and these will generally sample better if compression is applied sensibly and creatively.
Sounds of constant volume, such as wind or bowed string instruments, vocal 'aaghs' etc, do not generally benefit from compression and indeed can suffer, from the noise point of view, if compressed.
A further complication is that some ADCs tend to compress a sound anyway, without you knowing it. Early 8-bit ADC systems, for instance, sometimes used a logarithmic law to govern the way in which the analogue signal was digitally converted. The audible effect of this was compression.
Today's 12-bit and 16-bit ADCs tend to be linear.
So, like any audio processing, compression can be useful when used in the right amounts in the right places.
(PS. Just for the record, the theoretical dynamic range of the different bit-resolution samplers that use a linear ADC is as follows: 8-bit = 48dB, 12-bit = 72dB, 14-bit = 84dB, 16-bit = 96dB.)
Q: I notice that the new DX7II and the TX81Z both off er the facility of using 'microtonal scales'. What are they, and are they likely to be of any real use in my music?
Errollyn Wallen of Wallen Parr Music & Production answers:
A: Basically, microtones are any musical intervals smaller than a semitone, such as quarter tones and sixth tones. Musicians other than keyboard players have always intentionally used microtones - and certainly pianists unintentionally have, when their instrument is 'out of tune'.
A violinist will play, for example, an F sharp note slightly sharper than a G flat, likewise singers and woodwind players, but keyboard players have been mostly restricted to the fixed 12-note (equal-tempered) scale. I say 'mostly' because way back in the 1920s some composers were working with pairs of pianos tuned a quarter tone apart - and the unfortunately named Fokker company had a 31-note organ... honest!!
We must remember that as far back as the 1890s, microtones were being used on their own account rather than just for subtlety of tuning (as with our example violinist). Certainly if Harry Partch, who built his own percussion and wind instruments with a 43-notes per octave scale, were alive today, he'd be having enormous fun with the new DX7II and TX81Z.
Practically speaking, the inclusion of microtonal scales on these new Yamaha instruments also allows far greater flexibility and authenticity when playing some of the non-Western instrument sounds (eg. sitar), since they need not be tied to equal-tempered tuning.
So, keyboard players and composers alike-think 'out of tune' and explore those cracks between the black notes!
Further info about the extensive range of Gateway audio courses can be obtained by sending a SAE to: (Contact Details)
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!