Bowers & Wilkins launched Panorama 2 this week and we asked Senior Product Manager Mike Gough to explain how soundbars work.
There is no doubt that discrete speaker systems, be they 5.1-channel or more, will give you a better surround experience than any soundbar. If this were not so, manufacturers would have given up discrete systems long ago. Of course, not everyone is willing to accommodate 5 speakers or more around their room, so soundbars begin to be an attractive solution if they can offer a front sound stage that is wider than the speaker’s dimensions and some sort of surround effect. They are analogous to ventriloquists in making you imagine the sound is coming from somewhere other than its actual source. How well they do it depends on how they go about spreading the sound around the room and how amenable the room itself is.
We’ll look at the different approaches to soundbars in the next section, but it should be borne in mind that they generally operate best in regular shape rooms and when the soundbar is half-way along the wall. A degree of symmetry is a good thing and it’s more important at the front than behind the listeners, but that’s not to say that they don’t work at all if symmetry is missing – just not quite as well.
Types of soundbar
The simplest type of soundbar is the so-called LCR (left-centre-right) design. If you start with conventional separate front left, centre and right speakers and simply mount them all in the same cabinet, that’s it. They’re generally cheaper than other types because they can be passive and they have three separate speaker-level inputs that you connect to your multi-channel amplifier in the normal way.
For aesthetic reasons, all soundbars are designed to be more or less the same width as the television they work with, so it’s not surprising that the width of the soundstage with simple LCR soundbars is narrow. With discrete speaker systems, you normally have the left and right front speakers some distance from the sides of the screen and this helps keep the audio soundstage in scale with the picture; but this is not the case here. Furthermore, because there is no post-processing of the sound, the front channels are all you get. If you want any surround information at all, you are back to adding separate speakers behind the listeners.
This class of products uses an array of small drivers, to which may be added extra, larger drivers for bass reproduction. Complex digital signal processing (DSP) sends all channel signals to all the drivers in the array and adjusts the delay time to each driver such that a different beam of sound is radiated for each channel. The centre channel information is fired directly towards the listeners, while the left and right front and all surround channels are directed to bounce off the walls of the room and be reflected to the listeners from the appropriate angle. By their very nature, these must be active speakers.
Here is a simplified diagram of what happens. We start with an array of small drivers directly in front of the listeners and each one is fed the centre channel signal equally. A broad beam of sound is directed to the listening area. There is some spreading out, of course, but we are keeping things simple for illustration.
Now we add the front left and right channels. By adding progressive delay to adjacent drivers, it is possible to steer two beams towards the side walls and the dotted speakers show the position of the virtual drivers that define where the sound seems to come from. Note how these are necessarily in front of the array.
Now, we’re going to reduce the amount of delay between adjacent drivers and create two more beams that will carry the left and right surround channels. Here each beam is shown being bounced off two walls and again the dotted speakers are positioned where the sound seems to come from.
Beaming can only happen when the wavelength of the sound is comparable with or less than the dimensions of the array, so it gets more precise the higher the frequency. At lower frequencies, the radiated sound gets progressively more omni-directional. Early versions of the array had substantial height as well as width and this enabled vertical as well as horizontal steering. Indeed, the surround channels were bounced first off the ceiling, then the rear walls to reach the listeners from behind. However, users preferred less height to their speakers and so most current versions steer the beams only in the horizontal plane.
Modern sound projectors have an auto-setup routine in an attempt to cope with irregular-shape rooms. A microphone is placed at the centre of the listening area and the direction of the beam is automatically adjusted until a maximum level is found. Of course, the microphone only occupies a very localised position, so the setup gets more approximate the more spread out the listeners are.
Another limitation of sound projectors relates to the small size of the drivers that make up the array. They have to be small and mounted as close together as possible for the beam to be steered properly and this means that they have limited output capability. True, there are a large number of them, but a compromise has to be made between the maximum loudness and the lowest frequency that the array can be used to.
Soundfield shaping systems
Panorama falls into this general category, which is based on the acoustic antenna principle. Here is a simplified representation of the principle, using two drivers with specific directivity and mounted at a specific angle to one another. With some clever DSP modifying the signal to each, you can create a null in one direction and a main radiating lobe extending in a different direction. By modifying the DSP, you can define the angles of both the main lobe and the null. The null is always directed to the listening area and the main radiating lobe is given a different direction, depending on whether the front or the surround channels are being processed. In the sound projector, all channels are processed through all drive units that make up the array. Here, each pair of drive units handles the front and surround channels of one side only.
Let’s now assemble a complete soundbar. There is a 2-driver acoustic antenna at each end of the soundbar, one producing the left front and surround channels and the other producing the right front and surround channels. The front channel lobes are coloured green and the surround lobes red as before and, again, the dotted speaker outlines represent the virtual sound sources of the sound that gets reflected back to the listeners. You can see that they are more diffuse than the virtual sources of the sound projector, which makes this design much more tolerant of both listener position and room shape. The virtual speakers for front left and right are also further back than in the case of the sound projector, more closely resembling the discrete speaker situation.
In the centre, you can see the yellow lobe representing the centre channel, which is produced in the conventional way by forward facing drive units.
Soundbars are no different to any other speaker design in that products utilising the same general principle can be differentiated by their sound quality. It comes down to the design of the drive units and enclosure. In the same way that some 6-in 2-way speakers sound better than other 6-in 2-way speakers, so some soundbars sound better than others, regardless of the general principle used.
As in the original Panorama, we use high quality drive units, including our trademark tweeter with Nautilus Tube loading. The centre channel comprises two 3-in midrange drive units each side of a central tweeter. This is a typical centre channel complement and the drivers are dedicated to that important role.
The acoustic antennae at each end of the enclosure each comprise two 3-in full-range drivers. All the 3-in drivers are capable of producing realistic levels down to the frequency at which all channels cross over to the two dedicated bass drivers. These are not shown in the diagram above as they form no part of the general principle, but they are situated between the centre cluster and the outer antennae and handle the low bass of all channels. Indeed, they do it so well that you don’t have to use a subwoofer. Of course, there is the ability to connect one to the Panorama 2 and that can make the system both extend to extremely low frequencies and also play a whole lot louder – useful for larger rooms.
All the above applied to the original Panorama, so let’s now look at some of the differences in this later version.
Panorama 2 now has full HDMI connectivity. Rather than just connecting the audio part of the signal to the soundbar, you can connect an HDMI cable from your source device (eg Blu-ray player) to the Panorama 2. A further HDMI connection is made between the Panorama and the TV. Apart from simplifying connections, this allows an on-screen display for setting up the Panorama 2. The ‘pebble’ remote handset has been modified to enable on-screen control.
In addition to the remote, there remains an on-board control/display panel, but this now has touch controls and a proximity sensor. The display is off in normal use to avoid distraction and the proximity sensor brings it back to life as soon as your hand approaches the control surface.
We have reduced the listening modes to two – music and cinema. These correspond to the wide and surround modes of the original model. Both these modes are better rendered by changes to the DSP and improvements to the drivers. The bass drivers have a longer linear excursion capability, lowering distortion and allowing higher playback levels. The dispersion of the 3-in drivers is wider, which gives a more gradual, seamless blend between the various channels. No longer are you aware of separation between the front and surround channels. They blend into a more continuous 3-dimensional sound space.
Finally, and in line with our ‘green’ credentials, the power supply has been redesigned with lower standby consumption – less than 0.5W.
Senior Product Manager, Mike Gough.