LD#03: The soundtrack comes after you... Revised: 08 Feb 95 1. Introduction ____________________________________________________________ Surround sound is potentially available from stereo VHS/UHV/IRD/cable, MTS broadcast TV, stereo video tape, laser video disc and a few audio CDs. The topic is included in my LD (laserdisc) article series because surround is frequently the next enhancement sought after obtaining an LD player. There has been a fair amount of discussion over the years in the Usenet Netnews groups rec.video, rec.audio and rec.arts.movies on the subject of surround sound. Much of it has concentrated on theory; why recovery of 3 or 4 channels of sound from 2 channels of data is [im]possible, what artifacts are introduced, etc. The purpose of this article is to give you an idea of how surround relates to the rest of your home entertainment components, and what benefits it conveys. Article Contents: 2. The Experience of Surround - what does it sound like? 3. Putting Surround in Perspective - how important is it? 4a. How does matrix surround work? - how is it encoded? b. How does discrete surround work? 5. The Hierarchy of Home Surround - a travel guide to the buzzword maze. 6. Surround System Configuration - what hooks up where. 7. How to select a matrix decoder/processor. 8. Some surround titles - suggested demo titles. 9. References - tutorials and test reports. 10. Some related surround traffic from other contributors (>) & my replies. Some History: Surround sound is not new. Disney's "Fantasia" (1940) had discrete 6-channel sound, as did the Cinerama series of movies in the 50s. Surround sound in the home is not new either. 15 years ago it was called "quadraphonic". Indeed, the "Dolby Surround" system of today is not much different from the CBS "SQ" matrix system of a decade ago. "Dolby Stereo" = "Dolby Surround" = "Dolby MP", for home purposes. Many of the "stereo" soundtracks used for your laser video discs and VHS tapes have been encoded for surround since the mid-70s. This is because the encoding for "Dolby Stereo" (Dolby Motion Picture matrix, or simply Dolby MP) is the same as for home "Dolby Surround". It is easier for video producers to simply transfer the encoded signal from the theatrical audio master (or release print) to the video master, than to decode to simple stereo or go back to the pre-encoded audio elements and remix. More recently, several discrete, or "5.1 channel" surround systems have been fielded theatrically. These use a compressed digital audio format, either on the 35mm print or on separate CD-ROMs, to provide increased channel separate. One of these schemes (AC-3, aka Dolby Stereo Digital), threatens to invade the home. Unfortunately, most of the press coverage of surround has focused on technical specifications, and none that I have seen attempts to give you any idea of what surround actually *sounds* like. 2. The Experience of Surround ______________________________________________ * It may be a very long time before we have convincing 3-D video, but effective 3-D audio is here today. The audio portion of a surround presentation fills the entire listening space. * The difference between surround sound and conventional stereo can be as dramatic as the difference between stereo and mono. * To demonstrate this, during a sequence when the surround channel is active, try switching "effects OFF" on the decoder or processor. The consequence is that the sound collapses to the front of the room. Switch it on, and sound floods the room. * Observation: you can notice the *difference*, but on a tastefully mastered surround program, you often don't consciously notice the presence. The program is simply more involving. The psychological distance between you and the program is reduced. The image may still be at arm's length, but the sound joins you in the room. * Of course, on a tastelessly mastered work, herds of objects noisily zooming out into the room can become an irritation, but films like that are apt to be infested with myriad artistic defects of other kinds. The surround effect is almost subtle. Unless you have been informed that you are listening to a surround-encoded program on a surround system, you are likely to simply have a more engrossing viewing experience, without necessarily knowing why. You quickly get accustomed to hearing "image foreground" sounds from around you, after all, in real life, that's where they come from. However, the general public is not clamoring for surround, 70mm, wide aspect ratios or 60 frames/sec. in the theatres, because they are not consciously aware of the contributions made by these technologies. None, including surround, is quite as profound and easily identifiable as the difference between, for example, silent-vs-sound or B&W-vs-color. Surround may not be getting the public attention it deserves, but then, the lack of technology awareness is desirable in art. You don't want the medium to overshadow the message. 3. Putting Surround in Perspective _________________________________________ It is possible to create a home surround theatre whose sound exceeds that of your local 35mm "Dolby Stereo" movie hall. You will soon be able to match the sound of any AC-3/DSD halls that might exist in your area. However, surround sound is NOT the first step in a home theatre. If you are watching VHS tapes on a 13-inch department-store TV and listening to the audio through the TV, or even the 5-inch speakers of a $100 discount store "rack" system, then begin your upgrade elsewhere. Don't get surround until you have high quality in the following other areas... * Audio signal: Tape: The absolute minimum for tape is VHS linear (analog) stereo. VHS linear mono is incapable of surround, and you may not be satisfied with linear stereo. VHS linear may provide seriously inferior surround due to poor azimuth alignment at the head, and lack of Dolby-B on some decks. VHS HiFi is an improvement over VHS linear, but HiFi has compression and 60Hz cyclic artifacts that may eventually bug you. Broadcast: MTS stereo can provide acceptable surround, IF (big if), your TV has a "real" MTS decoder (with dBX) and your broadcaster isn't trashing the signal. Satellite stereo is generally of high quality. LD: The finest video sound is from laserdisc. Any LD player will provide FM stereo, and virtually all post-1985 players also have digital stereo. CD: A few titles have been release that are specifically mixed for Dolby decoding. Ambisonic CDs are not likely to sound better in Dolby Surround than in plain stereo. DVD: Digital Video Disc plans to use AC-3 5.1 discrete. I have no information yet on whether or not DVD will offer 2-channel re-encoded matrix output. * Video signal: If you principally watch VHS tapes today, upgrading to LD, satellite or at least S-VHS with HiFi is apt to provide a higher return on investment than adding surround. * Monitor: I suggest at least a 23-inch display (whether direct-view, front- or rear-projected) with at least 350 analog lines (or 640 discrete pixels) of real horizontal resolution, composite video input, and capable of correct setup for geometry, size (overscan), black level, white level and chroma. View at distances of between 4 and 8 screen heights. * Audio chain: The main (front) channels need to have speakers with fairly flat on- and off-axis response, with no resonances, no breakup or distortion at moderately loud listening levels, and backed by an amplifier of adequate power that adds no problems of its own. The treble response needs to be flat to 15KHz or more, and the bass response needs to reach down to at least 35Hz - the lower the better, since film/video programming has much more deep bass than music. Unless you are using a subwoofer, the main speakers need to be able to cleanly handle a lot of low bass. If, for example, you can't tell the difference between CD and pre-recorded audio cassettes on your system, you probably need a complete audio upgrade. Note on MTS: VHF/UHF broadcasts encoded for NTSC-MTS stereo can carry surround, but the stereo signal is often ruined by the local broadcaster or cable operator, leaving you with a mono, [re]simulated stereo, out of phase or highly distorted stereo signal. Don't get surround just for MTS programs unless you are certain that you have reliable access to solid stereo programming, and you have decent MTS decoder in your TV (many so-called MTS decoders lack the required dBX decoding, and are junk, even in some "hi end" sets). Note on Mono: If you play a surround-encoded signal on a mono VCR or TV, or through a mono audio system, mixing the left & right together, any sound intended for the "surround" channel will be cancelled out altogether and will be inaudible. For this reason, some stereo surround material is labelled "non-mono compatible". In fact, the surround channel component of *all* surround program material is non-mono-compatible. 4a. How does matrix surround work? _______________________________________ Matrix surround encodes 4 channels (left, right, center,surround) into two normal stereo tracks. [Over]simply stated, sound intended for the "surround" channel is recorded in the normal left & right stereo channels, but out of phase with respect to each other. Anti-phase is "rear" (surround). In-phase sound, balanced in both left and right is "center" or "dialog". Other sounds are left, right or some blend. Some hall-dependent delay may be added (by the playback decoder or processor) to the signal sent to the surround channel so that listeners far back in the theatre won't hear the surround signal (esp. simulated echo) prior to the original front channel sound. This also ensures that any "front" sound that "leaks" into the surround channel is heard after the real front sound, improving clarity (the "Hass" effect). In the Dolby system, the surround channel also employs mild (5 dB) Dolby B noise reduction. Depending on the decoder, other signal processing and channel amplitude manipulation (steering) may be applied to attempt to cancel various signals out of channels where they aren't "supposed" to be. Normally the effect is subtle and effective. To hear the steering in action (and struggling), play a dual-audio program, such as a laserdisc with a monophonic soundtrack on analog channel 1/L and a commentary on 2/R. With Pro Logic, the sound will jump all over the place. (Of course, a really smart unit may just give up in the face of such mistreatment, and shut down decoding.) If you listen to an un-decoded surround program on an ordinary stereo setup, you may detect the out-of-phaseness of surround signals (particularly on headphones - see postscript). The soundstage may appear to be wider than your speakers, or you may actually have a psycho-acoustic experience of sound from behind you (I noticed this on the "LadyHawke" LD, prior to having a surround decoder.) Technically stated, the "Dolby Stereo" MP Matrix ENCODING looks like: Source Sounds As Encoded Lt Left ----> + -------------------------------------------------+---> Left Source ^ ^ Total | 5 dB | | 100-7K Dolby +90 deg Ctr ---> -3 dB Surround ---> -3dB ---> band ---> B NR --->| Source | Source pass Encode -90 deg | | v v Rt Right ---> + -------------------------------------------------+---> Right Source Total Pre-recorded "surround" programs have Dolby Lt and Rt composite signals in the left and right stereo channels. To extract left/center/right/surround, you need a DECODER. Since several people have asked: I do not know how the +90/-90 degree phase shift is handled. For sounds not correlated to L&R front, +0/-180 probably works as well. For correlated sounds, +0/-180 is probably safe for playback systems that have rear delay. Optimally, a frequency dependent shift is probably used. A pair of simple passive filter circuits might suffice. 4a. How does discrete (5.1) surround work? _________________________________ Theatrically, there are three competing systems of discrete 5.1 surround: - AC-3, aka Dolby Stereo Digital (DSD), 16 bit, 384 kbps, 16:1 compression - DTS, Digital Theatre Sound, 20 bit, 1.2 Mbps, 3:1 compression - SDDS, Sony Dynamic Digital Sound (?) All have the following general features in common: - The full channels are: left front, center, right front left surround and right surround - The ".1" channel is: subwoofer (limited bandwidth) - Sound is stored digitally - The digital data is compressed The AC-3 and DTS systems also: - Discard some audio data to acheive a higher level of compression Of the three systems, only AC-3 appears to be near delivery on home media (LD and DVD). LD media with AC-3 coding is already for sale. As of 2/95, no players with AC-3 decoders or RF outputs had been released. Full details of the AC-3 system haven't been revealed, but how it generally works is as follows: The 6 source channels are digitized, at 20 bits/sample, and apparently at 44.1 KHz. The subwoofer channel is low-pass filtered to eliminate sound above 300 Hz. The raw combined signal stream would be about 6 Mbps. Data compression is applied to eliminate redundancy, and psycho-acoustic masking rules are used to remove sounds which theoretically can't be heard anyway. The net result is a data stream of 384 Kbps/sec (about 1/4 of the standard audio CD rate). This data stream is easily stored in various nooks and crannies of different media. On 35mm film it occupies the space between the sprocket holes (at 320 kbps). On laserdisc, it replaces analog channel 2/R. Most pre-AC3 players will ignore the AC-3 signal (and finding no analog audio on 2/R, will feed the 1/L audio to both stereo outputs when analog audio is selected). 1995 and later AC3-compatible players will either: - decode the AC-3 signal internally, feeding six (6) audio ports, probably via a ganged connector. - provide an RF output for an external decoder, much like the "AFM" output available today on some industrial LD players. In either case, once decoded, your audio system will need to be able to accept up to 6 separate audio inputs. Many low-end surround systems expect and support only simple two-channel (matrix surround) inputs. Another unresolved question about AC-3 (and any other future home discrete surround) is, if you have less than 6 channels in your audio system, can the decoder provide: - subwoofer blended into L+R fronts? - split surrounds mixed for a mono surround setup? - center mixed into L+R for "phantom center" - mix-down to simple stereo, or even mono, where required? AC-3 decoders for LD, at least, are required to also provide Dolby Pro-Logic for decoding the normal digital and analog stereo tracks. DTS and SDDS have yet to provide credible evidence of their intentions in the home theatre market. DTS could theoretically replace the normal "RedBook" digital stereo baseband AM signal on LD, but would require double-stocking of LDs, and an external decoder would be required for existing player (assuming the had a raw digital output, as most digital-sound models do). SDDS relies on 2 or more external CD-ROMs in the theatre, and presumably could do so in the home, but this seems highly unlikely. Indeed, the successful acceptance of any discrete implementation on LD remains to be seen. I personally have no plans to replace my current LD player simply to obtain AC-3 compatibility. Traffic in the alt.video.laserdisc Netnews group suggests that long-term LD enthusiasts are confused about discrete surround and AC-3 in particular, and many seem to share my opinion about the value of an "upgrade". The average LD owner, and potential LD owners, are probably hopelessly lost, and will not be drawn to invest in it. 5. The Hierarchy of Home Surround __________________________________________ There are several implementation levels (and a large number of buzzwords and trademarked terms) associated with home surround. Here is a condensed taxonomy of home surround: Stage 0: Two-Channel Surround Emulation. At this level there are only two front speakers and two channels of amplification. A pre- or post- processor is used to tinker with the sound to produce ambiance or "surround like" effect at certain listening positions. Not all of these techniques are marketed as alternative "surround" methods, but some are, and you need to beware of them if what you want is movie surround. Examples include: Carver Sonic Holography, Hughes SRS, Q-sound and the "surround" mode of some low-end stereo(only) receivers. In my opinion, none of these are satisfactory for presentation of Dolby surround-encoded material. Stage 1: Two-Channel Speaker Bridging, also known as the "Hafler connection". Because the surround signal is largely L-R, you can simply wire one or more surround speakers and a potentiometer across the (+) terminals of the left and right front speakers. Don't attempt this without first checking that your amp allows bridged connections (some will self-destruct), and don't expect wonderful results, particularly since left and right signal can bleed into the opposing speaker. If you want to try speaker bridging, see the June 1991 issue of Audio magazine for tips. Stage 2: Passive Matrix. A simple decoder passes Lt to left, Rt to right, then isolates and subtracts the Lt and Rt, sending the Lt-Rt result to the surround speakers. This at least eliminates the L/R bleeding. Stage 3: "Dolby Surround". A branded "Dolby Surround" decoder: * sends Lt-Rt (surround) thru a delay line (typically 20 mSec), * then thru a 7 KHz low-pass filter (to keep natural and azimuth error caused left/right source phase noise from being heard as surround) and * thru a 5 db (vs 10 on audio cassette) Dolby B-type noise reduction circuit. * A master volume control and input balance controls are also provided. * Lt and Rt are isolated, and may also be summed (Lt+Rt) and sent to the Center output, but this is not required for Dolby cert. In any case, a maximum of 3 dB of separation is achieved between each adjacent pair of: left-center-right-surround-left. Stage 4a. "Dolby Pro-Logic". A Pro Logic" decoder starts with Stage 3, and: * replaces the simple Lt-Rt (surround) and Lt+Rt (center) extractions with an active adaptive matrix decode step. For signals intended for one output, this circuit attempts to cancel them in the others. * It also analyses the soundfield for signal dominance, and focuses the sound toward those outputs. The net result is that 30 dB of separation is possible between any two channels. Dolby Labs publishes a "Principles of Operation" pamphlet that goes into more detail. * Input balance is required. * Lexicon Pro-Logic decoders also include an "auto-azimuth" feature to correct for group delay errors (time shifts) between the two source channels. Shifts as high as 50 microseconds are reportedly common in film-sourced sound (45 degree phase shift at 5 KHz). Incidentally, don't bother looking for any "Pro Logic" recordings. Pro Logic is used only in the playback processing. The encoding (recording) of Dolby Surround always uses the "Dolby Stereo" (aka "Dolby MP") matrix described above. Pro-Logic ICs are not generally available to the public, although at least one "kit" based on the Mitsubishi M69032P has been reported in Australia. Stage 4b. When AC-3 matures, it will probably go here. Stage 5. "LucasFilm THX". THX affects the presentation of movies at several points. For the home, a THX-certified processor starts with Dolby Pro-Logic and adds: * Surround channel decorrelation - a digital pitch shift is used to make the (mono) surround signal "different" in the left and right surround channels. * Re-equalization of the front channels, to make the movie mix seem less "bright" in the home. * "Timbre Matching" - an equalization applied to the surround channel to make effects sound consistent when panned between front and surround speakers. * THX is also becoming a LaserDisc certification program. The first title with "THX quality" presentation parameters will be the director's cut of "The Abyss", to be released in 1993. THX specifies front speakers with a reduced vertical dispersion (to minimize ceiling reflections) and two side-mounted surround speakers configured for dipole radiation. THX also recommends equalization for the L-C-R channels. A THX equalizer will have 1/3-octave bands from 80 to 800 Hz, implemented as "interpolating constant-Q" circuits, and parametric equalization above 1000Hz and for the subwoofer channel. Reportedly, a THX-certified equalizer will include in-home calibration. THX is also a certification process for video programming, beginning with LaserDiscs in early 1993. As with Pro Logic, THX-branded audio material is not "THX encoded". THX mixes are merely a specific way of performing Dolby-MP sound mastering. A THX-certified LD presumably will have purist aspect ratio, source element/transfer quality, colorimetry/timing, etc. The telecine monitor is calibrated by LucasFilm and a "THX Vertical Interval Test Signal(TM)" is inserted during NTSC encoding of the signal. Various full-field test patterns are included after the end of the program on LD. Other moviesound and home surround terms: "Ultra Stereo", "Chace Surround" and "matrix surround" are DolbyMP/Surround- compatible anti-phase encoding schemes that do not bear the Dolby logo. You may encounter these terms on program material. They will work on your passive, Dolby, Pro Logic or THX decoder/processor. Hughes "SRS", Carver "Sonic Holography" and "Q-sound" are not, as far as I know, Dolby-compatible. They are 2-channel schemes that process the signal on playback (SRS, Carver) or prior to recording (Q) and attempt to simulate wider or 3-D sound placement with only the normal two front stereo speakers. The effect may be limited to a small "sweet spot", and I don't recommend additional Dolby Surround processing. I have SRS on my Sony XBR TV, and with or without my external Pro Logic decoder switched in, SRS-on is principally a "hum enhancer" and "listener phase torture device". SRS is no substitute for Dolby, as far as I'm concerned. I have heard Carver several times, and am left uninspired. Q-sound I have not heard extensively. Ambisonics is a matrix encode/decode algorithm apparently intended mainly for music reproduction. I have no technical details. Reportedly some surround receivers introduced in 1992 will have Ambisonics in addition to Dolby. Dolby A, B, C, S and SR are noise reduction processes that have nothing to do with surround except that Dolby MP uses a modified Dolby-B on the surround signal, and VHS linear stereo uses (additional) normal Dolby-B on both channels. If your deck lacks Dolby B on the linear stereo, you can expect mistracking of the modified (additional) Dolby B used on the surround signal. Dolby HX-Pro is a variable-bias technique for analog tape recording and has nothing to do with tape playback, much less optical media or surround sound. Dolby SR-D is an early name for Dolby Stereo Digital (DSD, aka AC-3). Cinema Digital Sound (CDS) is another reduced bit rate (lossy compressed) digital sound encoding format for 35 and 70mm filmstock. CDS replaced the existing optical tracks. CDS failed in the market, and is probably gone for good, although it appears in some on-screen credit rolls. "AC-2" is Dolby's term for their family of 192 Kbps reduced-bit-rate data stereo coders. "AC-3" is is their "5.1 channel" surround version (320 Kbps). Dolby has been requesting that developers of next-generation media leave at least 320 Kbps of bandwidth available for AC-3. "THX" is a LucasFilm trademark for several things, three of which are directly related to home surround: 1. "THX Theatre" - THX is a certification process. Theatres bearing the logo are periodically tested to ensure that they meet LucasFilm standards for audio equipment, environment and playback of surround- encoded film. 2. "Re-recorded in a THX theatre" - THX logos on films and recordings indicate that the final DolbyMP-compatible mixdown was done with the recording console and engineer located in an actual THX-certified theatre. This is intended to ensure that the film audio will playback in a consistent and predictable manner in all THX theatres (and in homes equipped with THX certified components). 3. THX crossover - LucasFilm lists recommended audio components for THX theatres. They also make a crossover, bearing the THX brand, which is only used in actual motion picture theatres. 4. Home THX - LucasFilm has a testing and certification process for home audio equipment. Those models which are submitted by the maker, and pass the tests, may exhibit the branding. THX branded equipment provides the promise of effective home theatre, but can still sound hideous if improperly set up and calibrated. Some THX-branded equipment includes dealer installation and adjustment. For amplifiers, THX merely provides an assurance of high power and quality. The first THX-cert LD player (from Runco) was released in late 1994. 5. THX certified surround decoders, equalizers, main speakers and surround speakers, on the other hand, must provide specific THX required functions (listed in Stage 5 above), as well as high general quality. 6. THX certified LaserDisc media (described earlier). The future. Your humble prognosticator predicts: AC-3 on LD will not be a major factor in the market. What may be a major factor is DVD (which uses AC-3). DVD threatens to make major inroads into the VHS rental and sell-thru markets, as well as capture some share of the LD market. If this happens, the makers of home theatre audio hardware will shift rapidly to providing equipment with discrete inputs for 5.1 channels. AC-3 has not been extensively A-B (much less A-B-X) compared to high-quality matrix or less compressed discrete systems like DTS and SDDS. Until critical independent observers test and report, it is difficult to predict the success of AC-3. My personal opinion is that the bit rate is too low for acceptable fidelity (or put another way, the 16:1 compression ratio is too high). AC-3 may flop. Regardless, a Dolby Pro-Logic or THX matrix decoder will remain a viable investment for the rest of this century. 6. Matrix Surround System Configuration ____________________________________ Here is what a largish Dolby surround setup looks like in the home. You can easily get by with only four speakers/channels (I do). ============== Screen _ ....._..... _ | | : | | : | | /___\ : /___\ : /___\ Left Front : Center : Right Front :.........: Subwoofer : | : : _____ | _____ \ / : \ / |_| | |_| | | Surround : Surround | | /___\ | /___\ Audience : | : | : _____ | _____ \ / : \ / |_| | |_| Surround : Surround Room: Any size room is acceptable if the decoder has adjustable rear delay. With fixed delay, you must get the surround speakers properly located with respect to the front speakers and the delay value. See "surround" speakers below. Main (Left&Right): Uniform response from 35Hz (or lower) to 15Kz (or higher) is essential. Reduced vertical dispersion is desired to minimize ceiling reflections. This probably means that my Vanderstein 2Ci's, which have the midrange radiator tilted up, are not ideal for moviesound. Solid bass tolerance (if not response) is necessary in the mains. Movie sound has a great deal of low bass. Unless you are using a subwoofer, make sure that your intended speakers won't break up or be destroyed by challenging program material. The LaserDisc of "The Dream is Alive" (Lumivision LVD9019) makes an excellent test. Caution: the shuttle launch sequences in this program have destroyed speakers. Center: In both the theatre and the home, only a small portion of the audience is sitting near the centerline. Those near the sides might hear center channel sound (sounds equal in both left & right speakers) as coming from the front speaker nearest their side. If the decoder has a "center" output, it attempts to isolate in-phase, balanced sound, usually dialog, which it thinks should be "front center". It sends it to the center output and attempts to cancel it from the other outputs. That way, everyone hears "dialog" from the screen center speaker. Further, if you have no center speaker, and rely on the "center" sound to consist of an equal-and- in-phase signal emitted from both left/right front speakers, the sound waves will not arrive in-phase at all listening positions. This is the so-called "comb filter effect". For example, a 6-inch difference in distances to L/R speakers results in a 180-degree group delay (and total cancellation of the direct wavefront) at 1000Hz (and many other frequencies). LucasFilm reports that center- channel dialog is easier to understand if it comes from a single speaker. A THX-certified decoder, and many (but not all) Pro-Logic decoders, will roll off the center bass below 100Hz (-3dB @ 80Hz, -10dB @ 50Hz). If the bass is rolled off, it will be delivered to the subwoofer channel, or merely left in the L and R main channels. Although matching the L-C-R speakers is ideal, if you are certain that your decoder rolls off, you can use a smaller and less powerful speaker for "center". I am not yet using a center speaker, and can't contribute further comment. Another case for a center speaker is horizontal sound pans. If the speakers aren't identical, the timbre, and at some frequencies, even the phase of the sound will vary, providing a very distracting sonic image. If you don't plan a center speaker, the recommended positioning of the L/R mains is "close" to the screen. If you must place the mains far from the screen, consider using the TV's own speaker(s) as the "center". Unfortunately, most TV speakers are junk, so you'll have to judge the efficacy of this by trial. Be careful with "close" - the magnets in many speakers are powerful enough to skew the geometry and color of inadequately shielded TVs. Subwoofer: Although film sound has much more bass energy than music, due to sound effects, the case for a subwoofer in a surround setup little different than for a normal stereo setup. If your front speakers have wimpy bass, and you don't want to upgrade them, and you have lots of money, get a subwoofer (and maybe an amp to drive it). If your system can reproduce cleanly down to 40Hz, you are probably OK as is. The subwoofer may generally be located anywhere in the room. Pay attention to where the low-pass crossover is. Having a filtered subwoofer output in the surround process *plus* a filter or crossover in the subwoofer itself is not a "sound" idea. Surround: Notice that the side and rear speakers are *all* labelled "surround". In a textbook Dolby theatre setup, there are a number of them (typically 8) spaced around the hall, they all emit the same signal and they all point toward the audience. The point of having multiples is that each thus runs at a lower volume, the surround field is more uniform, and listeners near an individual surround speaker won't have their attention drawn to it. Theatres also use multiple surround speakers to achieve coverage. The LucasFilm-recommended number of home surround speakers is TWO, located to the SIDE, and not behind the audience. The rears are optional, but if present there must also be side speakers. Dolby recommends that the surround speakers be located 5 feet closer to the average listener than the front speakers, and that the "surround" signal be electronically delayed by 20 milliseconds (for a net arrival delay of 15 mSec compared to "front" sound). The Dolby publications "a listener's guide" and "Pro Logic Principles of Operation" both include distance-time nomographs. Generally, T = Nd + Df - Ds, where T is the delay setting, Nd is the net delay time in milliseconds (15), Df is the distance in feet from the listener to the nearest front speaker, and Ds is the distance in feet from the listener to the nearest surround speaker. The Shure recommendation for Nd is 20. I suggest using the Dolby value (15), and if T works out to be a value that you can't set precisely because of fixed delay steps, chose the next higher delay setting. Unless you are using a THX processor, or a decoder that otherwise differentiates left and right "surround" signals (in movie or ambiance modes), a single channel of surround amplification will suffice. One or multiple speakers may be used. Whether to wire in series or parallel depends on the impedance of the speakers and amp. If the two surrounds are at significantly different distances from the audience, separate amplification channels are needed to match levels. Surround speakers: You may be able to get by with modest surround speakers. In the Dolby mode of your decoder, the sound sent to these speakers is rolled off above 7 KHz, and although rolled off below 100 Hz during ENcode, it is NOT rolled off below 100 Hz during DEcode. Any deep bass naturally out of phase in the original left and right sources will appear in in the surround channel (particularly if the decoder/processor has a subwoofer output). A case can be made for matching the speakers all around. Several people have reported significant bass energy from their surround speakers, and some decoder/processors send full-range material to the surround speakers when in proprietary (non-Dolby) music surround, ambience or venue simulation modes. There have also been press reports that some film producers are pre-emphasizing the bass in the surround mix, so that it will still be present after the encoding roll-off. With respect to matching all-around, on the other hand... The LucasFilm/THX recommendation is that the high frequency radiators of home surround speakers NOT be pointed at the audience. The original formerly THX-certified speaker from Cambridge SoundWorks ("The Surround", $400/pair), for example, has its bass cone pointed at the audience, but has two mid/high cones per unit, wired out of phase (dipole) and pointing sideways. The newer, less powerful "Surround II" ($250/pair, not THX) has dual sets of elements, all in dipole, and nothing pointing "forward". The Alantic Technology Surround speakers use a slightly different scheme, wherein the drivers face each other at a 90 degree angle (instead of away at 180 degrees). They are still wired anti-phase. and people who own them report being pleased with the performance. If you end up using ordinary forward-firing speakers, and find that you are too aware of them, try facing them straight up (and re-adjust channel level matching after rotating them). You typically don't need a 14-inch woofer or thermonuclear tweeter for the surround speakers, or much amp power for that matter. Any decent bookshelf speakers and well-mated amp will do. I'm using a retired 55 W/ch receiver, driving a pair of CSW Surrounds. The dipole configuration does diffuse the surround more than the traditional speakers (CSW Ensemble I's) I used to use. Further note: If the speakers and amps are not all identical, it will not be trivial to ensure that they are all in phase and balanced. For phasing, I suggest testing one pair (each non-identical) in a simple stereo setup (with a mono signal), and correlating the polarity markings on the binding posts. Be sure to use the eventual amp channels for this, as some amps invert the signal. 7. How to select a matrix decoder or processor _____________________________ Step 7.1: Is your system ready for a decoder? You may need (or want) a new main receiver or amplifier. The surround process requires exporting the raw stereo-matrix signal at the pre-amp (line) level in the receiver/amp, then feeding the decoded front signals back in at that point. The input stages of the receiver/amp handle the matrix signal; the main output stages handle only the decoded "front" signal. Receiver or Amp __________ _________________ | Surround |---Lt---| Pre :: Power |---L---|Spkr< | Source | | :: | | e.g. LD |---Rt---| :: |---R---|Spkr< | VCR,CD | | Ext Proc Loop | `----------' `--Out------In----' | | ^ ^ v v | | .------------------. .------. | In Front-Out | | | | | | | | Sur Left |-->| Rear |---|Spkr< | | | | | Surround Decoder | | | | | | | | Sur Right|-->| Amp |---|Spkr< | | | | | Center SubWoof | | | | Out Out | `------' `------------------' | | v v .-----. .-----. >Spkr|---| Amp | | Amp |---|Spkr< `-----' `-----' Dialog SubWoof If you do not have "external processor" capability, but have a separate "record in" selector switch, you'll need to: - route the Surround Source into the "LD" or "VCR" input as usual, - set "record source" to select that input, - route the record-out jacks to the decoder, - route decoder front-out back into "AUX" or "TAPE2" or a similar unused line input, and - select "AUX" or "TAPE2" on the receiver/amp main selector switches. Another work-around is to: - feed the surround source (if you only have one) directly into the decoder Lt and Rt inputs, - feed the decoder "front" line outputs into the main (stereo) receiver or amp (AUX or other line-level in), and - feed the "center" "surround", "subwoofer" outputs directly to the secondary amps line-level inputs. Step 7.2: Pick a Decoder or Processor (or Receiver with integrated decoder). As far as brands and models, I cannot help you very much. My only exposure to surround has been via the Lexicon CP-1, which has both certified Dolby Pro Logic and a variety of other modes (although not THX - for that you need the CP-3). I can't really say whether or not the lack of auto-azimuth, use of ordinary Dolby or simple matrix decoding would be disappointing by comparison to Lexicon's all-digital Pro Logic. I would look for the following features (prioritized): * Pro Logic (adds less than $100 to new receivers nowadays). * Master volume control. * Auto-balancing on input (for programs recorded out of balance). * Adjustable rear-channel delay. * All calibrations from front panel and/or remote control. * Auto-test-tone program for calibration (more below). * Non-volatile storage of adjustable parameters (below). * Effects defeat (below). * Auto-azimuth to remove group delay of source program channels out of phase or independently time-delayed (as from SVA optical film tracks or when sharing a single DAC on LD). * THX certification. When you demo, I suggest starting at the top so that you have a standard to shoot for in a lower-priced decoder. Have the salesperson run through the setup procedure, and listen to a surround test disc (like Reference Recording's LD-101, "A Video Standard"). This will show you how much trouble the process is (or isn't) and more importantly, will ensure correct store setup. Far too often, I have heard simple stereo setups in stores that are out of phase. I estimate the chances of a correct surround demo at about 5%, rising perhaps to 50/50 at a "high end" store or "video salon". If you are considering getting an integrated receiver/decoder, I suggest getting ONLY a receiver with the Pro Logic or THX brand, as it may otherwise be difficult or uneconomical to upgrade a simple Dolby model later. There are some things to watch for, lest you end up with missing or duplicate components in the system and/or high "hassle coefficient": * Tuning decoder/processor input gain, balancing the outputs, setting surround delay, etc., are critical for acceptable performance. Does the decoder/processor have easy step-by-step instructions? Does it include a built-in pink noise generator for matching levels (or a separate tape/LP/CD with such a signal?); if not, consider getting a copy of the Reference Recordings LD-101 "A Video Standard" LD. * Does the decoder/processor or receiver supply its own surround channel amp(s)? If so, how many, and with what power? Is the power adequate for the speakers selected? Is the impedance matched to the number of speakers? * If a subwoofer output is provided, is it producing flat response, or does it incorporate a low-pass filter? What does the subwoofer itself require? Does the subwoofer include its own amp? How seamless is the response overlap between the subwoofer(s) and the bass response of the front and dialog (if any) speakers? * If your main speakers are driven by an integrated amp or receiver, does it have an "external processor loop" that allows separation of the pre-amp and power amp? If not, you may encounter complications in signal source selection and front/rear volume balancing. * Is there a single master volume control for all channels, controlled by the remote? Having that control duplicated on the front panel of the decoder/processor is a plus; a servo-driven ganged potentiometer is ideal. (The Lexicon's have only volume buttons , no knobs. The CP-1 has only remote volume. All have "mute".) * Are "effects" defeatable, allowing simple front-only stereo/mono? Can you bypass the decoder/processor altogether for critical ordinary stereo listening? * Does the decoder/processor have user-alterable presets for Dolby decoding, vendor-unique decoding, stereo ambiance enhancement and any other modes you will frequently use? Are the settings non-volatile (preserved thru power-off)? Finally, a feature to watch out for. If the decoder does not bear the double-D [)(] Dolby logo, find out why. The missing logo indicates that the vendor is unwilling to submit their design to Dolby for inspection, change requests, re-submission, etc. leading to an official approval. There are reasons why this might be (which also apply to THX): A. They focused their design on low-cost and/or time-to-market, and were unwilling to pay the royalty for using the Dolby logo, and/or put up with the certification delay. B. They don't like the Dolby spec, and think they have a superior decoding scheme. Of course, they could have done both their own and Dolby's. If the decoder (alone) sells for more than about $500, this is probably the explanation. C. Their decoder is too primitive and/or low in quality to pass Dolby qualification. It may also lack even simple processing, like surround channel delay. If the decoder is built-in to a receiver, monitor or other component, listen carefully. If possible, A/B-it against a quality stand-alone decoder. Step 7.3: Install, calibrate and enjoy your surround system. Unfortunately, many calibration programs jump from channel to channel, never turning on pairs simultaneously. It is very difficult to accurately set levels this way, particularly if you must leave the central seating position to make the adjustments. I use a sound pressure level meter (Radio Shack 33-2050, about $40), parked head-height at the listening center. Since only comparative levels matter, you can also use a microphone feeding any metered recorder. Adjust the record level (in PAUSE) to about 0dB, and set levels for all channels. Incidentally, if using "phantom center channel mode" (no center/dialog amp/speaker), it may not be possible to get the "center" channel's level to exactly match "left" and "right". And since someone asked... The setup for "master volume" on an outboard processor (like a Lexicon) is: - set input gain on processor so that 0dB test tone from input media on main non-variable source (e.g. LD player) hits max on processor - set processor master volume to max - adjust main amp gain for "loudest level you'll ever listen at". "The Abyss", "The Dream is Alive" and "T2" are good setup sources, as all have below normal digital track amplitudes (so that peaks don't clip). - reduce master vol to a nominal level and balance surround amp gain - stick "No Touch" labels on amp gains - use only processor master volume for system volume Step 7.4: Use of a surround system for music. If after reading all of the above, you suspect that there is an awful lot of processing being done on the original stereo signal, you are correct. Do you want to have all that switched on when playing ordinary stereo music on the same system? I suggest "no", unless the music was specifically recorded for surround (as a few CDs have been recently). When playing music on my system, I switch from Pro Logic mode to "small hall ambiance". If your decoder doesn't have any alternate modes, it is doubly important that it have an "effects off" mode. You may not like what passive matrix, much less Pro Logic or THX, do to non-surround stereo music, particularly if you are critical audiophile. 8. Some matrix surround titles _____________________________________________ If you get an opportunity to demo surround, make sure you are using source material that is worthy of the system. Be advised that: a. Many video sources with surround sound aren't so identified on the media jacket. Sometimes your ears or the "Dolby Stereo in Selected Theatres" that appears in the trailing credits are your only clues. b. Conversely, the appearance of "Dolby" in the trailing credits is no guarantee that a Dolby-ized stereo master was used for the video release (although it is rare that this is not the case). c. Even if "Dolby" or "surround" appears on the jacket, the effect may be less than dramatic, and may be largely ambience and echo. The following laser video disc (LD), from the IMAX movie, is what I use for surround demos: The Dream is Alive {CAV} (Ferguson, 1985) Lumivision LVD9019 The following two laser disc titles also have very effective surround programs. I cannot vouch for the non-Criterion pressings of "Ghostbusters". I cannot tell you anything about tape editions. "The Abyss" is also the first THX-certified LD. The earlier 1561-85 theatrical edition also has an excellent soundtrack, but is not THX certified. Empire of the Sun {CAV} (Spielberg, 1988) WB 11844 Empire of the Sun {CLV} (Spielberg, 1988) WB W11573 (P-51 attack scene) Ghostbusters {CAV} (Reitman, 1984) Criterion CC1181L Ghostbusters {CLV} (Reitman, 1984) Criterion CC1182L ("Slimer" scene in hotel) The Abyss: Special Edition (CL/AV) (Cameron, 1989/93) Fox 1988-85 There are several test discs available for calibrating systems (and verifying that your dealer's demo system is correctly set up). The most easily available is: A Video Standard (Kane, 1989) RR LD-101 The following LDs also have surround programs. Although they are somewhat less dramatic than those above, they are more pronounced than several other "surround" discs I have examined. Ben-Hur {the current letterbox edition} (Wyler, 1959) MGM ML101525 Die Hard (McTiernan, 1988) CBS/Fox 1666-80 Dragonslayer {letterbox} (Robbins, 1981) Bandai LA098L14046 Dragonslayer {cropped} (Robbins, 1981) Paramount LV1367 LadyHawke {current letterbox} (Donner, 1985) WB 12370 LadyHawke {cropped} (Donner, 1985) WB 11464LV Ruthless People (Abrahams,Zucker, 1986) Touchstone 485AS Star Trek III - The Search for Spock (Nimoy, 1984) Paramount LV1621 Star Trek III - The Search for Spock[WS] (Nimoy, 1984) Paramount LV12954 Star Wars {letterbox} (Lucas, 1977) CBS/Fox 1130-85* The Witches of Eastwick (Miller, 1987) WB 11741A/B One title to avoid: Dolby Technologies: How They Work Pioneer 05458 Although a useful tutorial, it contains NO demo material. Many of the articles referenced below also list demo titles. * This is the new letterbox CLV edition with digital sound. 1130-80 (CLV) and 1130-84 (CAV) are cropped, although probably still surround. A THX certified edition is expected in 1993. 9. References ______________________________________________________________ Available free from: Dolby Laboratories 100 Potrero Avenue San Francisco CA 94103-4813 Write for: "Dolby Pro Logic Surround Decoder - Principles of Operation" "Dolby Surround - a listener's guide" "Heard Any Good Movies Lately?" (a list of Dolby Stereo films) "Question about Dolby Surround" "What is Dolby Surround" ____________________________________________________________________________ 10. Some related surround traffic from other contributors (>) & my replies. > Whenever possible, I use headphones when watching a movie on laserdisc. > I use a good pair of Sony headphones, the enclosed kind that kill any > external sounds, model MDVR-6 or something like that. Some movies are > truly awesome this way. How does [matrix] surround sound compare to > headphones, in terms of the listening experience? I don't watch video with phones on, but for the purposes of answering this query, I got out the MDR-V6's and gave it a try. Generally: Normal stereo speakers: Wall (or "stage") of sound in front of you. Surround speakers: Field of sound all around you. Stereophones: Line of sound between your ears. The headphone experience is one of having all the audio happening inside your head. This is certainly "different" than stereo speakers. Whether or not it is "better" is a matter of taste and exposure. On an ordinary stereo program, once the brain has learned to associate the sounds with the visual action, there's not much difference between phones and speakers. Stereo video soundtracks often share the same problems that musical works (esp. early stereo recordings) have when heard via phones. If individual sounds are fed into the mix (say, the left channel) without bleeding some reverb (into the right), they sound artificial. This is all too common on movie sound. The effects sound pasted on, and not part of the program. In a speaker (or theatre) setup, this is not a problem, because the room adds the necessary blending and ambiance. On undecoded matrix surround programs, however, I noticed two more things: 1. The anti-phase encoding of surround information can be distracting. 2. There is no sense of "front" and "back". On (1): An effect that is supposed to be a sound moving from front center to rear center is a sound that starts in-phase and shifts to anti-phase. If you are sensitive to phase (as I am), it is slightly annoying. The "location" of the sound shifts from head center to "both ears at once". It does not move front-to-back. On (2): Given a stationary head and sound source, the way that the human auditory system determines "front" and "back" is through subtle amplitude, phase and group delay differences between the sound at each ear, plus frequency contouring, local reflections at each ear, and bone conduction. The head and outer ear re-shape the sound spectrum, based on the direction of the source. This is why, in real life or in a surround speaker setup, sounds from behind you are experienced as coming from behind you. Real life surround is therefore completely different than matrix/Dolby encoded surround. The encoded signal is not at all naturally directional to the headphone listener. Building headphones with 4 speaker assemblies doesn't help much either. Vendors (even Stax) try this from time to time with little success. It is possible to largely re-create a 3D listening experience for headphones via a special sub-species of stereo known as "binaural" recording. This requires a two-track recording typically made using an anatomically correct human head dummy, with microphones in the ear canals. The resulting work preserves all the directional cues imparted by a human anatomy, although not the exact same anatomy that your brain grew up with. Obviously, the work must be heard via earphones (optimally, the intra-aural in-the-ear type, so as not to add any further anatomical processing). Some radio dramas (including Stephen King's "The Mist") have been recorded binaurally. Very few musical recordings are made this way, and to my knowledge, NO video programs have binaural tracks. The apparent bottom line: * For ordinary stereo programs, headphone listening is an acceptable alternative to speakers, perhaps even preferrable, depending on your tastes and listening environment constraints. * For surround programs, headphones are no substitute for speakers unless the program is binaurally recorded (and no video material is). Bob Niland /-------------------------------------------------------------------------/ > I think these laserdisc titles also have at least some great surround > utilization: > Back to the Future > E.T. - The Extra-Terrestrial > Poltergeist > Star Trek II - The Wrath of Kahn > Young Sherlock Holmes > These titles at least have great overall sound and may have good surround > too (I can't remember for sure about the surround part): > Predator > Robocop > Star Trek IV - The Voyage Home > Starman > 2010: The Year We Make Contact - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - > The absolute *BEST* disc for surround sound is _For All Mankind_. The > launch of a Saturn 5 can't be matched, and stage separation ... WOW! [Note - A correspondent who was working at the Cape at the time reports that although impressive, the "surround" sounds on this disc are completely faked. - rjn] > A low cost way to get into surround sound is with the Radio Shack decoder. > You can get it plus a pair of speakers (Minimus 7's?) for about $200. If > you switch it to use all of it's amp power for the back channel speakers > and use your hifi for the front, it works pretty well. It only has a few > watts of power (15 or so?). My front channel is a pair of Altec-Lansing > 15" Voice of the Theater speaker systems driven by a 130 watt per channel > amp. /-------------------------------------------------------------------------/ * For surround programs, headphones are no substitute for speakers. >> Would one of those "quad" headphones work for this case? (I remember >> the four speaker headphones being sold at the peak of the quad fad.) > I suspect it depends on how fussy you are. I have a pair of Koss quad > headphones from the old quad days. My father was into that non-fad and I > ended up with them, along with his old receiver. The headphones did > produce a kind of four channel effect, although not the same effect that > was produced by the speakers. In the same manner, I can decode surround > sound with the QS matrix decoder in the receiver, but sounds tend to > wander about the room and separation is not as good as in the theaters. I > would guess that I have something nearly as good as the low end surround > sound decoders, but not nearly as nice as the new pro logic decoders. /-------------------------------------------------------------------------/ re: >> * For surround programs, headphones are no substitute for speakers. > Would one of those "quad" headphones work for this case? (I remember > the four speaker headphones being sold at the peak of the quad fad.) Well, as I said in the quoted article: >> Building headphones with 4 speaker assemblies doesn't help, either. >> Vendors (even Stax) try this from time to time with little success. Let me elaborate. Keep in mind that this is just a "thought experiment". I have no actual experience with 4-element headphones. The directional cues that occur to me are: * Ear effects: (the "pinna" is the external part of the ear) directional response envelope local reflections in the pinna * Head effects: (comparing one ear to the other) diffraction of sound when one ear is masked. arrival-time differences response/amplitude differences * Body effects: feeling louder and/or lower frequency sounds on one side * Conduction: A fair amount of mid- and low-frequency energy reaches the inner ear via bone conduction. Dummy microphone heads are even being made with false skull bones for this reason. * Environment: With "real" sounds, the location of the sound with respect to the listening environment is stable as you move your head. Even when you are "still", your breathing alone is moving your head enough to cause several degrees of phase difference at voice frequencies, not to mention shifting your position in any room standing wave patterns. With 4-element phones, assuming that the "rear" elements are actually at the rear of the housing, the only cues preserved (maybe) are the pinna effects. The head, body, conduction and environment effects are totally lost, because the sound from any single element is heard only in one ear, and the relationship between ears and sources is constant. I'm not even sure that the pinna effects are detectable, given that the sound from any element is bouncing around in the phone housing as well as off the ear. I suspect that it is possible to detect a difference between 2-element and 4-element operation of quad-phones, but I'll bet that quad-phones provide a very unsatisfactory surround experience, beyond allowing you to employ the decoder to remove the anti-phase and Dolby-B encoding of the raw stereo signal. Regards, 1001A East Harmony Rd. Bob Niland Suite 503 Ft Collins CO 80525