We are often asked which of these two techniques proves best to optimise sound coverage across a room. As ever there are no fixed rules, as both suit different situations. But it is important to understand the differences to help you design the optimum system. This article has been written based around ‘typical’ speaker systems, to give you some insight into the ‘typical’ challenges. It is important to note that, through continuous development, the Hill M-Series now allows you to sidestep many of these restrictions.
So to get us started, first a brief run down of the advantages and disadvantages of a distributed / line array based system.
Good gain before feedback -
Even distribution of high frequencies -
Even distribution of suspended weight through a venue -
Adjustable horizontal dispersion -
Listener is close to sound source (High speech intelligibility) -
- Requires more complex processing
- Uneven distribution of low frequencies
- Lower total SPL
- Confused psychoacoustic audio image
- Additional rigging cost and longer installation time
Optimal psychoacoustic audio image -
Most efficient configuration -
Coherent wavefront -
Simple rigging -
Even coverage of all frequencies -
- Requires careful and accurate positioning
- High single point load
- Fixed horizontal dispersion
- Requires height to deliver best results
When looking at this in detail lets consider three aspects; sound quality, room dimensions and room acoustics.
When trying to attain the best possible sound quality, we find the best approach is to get as close as possible to recreating the original sound source and project it evenly across the seating area with minimal conflicting reflections and anomalies. It’s easy to do that when someone is sat in the ‘ideal seat’ with both line-array and distributed methods, so instead let's consider the worst affected seats.
For a line-array, this will tend to be the seat placed at the edge of the sound field; perhaps the furthest away, or far off to one side.
The line-array principle seeks to overcome this by using the natural ‘coupling’ of many speaker elements placed one above the other to form a wavefront that matches the room shape. This is very adjustable in the vertical field, and typically more limited in the horizontal field. This effectively means the room acoustics becomes the ‘deciding factor’ in attaining exceptional sound quality. For a distributed system the opposite is true. Instead of working with the rooms acoustics, an attempt is made to overcome the room. Many speakers are spread around the venue, closely positioned to the listeners. This proximity ensures that the direct sound from the speaker is more prominent than the reflected and affected sound. However the consequence of this is that there are now multiple sound sources. So even in the ideal acoustic environment it will always be possible to hear multiple sounds coming from different directions. This is especially true for low-to-mid frequencies, which are typically omnidirectional in a ‘point source’ speaker enclosure. The images below shows this in practice with the 1st and 3rd distributed speakers negatively contributing to the overall sound field for the listener.
In this image, we can see with a line array that the person or persons within 'ideal listening' location is receiving audio from one source, and in one direction. This is known as a coherent wavefront and allows multiple cabinets to combine their outputs for ideal scenario coverage.
In this example however, the source of the audio has been separated across multiple speakers to create a distributed system. If done incorrectly, a distributed system can cause confused wavefront which prevents the listener from receiving optimum audio.
So if we have a reasonable room acoustic, deploying a Line-Array based system will normally sound better. However, sometimes there are other reasons why a Line-Array is either not possible or it becomes impractical. Let’s explore these challenges through looking at the room in more detail. (the room is often the most significant factor).
There are two main parameters to consider, the first is the venue’s dimensions and the second is the venue’s natural acoustic.
A Line-Array requires a minimum height to deliver even coverage front to back across a room. The Line-Array ‘effect’ only happens when the total height of the array starts to become significant in relation to the wavelength that you are wanting to control, typically the midrange Line-Array effect starts to be beneﬁcial at 4+ enclosures in a single hang. Before that you have a point-source system. To achieve an even SPL with the Line-Array effect, more enclosures are added to the top of the array. This unfortunately increases the required ceiling height even further. With these two parameters brought together you will typically need a minimum height of 5m to deliver acceptable results, with further improvements becoming available as you increase the ceiling height. A typical optimum would be to position the lowest enclosure at a height of ~5m. So what do we do if the ceiling height is only 2.7m or even 2.4m? Well a Line-Array just won’t work, no one will be able to see, and the front row will be incredibly loud in comparison to the back! So in this scenario the distributed systems really wins. Occasionally we are faced with a room that has both properties (balcony, mezzanine, theatre etc.) in that case simply apply both techniques together. This can work really well, as long as the same technology can be employed in each area, because it is vital that every seat ‘sounds the same’ (see different systems in one venue).
Combining line array with distributed system to create an efficient wavefront.
If we have a poor to moderate room acoustic (especially a high RT60) it is common to see fully distributed systems, for example a Cathedral PA system. Is it possible to maintain the advantages of a Line-Array (coherent wave-front, simple rigging, excellent audio-imaging, etc) in this environment, and achieve the same level of intelligibility? The key to achieving success is not just managing the direct sound, but ensuring the off axis audio is also superb. In a Line-Array the vertical and horizontal dispersions are calculated independently of each other. The vertical (downward) off-axis audio is not of too great a concern in relation to the FoH sound as it is mostly focused at people and the stage (a problem for achievable gain before feed back, but not direct concern to us here). The upward traveling sound waves are however a big issue. These waves will reﬂect straight off the ceiling and into our audience (and probably the Mix position too). So great control is needed in this region. Above 2KHz this is typically managed directly by the HF Waveguide. From between 80Hz and 2KHz the off axis sound is managed by a combination of array height (quantity of enclosures employed) and inter-cabinet splay angle.
Array vertical coverage control, inter-cabinet splay dependant.
In the Horizontal ﬁeld we need to cover the seats at either far side, but in the areas beyond that the sound will simply be hitting walls & creating unwanted reﬂections. With this in mind a rapid drop-off of sound outside the main ﬁeld is essential, and it’s also important to ensure that any sound that propagates sideways (and is therefore reﬂected back into the seating area) sounds great and is free of anomalies. Many cabinets struggle to achieve this, especially looking at the midrange regions (typically 180deg at 400Hz, moving to 40deg at 1KHz).
Typical horizontal coverage control problems for dipolar designs. (Left: no control, 500Hz down) or (Right: anomalies in mid-range area 400-1.5KHz).
A Line-array approach is nearly always preferable unless there is a restricted ceiling height (4m or less). At that point a distributed approach should take over. If the two are to be combined then the same technology should be employed in both areas. This requires the speaker to have a low self height, and it must be able to operate in both array and point-source modes. If there is a room with a poor acoustic then a line-array speaker with exceptional pattern control is optimal. If the ‘throw’ of the system is greater than 40m then the introduction of a secondary line-array (quasi distributed) could be considered. The final factor to consider is gain-before-feedback, which can be effectively managed through array placement, and optimal coverage pattern control.