DIY Fret Work - Setting Up Your Guitar
Maintenance / Repair / Modification
(STG Feb 97)
Magazine Archive
Home -> Magazines -> Issues -> Articles in this issue -> View
A Matter of Interpretation | |
Understanding SpecificationsArticle from Sound On Stage, April 1997 | |
Hugh Robjohns gets boxed in by facts and figures.
Music-making increasingly relies on vast arrays of technical equipment, and whether we are talking about kit for a live performance or a demo recording, the question is how do you select the most appropriate equipment from the enormous range on offer? Reviews in magazines such as this are obviously a good place to start, as are personal recommendations from existing users, but in the end, it comes down to three things: technical specifications, preferences about operational ergonomics, and what it actually sounds like. I can't help with the last two, but I can shed some light on the first, which will hopefully allow you to make some sense out of the collection of figures printed on the back page of equipment handbooks, and at least sort the wheat from the chaff!
The technical specifications provide information about what the machine does, how well it does it, and under what circumstances. With careful interpretation, the specifications can be used to assess suitability for the task in hand — both in terms of what it does, and where it can do it — and highlight any qualitative differences between similar products. However, there are many traps to catch the unwary: different assessment techniques, reference points, and even measuring equipment can have a profound affect on the figures. It is important to compare 'like for like' if meaningful conclusions are to be drawn from the figures alone.
The good news is that, by and large, modern electronics have reached a stage where it is actually quite difficult to design a duff product. For example, interfacing between various pieces of equipment is rarely an obstacle these days, and even the most basic designs achieve very respectable noise and distortion performance. However, there are still a few rogue products about, but their specifications give the game away if you know what to look for.
The kinds of specifications typically provided for audio equipment can be broken down into six areas: interface impedances and levels, frequency response, dynamic range, noise and distortion, crosstalk, and timing problems.
TECHNICAL SPECIFICATION
| Input impedance: | 10kOhm |
| Output impedance: | 600 Ohms |
| Maximum input level: | +8dBu |
| Maximum output level: | +8dBu |
| Frequency response: | 20Hz to 20kHz (-3dB) |
| Signal to noise ratio: | 85dB |
| Headroom: | +24dB |
| Mic input EIN: | -125dBu EIN |
| Output Noise: | 90dBu |
| Distortion: | less than 0.05% THD |
| Crosstalk: | 90dBu |
IMPEDANCE
At one time, the professional standard was to use 600 Ohm send and receive impedances, but these days, the norm is to send from as low a resistance as possible (typically 100 Ohms or less), and receive with a relatively high impedance, say 10,000 Ohms. This arrangement allows the signal source to provide plenty of 'oomph' to drive long connecting leads, and the receiving input presents an undemanding load.
However, there are some circumstances where a much higher input impedance is desirable, and the most obvious one is for instrument inputs, such as those designed for electric guitars. A guitar pick-up is not able to supply much of a signal, and so a high input impedance avoids loading it too much. Typical figures here might be of the order of 500 kOhms or more, but the danger is that the high impedance will make the noise performance worse.
However, there are some circumstances where a much higher input impedance is desirable, and the most obvious one is for instrument inputs, such as those designed for electric guitars. A guitar pick-up is not able to supply much of a signal, and so a high input impedance avoids loading it too much. Typical figures here might be of the order of 500 kOhms or more, but the danger is that the high impedance will make the noise performance worse.
INTERFACE LEVELS
Signal levels are normally specified in terms of decibels (dB) with a suffix to indicate the particular reference point used. In days of old, a term derived from the telephone industry was used, the dBm. Zero dBm was the voltage required to dissipate 1mW of power in a 600 Ohm resistor, which works out as 0.775V. Today, the professional audio industry avoids 600 Ohm send and receive impedances, but retains the same signal voltage as the reference — this modified version of the old standard uses the term dBu — so 'Zero Level' is 0dBu, and still 0.775 Volts. This corresponds to a reading of 4 on a broadcast standard Peak Programme Meter, or -4dB on a VU meter (ie. 0VU equals +4dBu). This relationship between dBu and VU meters is the basis for the nominal standard professional interface level of +4dBu (often misquoted as +4dBm).
Semi-professional equipment has long since standardised on a rather lower reference level of -10dBV (ie. 10dB below 1 Volt), which works out as roughly 300mV. In terms of dBus, there is about 8dB difference between the professional and semi-pro reference levels — normally well within the range of an input level control.
Semi-professional equipment has long since standardised on a rather lower reference level of -10dBV (ie. 10dB below 1 Volt), which works out as roughly 300mV. In terms of dBus, there is about 8dB difference between the professional and semi-pro reference levels — normally well within the range of an input level control.
FREQUENCY RESPONSE
The frequency response of a device is its ability to handle a range of sounds from the lowest notes to the highest, at a near-constant level. There is still great debate about the range of frequencies that the average person can perceive, but the usual figures quoted are 20Hz to 20kHz.
We generally tend to 'feel' rather than 'hear' sounds below about 30Hz. it is not uncommon to find equipment which can pass signals as low as 5Hz or so, especially digital systems — although finding a monitoring or PA system to replay these frequencies is a little more challenging!
At the high end of the frequency range, there are few people, who can reliably 'hear' signals in the region of 20kHz, and it is natural for the human hearing system to lose sensitivity to high frequencies with increasing age. However, there is some evidence to indicate that we may be able to perceive such signals by other means (as yet unknown), and this has encouraged some manufacturers to design equipment with frequency responses extending well up into the 50 or even 100kHz regions (although there are also a number of technical reasons why this design approach can be beneficial too).
For a frequency response to be meaningful, we need two things: the upper and lower frequency extremes and the tolerance limits indicating how uniform the signal level can be maintained between the frequency extremes (eg. +1dB). Where no tolerances are given, the general assumption is that the frequency extremes relate to the 'half-power' points where the signal level has fallen by 3dB relative to some reference frequency (normally 1kHz), although the 'flatness' of the response can only be guessed at...
We generally tend to 'feel' rather than 'hear' sounds below about 30Hz. it is not uncommon to find equipment which can pass signals as low as 5Hz or so, especially digital systems — although finding a monitoring or PA system to replay these frequencies is a little more challenging!
At the high end of the frequency range, there are few people, who can reliably 'hear' signals in the region of 20kHz, and it is natural for the human hearing system to lose sensitivity to high frequencies with increasing age. However, there is some evidence to indicate that we may be able to perceive such signals by other means (as yet unknown), and this has encouraged some manufacturers to design equipment with frequency responses extending well up into the 50 or even 100kHz regions (although there are also a number of technical reasons why this design approach can be beneficial too).
For a frequency response to be meaningful, we need two things: the upper and lower frequency extremes and the tolerance limits indicating how uniform the signal level can be maintained between the frequency extremes (eg. +1dB). Where no tolerances are given, the general assumption is that the frequency extremes relate to the 'half-power' points where the signal level has fallen by 3dB relative to some reference frequency (normally 1kHz), although the 'flatness' of the response can only be guessed at...
NOISE AND DISTORTION
Distortion is any unwanted change in the audio waveform, but in the context of equipment specifications, the figures usually refer to the total amount of harmonic distortion (ie. the creation of unwanted harmonic frequencies), or intermodulation distortion (the creation of sum and difference frequencies from two unrelated signals). Intermodulation is usually associated with power amplifiers and is definitely not nice — so the lower the figure, the better! Values below 1% are desirable. Harmonic distortion is normally unwanted too, and figures below 0.01% are normal with modern electronics. However, harmonic distortion often provides the sonic 'character' of a device, especially in valve-based equalisers, compressors, or microphone preamplifiers.
DYNAMIC RANGE
Dynamic range is the difference in level between the loudest and quietest things that can pass through the equipment. Very quiet signals may well remain audible below the noise floor of the equipment, and the highest levels will be determined by the available headroom (professional equipment can usually sustain signals up to +24dBu).
The dynamic range of real life is potentially enormous, and our own auditory system can cope with something like 140dB. However, there are few electronic systems that can match this; normally the dynamic range has to be reduced to something more manageable. Modern equipment should be able to achieve a dynamic range well in excess of 80dB, without too much trouble (although analogue tape recorders will usually be struggling to make 60dB without noise reduction). A 16-bit digital system should achieve something in excess of 90dB.
The dynamic range of real life is potentially enormous, and our own auditory system can cope with something like 140dB. However, there are few electronic systems that can match this; normally the dynamic range has to be reduced to something more manageable. Modern equipment should be able to achieve a dynamic range well in excess of 80dB, without too much trouble (although analogue tape recorders will usually be struggling to make 60dB without noise reduction). A 16-bit digital system should achieve something in excess of 90dB.
MIC INPUT NOISE
There are two common noise performance specs: microphone input stages are typically specified at around -129dBu EIN, and output stages are usually in the order of -90dBu. The 'EIN' after the input noise figure means 'Equivalent Input Noise'. It is derived by connecting a low value resistor in place of the microphone, and measuring the residual noise. The total voltage gain (in dBs) of the mic stage is then added to this to give the EIN noise figure.
The catch is that the value of the source resistor can have very significant affects on the EIN and not all manufacturers bother to indicate the value used! EIN is usually measured with 200 or 150 Ohm source resistors, since these are typical values for real microphones, but higher values will produce worse noise figures (some manufacturers use 600 Ohms), and a short circuit will produce spectacularly good results!
The catch is that the value of the source resistor can have very significant affects on the EIN and not all manufacturers bother to indicate the value used! EIN is usually measured with 200 or 150 Ohm source resistors, since these are typical values for real microphones, but higher values will produce worse noise figures (some manufacturers use 600 Ohms), and a short circuit will produce spectacularly good results!
OUTPUT NOISE
Output noise is usually measured with respect to the system's standard operating level, and figures around -90dBu should be achievable. In the case of sound desks, an important factor in the overall noise figure is how many channels are contributing to it? Some manufacturers quote with a single channel/group/master structure; others with all channels routed to the output. Some measure the noise with the channel faders open and others with them closed, so straight comparisons can often be rather difficult!
NO WEIGHTING!
One other point to watch for: noise is often measured with a special equaliser in circuit. Since noise is quiet, and as the frequency response of human hearing at low levels is far from flat, it can be argued that the noise measurement should take into account the ear's natural lack of sensitivity to low and high frequencies, and better acuity in the baby-squealing ranges, thereby representing its nuisance value! Thus, 'A-weighted' noise figures are often quoted, which will always be significantly better than measurements made under flat (unweighted) conditions.
CROSSTALK
In multi-channel systems, such as sound desks and multitrack recorders, crosstalk between channels is undesirable as it would reduce the usefulness of functions like over-dubbing! In a purely stereo situation, crosstalk is much less of a problem, because the two channels are directly related to each other. Ideally, in discrete channel systems, crosstalk should be below the noise floor, so figures in the region of -90dBu are desirable. In the case of a simple stereo recorder, anything over -50dB would be acceptable, but often the figure is much better than this anyway.
More with this topic
 |
 Confessions - ...of an English kit builder. |
 Modifying The Mu-Tron Bi-Phase - For Control Voltage Interfaces |
 Workbench - Signal Processors - the saga continues |
 Sample & Hold Resurrection - what to do with your analog sample and hold once you've gone digital |
 How To Make An EGG Mod |
 Workbench |
 String Attention |
 Balancing Acts - Cable Wiring Tips (Part 1) |
 Polarity Unveiled |
 Resistance is Useless! - Impedance & Resistance Explained |
 4780 Sequencer Modification |
Browse by Topic:
Maintenance / Repair / Modification
Publisher: Sound On Stage - SOS Publications Ltd.
The contents of this magazine are re-published here with the kind permission of SOS Publications Ltd.
The current copyright owner/s of this content may differ from the originally published copyright notice.
More details on copyright ownership...
Sound On Stage - Apr 1997
Feature by Hugh Robjohns
Help Support The Things You Love
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!
Donations for August 2026
Issues donated this month: 0
New issues that have been donated or scanned for us this month.
Funds donated this month: £0.00
All donations and support are gratefully appreciated - thank you.
Magazines Needed - Can You Help?
Do you have any of these magazine issues?
If so, and you can donate, lend or scan them to help complete our archive, please get in touch via the Contribute page - thanks!
If you're enjoying the site, please consider supporting me to help build this archive...
...with a one time Donation, or a recurring Donation of just £2 a month. It really helps - thank you!
@mu_zines
RSS Feed
Blog