----- Experience in your own room the magical nature of stereo sound -----



What's new


LX - Store




with Fitz





The Magic in 2-Channel Sound

Issues in speaker

Dipole models

Active filters

Amplifiers etc





Room acoustics

Stereo Recording and Rendering

Audio production




Your own desig












dipole speaker

Three-Box active
system (1978)



& Room




Sound recordings


Other designs

My current setup

About me

Site map




Digital Photo


Sea Ranch


My Daughter
the Jeweler


What's new


LX - Store


with Fitz






Surround sound

1 - Ambience recovery from a 2-channel CD
2 - Refinement of ambience recovery
3 - SACD and DVD-A use of the center channel
4 - A few SACD and their need for a center speaker


1 - Ambience recovery from a 2-channel CD

Many CDs, especially those of classical and jazz performances, preserve a sense of the acoustic environment in which the recording took place.  On a good two channel playback system this ambience information is perceived as depth and width of the sound stage. The ambience information can be extracted and fed to rear loudspeakers. Doing so increases the apparent acoustic space around the listener and significantly enhances the realism of reproduction and its enjoyment. 

In its simplest form of implementation the difference signal between left and right channels is taken, then delayed electrically by 15 ms to 30 ms and applied to two rear speakers. A refinement of this approach is described by Doug Rife, designer of  the MLSSA loudspeaker and acoustic analyzer.

I have chosen a Lexicon DC2 Digital Controller for my surround system. The DC2 is a preamplifier with remote volume control that can take analog inputs as well as digital inputs from CD and DVD  transports. It provides up to seven outputs for left-center-right front channels, left-right side channels, and left-right rear channels. The unit is immensely flexible and handles not only all home theater modes, but also provides a variety of music specific modes, including ambience recovery from stereo recordings.

My interest is in high accuracy sound and music reproduction. I am not using a center channel, because the main Audio Artistry Beethoven-Elite speaker system provides excellent imaging and a center speaker does not fit with the decor. I wanted the two side speakers and the two rear speakers to be visually and acoustically unobtrusive. Ambience information is played back at a lower sound level, so that small size speakers can be used, but they should reproduce frequencies down to at least 60 Hz. Since the speakers need to be at ear level height when seated, I made the stand part of the speaker enclosure. The speaker cones point upwards for wide and uniform sound dispersion. Construction details are provided on the Surround Speaker page in the Build-Your-Own section of this website.

The photos show the speaker setup in my living room. Visible are the front and the two side speakers. The two rear speakers are outside the first picture. 


Most of my listening is done from the center chair at the bottom of the above picture. For very critical listening I may turn around the chair closest to the front speakers. The side speakers are aligned with this chair. None of the four surround speakers are localized as separate sound sources, even when sitting on the sofa or in the right chair.

The left and right rear speakers are about 23 feet away from the front speaker. All four surround speakers are visually quite unobtrusive. More so than the photos might indicate.

The surround speakers face upward to generate a diffuse sound field in order to prevent sharp imaging of side and rear sounds. This is not a system for home theater, where sound effects are panned into side and rear channels. You can see from the size of the TV that video plays a minor role in this house. 

Almost all of my music is on CD, hardly anything on DVD-Video. "Music Logic" is my default mode for the Lexicon DC2. It can be enjoyable to listen to DVD-V surround sound on this system, if the sound effects in the side and rear channels are kept to a minimum. In the future I might add DVD-Audio and SACD capability, but compared to a high quality CD the sonic improvement from 24bit/96kHz high resolution audio is surprisingly subtle. Unless the loudspeakers and the whole signal chain are of the highest accuracy, probably very little is gained from the new formats. The greatest and most readily realized benefit in realism comes from the addition of two ambience channels to a conventional stereo system. Increasing this to four ambience channels provides a smaller gain.

For further study I recommend David Griesinger's paper "Surround: The current technological situation". 

The AES Technical Document "Multichannel surround sound systems and operations" is heavily influenced by home theater surround sound practices and needs to be updated for the new high resolution multi-channel audio formats. See, for example, my comments about center channel, subwoofer, loudspeaker performance and listening conditions. 


2 - Refinement of ambience recovery

After the ORION loudspeakers had replaced the Audio Artistry Beethoven-Elite speakers in the system above I discovered quickly that I preferred the normal 2-channel stereo playback over the ambience recovery surround sound that I had used previously for many recordings. The ORION gives such a natural presentation of the soundstage that the Music Logic mode of the DC-2 actually detracted from the listening experience. It affected imaging, sound color and bass response. Consequently, I did not use the rear speakers for many months.

There is a vast volume of well recorded 2-channel material on CD available. Since the ORION does such a satisfying job of playing it back, any ambience recovery system must not change the signal to the two front speakers. This can be assured by using a separate processor which derives the various surround signals from the CD player output. Of those only the signals for the side and rear speakers are then used. 

This approach works exceedingly well. It adds reverberant sound to the room and makes the walls disappear. The side and rear speakers cannot be localized, because the signals feeding them are decorrelated and the speakers have very wide dispersion. The rear speakers contribute little audibly, except that without them the side speakers would have to deliver higher output. Since all speakers use 4" drivers and are configured as full-range speakers in the A/V receiver, it is beneficial to share the load between them. The surround speakers need to have some bass output to simulate ambience. 
The effect is an increase in fullness of sound, without distracting from the imaging, and actually giving instruments more body. One must be careful not to set the volume level of the side and rear speakers too high. Their contribution is really noticed when they are turned off and the sound turns almost flat. It takes a while to find the correct Volume 2 level relative to Volume 1, but once found it is always the same dB number offset between the two volume settings, regardless of the recording. A surprising benefit of this system is that I reduce the volume level to the front speakers by 5 dB compared to the settings that I used in 2-channel mode. Yet the subjective loudness is unchanged and the sound is even richer. 
The setup above is slightly inconvenient because any volume change of the Preamp requires a corresponding volume change of the A/V Receiver to keep the surround speaker output in fixed proportion to the front speakers. If, instead,  the A/V Receiver takes its 2-channel analog input from the Preamp after Volume 1 has been set, then Volume 2 can stay at a constant setting to provide just the right amount of ambience.     

If you have a good 2-channel sound system, then here is a way to add to its realism. It would seem to me that the A/V receiver is not very critical, as long as it has the proper processing capability, since only background sounds are derived from it. Give it a try!


3 - SACD and DVD-A use of the center channel 

I had acquired the DPR 1001 Audio/Video receiver to experiment with SACD and DVD-A playback, and have learned a bit more about the use of the 6 discrete channels that can be recorded in these formats. While the two storage formats have been well defined at the bit level, there seems to be no universal agreement about what to record to the different channels, nor about the loudspeaker layout to play them back. In particular, I read that some recording engineers would rather not use a center channel for music, while others insist on it. Chesky, on the other hand, proposed 2 front speakers, 2 elevated side/front speakers and 2 rear speakers. This takes up the available 6 discrete channels. A universal player like the Pioneer DV-45A has 6 analog audio outputs for SACD or DVD-A multichannel sound recordings, which are labeled Front L/R, Surround L/R, Center and Subwoofer. The de-facto standard for multi-channel music playback seems to be the Home Theater 5.1 speaker setup. The multi-channel music demonstrations that I have heard from Sony, Meridian, JBL and others used a 3+2+.1 setup and have been disappointing in comparison to the 2+2+2 setup by Chesky/Muse/Avalon, which demonstrated DVD-A to the High-Resolution Audio Workshop during the 106th AES Convention in Los Angeles. The large number of installed 5.1 home theater setups is clearly driving the music recording industry now.

Most of the SACD and DVD-A discs that I tried contained a recorded center channel track. If this track is not used during playback, then the sound stage has a glaring hole in the center. The DV-45A can be configured for "no center speaker". In that case the center channel information is mixed equally into left and right front speakers and a stable phantom image is generated between the two speakers. But, when the sound stage is compared to the sound stage from the 2 track mix, that is provided on the same disc, then the 2 track version presents a more continuous image spread. Thus, I have come to the tentative conclusion, if a center channel has been recorded, it should also be played back via a center speaker, and not as a phantom center. This then demands a suitable center speaker to get the most from the new discrete multi-channel recording formats. 

In order to verify my tentative conclusion I will use the top section of the ORION, mounted on a stand, as a "small" center speaker. I do not see a need for a full-range or "large" center speaker, since the frequency range above 100 Hz is all that is needed to establish directional cues in this playback situation. In my room the center speaker will only be placed in position, when I actually listen to a SACD or DVD-A recording, since I have no esthetically satisfactory solution for a permanent center speaker. My current surround channel speakers appear to perform adequately for the type of music I listen to.  

My center speaker prototype uses ORION midrange and tweeter channel electronics, except that the 120 Hz highpass filters and midrange delay circuits are bypassed. The DPR 1001 is set up for a "Small" center speaker. The cabinet dimensions follow the drawings in the ORION documentation, except that the dress panel is lengthened for cosmetic reasons. I decided not to build a fully finished unit before finding out whether I would keep a center speaker. 

I have now spent a good amount of time listening with and without the center speaker. Actually, this speaker blends in its tonality so well with left and right speakers that I often had to walk right up to it in order to tell whether it was on or off. My conclusions after listening are not completely what I expected. 
First, the one time where the center speaker was of benefit occurred during the Mahler Symphony #4 recording of the San Francisco Symphony. The words sung by the soprano in Davies Symphony Hall were slightly better to understand. 
Secondly, recordings that did not use a center channel (Chesky) were not recognizable as such. Only by touching the speaker cone did it become apparent that I was listening to a phantom center. 
Third, switching between actual and phantom center on recordings that used the center channel did not produce significant audible differences. If anything, I preferred the phantom center and the associated smoothness of soundstage spread.  
Fourth, the side speakers should have some bass output and not be "Small". The sense of acoustic space and ambience depends strongly on the frequency range below 500 Hz. I used two of my surround speakers on each side in "Large" mode to keep distortion low. The distortion became noticeable on large choral pieces. 
    At this point I am not convinced that a center speaker in my system is worth it, even if it is placed out of sight most of the time. Sitting in front of such speaker is visually highly distractive and its benefit for multi-channel sound is marginal, given the satisfying sound stage that is set up by the ORION. 



4 - A few SACD and their need for a center speaker 

A number of people wrote to me about the reissue of RCA Red Seal Living Stereo recordings in SACD format and the necessity of a center channel speaker for playback. A number of the original recordings were made with 3-track analog tape equipment for playback over left, center and right front speakers. They were mixed down to two channels for LP issue and contained some wonderful classical music performances. Today the original, unprocessed three tracks of these historical recordings can be heard via SACD. 

I bought three of the ten available SACD:
1 - Chopin: Ballades and Scherzos, Arthur Rubinstein
2 - Mussorgsky: Pictures at an Exhibition, Fritz Reiner, CSO
3 - Tchaikovsky: Piano Concerto #1, Rachmaninoff: Piano Concerto #2, Van Cliburn

The following are first impressions and observations from comparing playback of the 3-channel recording to the 2-channel CD down-mix which is also on the SACD. I used the center speaker that is described above and which was entered into the AV receiver setup as 'Small' for bass management.

1 - The CD tracks are superb on the 'Rubinstein'. I had listened to them several times and weeks had gone by before I finally took the center speaker out of the garage and set it up again. Well, the SACD and its center channel added a real sense of solidity and body to the piano. Definitely worthwhile. This motivated me to find more of the RCA recordings and I got discs 2 and 3 from the local record store.

2 - The recording quality of the 'Pictures' is far below the 'Rubinstein'. I have not listened to the whole piece. It did not promise to be enjoyable. Track 7, for example, has horrendous modulation distortion. I moved on to disc 3 with trepidation.

3 - The 'Tchaikosky' immediately struck me as a clean, natural recording and curiously I switched to the 'Rachmaninoff', which is my wife's favorite music, thinking that I could give her a demo she would enjoy. I was surprised about the difference in sound stage. The three microphones must have been very widely spaced. The string sound is quite unnatural. On the CD cut the phantom image of the piano is swimming in space. The center speaker locks it in place and anchors the sound stage on the SACD cut. I suspect the weakness in image stability is due to small amounts of wow and flutter from the analog tape recorder. On such recordings a center speaker stabilizes the image, but also tends to draw the center stage unto itself, rather than leaving it behind the speaker. Preferably I listen with my eyes closed in order not to get my sonic mental image distorted by the visual presence of the speakers.

All three SACD benefited from using a center speaker, though in different degree. I then switched to two recent SACD recordings:
4 - Orff: Carmina Burana, Atlanta Symphony Orchestra and Chorus, on Telarc
5 - Mahler: Symphony #4, San Francisco Symphony

The first impression after the RCA recordings was one of greater stability and solidness of sound. Also the low frequency end was more realistic in its character. The center speaker seemed to contribute less to the image other than to widen somewhat the 'sweet spot' for listening. It also had some benefit in making the soprano in the last movement of the 'Mahler' more intelligible, though still not enough to follow the German text. But that is not different from a live concert. With these two recordings, though, the center speaker was visually in the way, because the sonic event was happening behind it. It is much more easy to ignore the left and right speakers as just sitting there and not being part of the sound picture when I listen to a good CD, but having three speakers looming there like a fence, though seemingly transparent, is disconcerting to me and I must close my eyes not to get distracted. 

Someone suggested to me to get the Clark Terry, 'Portraits' recording on Chesky Records to see if the trumpet can be reproduced with a 1 inch dome tweeter crossed over at a low 1440 Hz as I do for the ORION. This 1988 recording by Bob Katz has been reissued on SACD. The center channel is not used and so in this respect there is no noticeable difference between the CD and SACD tracks. The sound stage is that of a typical studio recording as opposed to the Bucky Pizzarelli, 'Swing Live' SACD's nightclub atmosphere, which to me is more pleasing. The trumpet does indeed tax the tweeter more than on most recordings. For example, playing the 'Autumn Leaves' selection of the 'Portraits', track 3, at respectably loud sound levels demands around 20 W peak from the tweeter's dedicated power amplifier. The trumpet is firmly locked to the right speaker, the piano to the left speaker while the bass and drums are located between the speakers with some depth and separate from them. The trumpet sound seems slightly more bodied on the SACD cut. I am quite certain that the trumpet sound character is not modified by tweeter distortion. It appears level independent. Also, I have listened to many other trumpet and brass instrument recordings (e.g. Wynton Marsalis) at high levels without any suggestion of tweeter stress. A 20 W requirement for the tweeter can be quite demanding, though, if a single power amplifier must drive a loudspeaker through passive crossover networks. Then these 20 W must be delivered on top of the simultaneously occurring and usually higher power draw from midrange and woofer drivers. This can lead to clipping distortion and even destruction of the tweeter unless the power amplifier has very high power capability and headroom.
The two mentioned jazz SACD contain fun music and are well worth owning. They are good examples of surround sound without a center channel, where the main purpose of the surround channels is the creation of acoustic space and not the placement of discrete instruments to the side or rear of your room.

My overall assessment for the need of a center channel with the ORION has not changed. If I sit for my critical or fully involved listening in the stereo 'sweet spot', the apex of an equilateral triangle, then the contribution from the center speaker - as compared to a phantom center - is marginal. I only found it worthwhile for the historic RCA recordings. If I listen from a greater distance, then the reverberant sound in the room becomes too dominant for the center speaker to make a significant difference. If I sit off to one side, then the sound is pulled less strongly to the near speaker, but that does not turn it into a place where I would sit to listen attentively, because the sound stage is too distorted. 

A further observation may illustrate the pro and con of a center speaker. I was listening to the wonderful SACD recording of Mahler's Symphony #2 by the San Francisco Symphony - of an outstanding performance which I had actually attended - and I noticed that the sound of the large chorus became more aggravating during loud passages than I am used to from 2-channel choral recordings and live performances. When I disconnected the center speaker and set the receiver to produce a phantom center, the sound and sound stage became smoother. The slight loss in intelligibility of the center stage soloists was compensated for by a generally more musical and believable image presentation without lump.

In 2-channel playback there is crosstalk from the left speaker to the right ear and from the right speaker to the left ear, which results in comb filtering due to time delays. The center speaker adds additional signals at the ears that are not necessarily time coherent with the other two speakers' signals, because of the overlap in the three microphone pickups. This may have caused the effect that was observed.  

For the time being I will leave the center speaker in my listening room rather than return it to the garage. I have moved it off to the right where it is mostly out of sight. I can easily place it back in the center should the need arise. But I suspect a center setup will not last. 

Indeed, it did not last, and  I prefer the ORION without it.




What you hear is not the air pressure variation in itself 
but what has drawn your attention
in the streams of superimposed air pressure variations 
at your eardrums

An acoustic event has dimensions of Time, Tone, Loudness and Space
Have they been recorded and rendered sensibly?

Last revised: 02/15/2023   -  1999-2019 LINKWITZ LAB, All Rights Reserved