datasheets.com EBN.com EDN.com EETimes.com Embedded.com PlanetAnalog.com TechOnline.com  
Events
UBM Tech
UBM Tech

Design Article

Comment


kdboyce

1/5/2012 4:35 PM EST

My apologies to Fred vis a vis O'Toole vs Toole.

No problems Brad. I ...

More...



bcarso

1/5/2012 2:28 PM EST

Thanks for your clarifications Ken! And yes, I wasn't paying too much attention ...

More...

Generating spatial audio from portable products - Part 1: Spatial audio basics

Ken Boyce, Audio Technologist, Texas Instruments Silicon Valley Labs

1/3/2012 12:20 PM EST

Acoustic Beamforming

The term "acoustic beamforming," when applied to loudspeaker audio reproduction, refers to the ability to direct audio waves in a particular direction other than the loudspeakers typical radiation pattern.

An ideal loudspeaker acts as a piston-type source set in an infinite baffle and has a radiation pattern that is dependent upon the frequency being reproduced relative to the speaker disk radius. The degree of acoustic sound beaming is related to the ratio of the radius of the loudspeaker piston to the wavelength of the sound.

At low frequencies, the loudspeaker's sound spreads out evenly in all directions in front of the speaker. This is the reason you can set a subwoofer at almost any location in the front of a room and hear it equally well from anywhere else in the room.

As the frequency is increased, the radiation pattern becomes more focused in front of the speaker, increasingly becoming a narrowing cone shaped pattern around an axis perpendicular to the loudspeaker face. The sound pressure level is strongest within the cone pattern and drops off rapidly outside the pattern. You can test this by listening to high frequencies from your hi-fi loudspeakers while moving from side-to-side.

Also, real loudspeakers are mounted in relatively small finite boxes, not infinite baffles. The edges of the box will cause diffraction of the sound waves resulting in a more complex radiation pattern. If the loudspeaker was placed in an open air setting, you would correctly hear all of the frequencies reproduced only if you listened at a position that was on the axis perpendicular to the front face of the speaker.

If the loudspeaker is placed in a room, the listening position is less critical since the sound is also being reflected from the walls and other furnishings at various angles. While reflections make it easier to hear from any position, the reflections arrive at different times and intensities than the original signal and result in sound that lacks clarity.

Acoustic beamforming in the extreme case attempts to direct the sound energy emanating from the speaker to a specific angular position within the room environment. Acoustic beamforming with one speaker is very difficult, so typical applications use two or more speakers. Using multiple speakers also allows the use of constructive and destructive combinations of sound wave energy to create certain directional patterns.

Multiple speakers, as seen in Figure 3, are placed in an array where the array size and shape also make certain directional patterns possible. The array is usually physically linear, but may also be built as two dimensional arrays such as curvilinear, planar, circular, or combinations thereof.

Figure 3: In speaker array beamforming, sound waves constructively interfere at desired locations in space.

As a general rule, spatial effects are more easily accomplished and discerned at higher frequencies. However, a large array along with speakers with very good low frequency response will allow better directional control over low frequencies.

The size and shape of the array partially determines the technique(s) to use to achieve the desired spatial effect. The other factor is the purpose for which the array is being built. For example, if the array is constructed for the picture-in-picture feature, allowing two viewers to watch two different TV channels (low use case), another technique would be needed if the purpose was to simulate a 5.1-channel surround sound environment (high use case) for several people seated in front of the TV. As a practical matter, the spatial audio system may have to support both methods (and possibly several others) in order for the consumer to have pleasurable listening experiences regardless of the TV program-mode.

Go to Part 2 of this article.

About the author:
Kenneth Boyce is Audio Technologist for Texas Instruments Silicon Valley Labs. He previously served as Technology Director for National Semiconductor's Audio Products Group, and before joining National, Boyce served as director of the Audio and Communications Division at Oak Technology, which developed initial implementations of AC-97 Codecs and Digital Audio Controllers. He holds a bachelor of science degree in electronics from West Virginia University.

For more articles like this and others related to audio design, visit Audio Designline and/or subscribe to the monthly Audio newsletter (free registration).





EREBUS

1/3/2012 4:47 PM EST

Having tested some of these algorithms twenty years ago, I can honestly say that the technology is amazing. The advancement in inner cranial and exo cranial perception has been a long time coming from the research labs to commercial products. I think everyone will be amazed with how sound can be manipulated to put you into a 3-D sound environment to match the 3-D video available.

This is really neat stuff.

Sign in to Reply



bcarso

1/4/2012 2:13 PM EST

Although I'm a big fan of such processing in certain applications (for example nearfield "personal" monitors, one needs to note the importance of the room in more traditional audio settings, which generally is neglected and for which is much more difficult to account. And it tends to screw up spatial processing of this sort.

But merely treating the room as a negative, as is suggested in this article, flies in the face of listener preferences and other psychoacoustic results, some of which are sometimes subtle but rather well-understood now, despite continued misunderstandings in the popular press and elsewhere. In particular, it is misleading to state, as the author does, that reflections per se result in reduced "clarity". In fact, however counterintuitive, the reality is quite the contrary. See Toole's book Sound Reproduction for a comprehensive treatment of these issues.

Brad Wood

Sign in to Reply



zeeglen

1/4/2012 5:12 PM EST

Good observation Bcarso. Here is a fun experiment for a rainy day:

Start with the typical residential room with drywall (acoustically reflective) walls and ceilings and maybe some curtains and floor carpet acoustic damping for good luck.

Drive a single speaker with a sine wave about 300-400 Hz or so at comfortable loudness.

Plug one ear with an earplug.

Move around and see how many places exist where you can completely null the perceived audio tone with the position of the unplugged ear. Caused by reflected (standing) waves and destructive interference, the wavelength IIRC of 350 Hz is somewhere about a meter.

This effect exists with normal listening with both ears, just harder to notice.

Good article, but room acoustics do play a big part.

Sign in to Reply



bcarso

1/4/2012 8:47 PM EST

Indeed. If we were to take the "reflections diminish clarity" assertion seriously, we'd all be seeking out anechoic spaces for our listening enjoyment. For those of you who have had access to a good chamber, you know how absurd that would be.

Sign in to Reply



kdboyce

1/5/2012 1:02 AM EST

There was no mention of anechoic chambers in the article, and I definitely do not advocate them as a desirable listening environment. . They are, however, a good environment for making specific audio measurements. Anechoic chambers are also known as “dead rooms” while the most pleasant listening environment is a “live room”. The terms refer to the fact that a live room sounds more realistic while a dead room is, well… dead – no life.

There is no perfect technique to accurately reproduce sound. All methods have drawbacks and what one enjoys another will not. From a psychoacoustic point of view, practically all reproduced sounds we listen to don’t faithfully reproduce the original and are enhanced in some way for our listening enjoyment (or tolerance). This does not stop people from using the techniques, however. Witness a whole generation that only knows MP3 music quality and you know what I mean.

Sign in to Reply



kdboyce

1/4/2012 10:58 PM EST

I suppose the best statement would have been: “While reflections make it easier to hear from any position, the reflections arrive at different times and intensities than the original signal and CAN result in sound that lacks clarity.”

The intent of the statement in the article was not to explicitly “treat the room as a negative”. Rather it was to point out that the room does affect clarity, usually negatively. Because of this fact, much effort, including material in O’Toole’s excellent book, is expended to improve room listening environments for consumers.

As a practical matter, reflections (or echoes) are a very important part of acoustics as they help us estimate the direction, and distance of sound objects. But no reflecting surface is perfectly flat and that fact alone will blur the reflection, e.g. “smear” the sound. It becomes more complex as the number of reflecting surfaces increase.

Reflections can obscure the true source of a sound under certain conditions and reduce intelligibility. Longer echoes are generally a less offensive than shorter echoes. Echoes can cause phase interference which results in either reinforcement or partial cancellation of a sound at a particular frequency. For music, this can occur over a wider frequency range than for speech. When similar complex sounds in different phases interact, the effect is called comb filtering and is almost always undesirable as it obscures detail and harms intelligibility.

Sign in to Reply



azskibum

1/4/2012 4:12 PM EST

Nice overview Ken. I'm looking forward to reading Part 2.

Sign in to Reply



Frank Eory

1/4/2012 4:15 PM EST

Oops, wrong login -- that was me. Again, thanks for a good article Ken.

Sign in to Reply



kdboyce

1/5/2012 1:31 AM EST

I want to address a couple of other points Brad made.

1. Yes... the room situation can screw up spatial audio effects, especially if you are using the technique to try to reproduce a 5.1 or 7.1 audio system. In this case you may want to do some specific beam forming in order to create wanted reflections. Some makers of such systems also advocate having a microphone at the desired listening position, and use the data it picks up as a means of measuring the room characteristics and adjusting the algorithms accordingly.

However, the article premise was on small product use cases where space constraints did not permit even a reasonable stereo sound field. In this case, the spatial audio processing could enhance the listener experience and it was not intended for a multi-person audience.

2. As a closet musician, I am were aware of the improvement of synthesized sounds that can be accomplished with the judicious use of early reflections and reverberation as well as equalization. To me, a "dead" piano (or 'dry' as they say in music lingo) is useless - not real sounding. So reverb and reflection echoes are added (make the sound 'wet') to make the piano sound more realistic as if it were in a real room. All 'dry' or all 'wet' makes Jack a dull boy, so there is always an optimum mix of 'dry' to 'wet'. Whether this is accomplished totally electronically or partially by the room makes no difference so long as the resultant sound sounds more realistic to the average person. If so, then they can enjoy it. If not.... there is always complaints.

Lastly, I want to thank everyone who has read the article and commented on it. I hope my replies have helped.

Sign in to Reply



bcarso

1/5/2012 2:28 PM EST

Thanks for your clarifications Ken! And yes, I wasn't paying too much attention to the "portable" in the article's title --- my apologies :) I am on edge a bit I guess when I read things about rooms, particularly after some recent pronouncements about how poorly understood loudspeakers and rooms are (AES Heyser lecture by a certain prominent hifi magazine editor).

Floyd is just Toole btw. I joked with him when he was persuaded to leave NRC in Canada and join Harman International, that he had decided to become a capitalist Toole :)

Sign in to Reply



kdboyce

1/5/2012 4:35 PM EST

My apologies to Fred vis a vis O'Toole vs Toole.

No problems Brad. I am glad you read the article thru and your comments gave the opportunity for further clarifications.

Sign in to Reply



Please sign in to post comment

Navigate to related information

Datasheets.com Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)