Welcome to ISCE

ISCE Featured articles

ISCE members receive a quarterly magazine which features technical articles, advice and industry news.   Here are a couple of examples of typical technical features found in the magazine. Joining ISCE will give you access to all future and many past articles on our website.

A guide to the wonders of the wiggly line

Joules Newell MInstSCE MIOA MAES

Part 1 – Measurements never lie – or do they?There is much magic and myth that can surround the process of system tuning and analysis, To many, it is the realm of the expert, yet so many times some of the so-called experts can be unaware of the true nature of what they are doing. Here we will touch upon the issues and pitfalls of measuring and aligning sound systems within the average acoustic space.

So what’s the big deal that makes it all so complicated?

If we take a calibrated voltmeter and place it upon a 12.003 volt supply we will absolutely expect it to read 12.003 volts. Similarly if we take a calibrated water pressure gauge and place it on a 20.5 PSI vessel we expect it to read 20.5 PSI. It seems only reasonable that, from this, many people expect us to take a calibrated acoustic analyser or sound level meter and have similar accuracy within the venue we wish to measure. Unfortunately, many other factors can interfere with our measurement and thus it is not quite such an accurate a process as the inexperienced person expects it to be. This has been the cause of much consternation amongst audio folk, more so as acoustic analysis has become more prevalent and moved into the hands of those with less understanding of what may be affecting their results, who expect a definitive answer from the equipment.
For the purposes of this article, we shall be concentrating on the most common of all circumstances, which is a system in a room, hall, or similar acoustic space. It is the circumstance in which we usually use the system, and the way in which the customer will interface with the system. There are substantial differences in free field, or anechoic testing environments, but this article will not enter into that realm as it does not represent in any way what we do in our day to day application of electro-acoustics.

What is a test rig?

Firstly we shall take a brief look at the test rig. There tends to be much fuss and brand snobbery when it comes to test equipment, and many misguided cases of “Well, mine costs more so the measurement must be better”. If the measurement depended solely upon the test equipment this may well be the case, but most of this fuss is unwarranted once we are out of the laboratory and under the constraints of real world environments; here the test rig accuracy is limited by so many other factors.There are a few basic rules of thumb when it comes to choosing a test device. The first criterion should be that what we are using is from a reputable source, a source that can at least be said to have implemented the calculus and algorithms in a manner that processes the raw data correctly, and has taken care to get the signal into the unit without alteration.

Figure 1 - Typical Collection of Kit.Figure 1 – Typical Collection of Kit.

One should beware, as there are quite a few pieces of software out there that the internet forum brigade are touting as gospel which, quite frankly, are poorly implemented and are unable to produce results with any degree of accuracy. It is best to stick to something that has a genuine specialised developer behind it that can support the product. There are a range of such professional suites of software ranging from iPhone/ iPod/ iPad apps (must use ‘pro’ interface), through various PC programs right up to dedicated hardware devices that cost as much as a house.For reasons I will expand on later, there is little benefit in extreme accuracy for our purposes. What is more important is clear, concise, and prompt display of the results.

A test microphone and interface should at least have somewhere near a flat frequency response, but few are truly flat, and of sufficiently low noise floor. What we usually find popular and reliable is a mid-priced laptop based software package, a simple USB audio interface and a flat response measurement microphone. The fewer adjustments that can be made in the signal chain, the better. New to the scene, but similarly featured Smartphone / PDA packages with their own external preamplifier and measurement microphone are also proving to be of excellent standard and value, but more importantly highly portable, so as to enable easy capture of more data. If we are to do noise level calculations then some form of reference calibrator is advised, just to verify that we are not working with a misaligned measurement tool.

If you are thinking in fractions of decibels then you need to head back to the lab. Real world environments rarely allow for such precise resolution for any measurement; there is so much acoustic interference and interaction going on that it is hopeless trying to nail down a measurement so tightly. In summary, any reasonable test rig will do, it needn’t be laboratory standard, even for the most important measurements. What is more important is ease of use, and clear display of live results. Sixteen bit resolution is more than adequate.

The measurement

The first thing we need to understand is that the measurement microphone capsule occupies an incredibly small point in space, and only that particular point in space. Any measurement we take will only be representative of that precise point in space and not a room as a whole. It is highly probable that what we have happening across the spectrum at that point in space is only happening at that point and nowhere else.Acoustic path variation

So how is this? Let us take the example of the most basic anomaly of the floor reflection. We have the direct path of the audio propagating from the loudspeaker and reaching the microphone, right on time, as we would expect, but a quantity of that spherically (or fractionally spherical) propagating wave will inevitably bump into the floor or any other hard surface. When it does this, it will reflect, following the fundamental laws of reflection and continue along its way. Unfortunately, having participated in this merry diversion along its way to the microphone, it will then arrive somewhat later than the rest of its friends who left at the same time but took the direct path. This late arrival will then be out of time (phase) with the direct arrival and will thus interfere with the amplitude measurement you are taking. We can find exactly what frequency/wavelength this late arrival will coincide with, by some basic triangulation maths.

If we are to move the microphone further from, or nearer to the source, these path lengths will change (see Figure 2). As this unwanted anomaly is a pure function of microphone position, and not the source, we find that by moving the microphone in space we can alter the frequency and intensity of the interaction we see on our analyser display, assuming that we have sufficient display resolution. Any multiple
source arrivals, be they from surface reflections or multiple loudspeaker sources, even from within one loudspeaker cabinet, will have similar effects upon what we see with our measurement system.

All of these anomalies are a function of arrival-time differences at the measurement point; they are purely a function of time and space, and are nothing to do with the tonal qualities of the source itself. Should we move to a different point in space we will again have a completely different set of arrival times, and possibly different reflection sources. What we are dealing with when measuring systems in real life spaces is a massively complex combination of delayed wavefront arrivals that are dependent upon a huge number of external factors such as atmospheric conditions and physical presences. In some poor spaces, these arrivals may get very dense and unpredictable. These anomalies alone make accurate system amplitude correction within an averagely-reflective space a complete minefield. Where we look at our once trusted analyser display, we realise now that what we are seeing is not in any way an accurate representation of the sound of our system within the room as a whole, but rather only at the few square millimetres that the microphone capsule occupies. If we are unlucky, it may even be that we have placed our microphone in such a position that what it is measuring is absolutely nothing to do with what the listener is going to experience. We may have unfortunately placed it within a minuscule point in space where we have a total cancellation, or massive summation.

There is a second issue when trying to resolve these problems with system tuning. As what we see is a function of acoustic signal path length variations (timing differences) it is not a tonal (amplitude) issue due to the source. In many ways, this prevents equalization from having any effect at all. If we had perfect cancellation at one frequency, no amount of increased input would correct for it; where the cancellation is more slight we find the amount of additional input is not proportional to the correction, this usually means we are severely damaging the response elsewhere. What we are attempting to do is correct a time-domain fault in the frequency domain.



Commissioning a sound system for G Live

Jon Raper MInstSCE, R K Sound Engineering

F or most of us the concept of commissioning a sound system brings thoughts of a slightly stressful experience where we try to convince a consultant that we have done a wonderful job. In July, I was part of a commissioning process that was a very different experience.
G Live is a multi-purpose hall for Guildford and R K Sound Engineering installed an acoustic control system (ACS) to the specification of Peter Mapp Associates. After all the initial commissioning and setting of presets, the day had come to try the hall with a 60-piece orchestra and an audience.

First job was to see how the hall acoustics changed with an orchestra on stage. An impulse response was taken with the natural acoustics and then at the acoustic pre-set most suitable for a symphony orchestra. Later these could be compared with the measurements taken in an empty hall. The orchestra then played some of their repertoire while various ears belonging to musicians, acousticians and management sampled the fare throughout the auditorium. So far so good.

While the orchestra disappeared for a short break the audience of about 400 filed into the lower section of the auditorium seating; the entire hall seats just over 1000. This enabled us to take some measurements to see the effect of the audience on the room acoustics. When the orchestra returned we could take some further measurements with both the audience and the orchestra in situ.

And then the music. We had a chance to hear the orchestra playing some Beethoven, Elgar’s Cello Concerto, Stravinsky’s Firebird, John William’s Star Wars and the ever popular, Pomp and Circumstance. We finished with a vocalist leading Land of Hope and Glory accompanied by the audience; very ‘Last Night of the Proms’.

The local authority are keen to use the hall for classical music as well as other popular entertainment and they already have a variety of major orchestras, brass bands and choirs booked for the coming season. Therefore it is important that the hall acoustics can deal with a wide variety of acoustic expectations. The ACS system with its pre-settable configuration augments the natural acoustics for a variety of performance requirements thus maximising the potential of the venue.

The comments from the conductor and orchestra were very complimentary about the assisted acoustics. Their verdict was reinforced by a show of hands from the audience. What was intended to be an open rehearsal by the orchestra had become an orchestral performance with additional witty introductions from the conductor. The audience’s enjoyment of the experience was made abundantly clear at the end.
System commissioning does not come any more rewarding than this.

If you like what you see, then why not join the ISCE?

Click here to learn more about ISCE membership