Qsound

July 28, 2019
 by Paul McGowan

Qsound is a technology intended to reproduce surround sound with only two speakers. The results can be rather spooky. Sonic images appear in 360˚ patterns around the listening position. That’s quite remarkable given their two-channel source. The classic example of Qsound is Roger Water’s Amused To Death.

If you’ve never heard this work on a decent stereo system it’s worth the purchase just to experience surround sound from your stereo.

There’s a catch. Qsound only works for one person at a time. That person must be seated dead center between the speakers to appreciate the surround effects. Move off-center even just a bit and we get standard stereo again.

According to the Wikipedia article: “QSound is essentially a filtering algorithm. It manipulates timing, amplitude and frequency response to produce a binaural image. Systems like QSound rely on the fact that a sound arriving from one side of the listener will reach one ear before the other and that when it reaches the furthest ear, it is lower in amplitude and spectrally altered due to obstruction by the head. However, the ideal algorithm was arrived at empirically, with parameters adjusted according to the outcomes of many listening tests.”

That’s correct but a much simpler way to understand what’s going on would be to focus on phase differences, which is what Mathew Polk and his team at Polk Audio did years before in their SDA products. Using cross-connected stereo speakers and an additional set of midrange drivers, a bit of out of phase left was fed into the right channel’s “dimensional” drivers, and vice versa—the stereo width was extended beyond the room boundaries because the left ear got a slightly delayed out of phase signal from the right channel, and so on.

The same technique called Interaural Crosstalk has been applied in multiple ways: Ambiophonics, Carver’s Sonic Hologram generator, among the many.

Essentially, all these systems were precursors to what today we use to direct sound in big PA systems so every seat gets the same image and at the same amplitude.

There’s a lot more than just two dimensions possible with 2-channels of audio.

Just ask any audiophile about depth, width, and height for starters.

Subscribe to Paul's Posts

22 comments on “Qsound”

  1. Amused To Death is a great album and the first time I heard track 3 was on a set of Quads, and the piano floating above the right speaker was so realistic to me that I new this recording and the Quads were special. I closely get it with my old AR-58’s, but not quite.

    I am still quite happy with well recorded and mastered stereo recordings and to me I use what I know are well recorded piano recordings to determine if a speaker is accurate and can try and recreate the sound of a real piano in space. If it does, I’m in.

  2. One classic example of Qsound is Madonna’s “The Immaculate Collection” which was done two years before “Amused to Death”!
    Other s are “Power of Love” by Luther Vandross, “The Soul Cages” by Sting and “Spellbound” by Paula Abdul – all of them before RW.

  3. To the uninitiated listening to a Qsound mixed album at a dealer speaker demonstration could well be very misleading.
    The Pink Floyd album where I think the phase technique is convincingly applied is on the vinyl boxed set of Pulse, it perhaps gives an exaggerated sense of scale (I wouldn’t know as I wasn’t there at the time) but it certainly is exciting for the listener. I would imagine some purists frowning at the use of such techniques but to me its harmless fun.

  4. Fascinating read from Paul, however, for the pieces of music where ‘everyone’ involved from creators, artists to producers want to fill a room with precise location in my opinion SACD 5.1 is amazing.

    Before Paul or others quotes only the stereo pair in even good surround sound systems can be good for music I beg to differ. There are several companies now like Monitor Audio that make amazing sounding surround systems. Yes, centre and back speakers may not be huge like the the stereo pairs, but they are certainly getting bigger and sounding better. With a high end AV Amp even stereo sources can sound very spatial (back front, as well as up down and left right) and help extend the listening position for more than one person in a sweet spot. The simple multichannel stereo ‘processing modes’ for music have improved massively, helping ‘normal folk’ without IRS V and similar massive (in cost and size) Systems achieve more than was possible some years ago. As Paul and PS Audio is using for the new ANx Speakers these AV Amps have DSP processing to ‘help normalise the room’, but for all speakers. The high end ones are typically ‘light touch’ as Paul likes in terms of frequency response, even tweakable by the user.

    Much to my surprise I’m listening to 5/6/8 speakers, intended for AV ‘movies’ and Multichannel Music, for stereo music sources as they just come alive when compared to just the larger stereo pair. The IRS V and the ‘near like’ would I’m sure sound ‘better’, but that’s not an option for almost everyone. So, for me, this is my Qsound in that ‘processing and more speakers’ makes the music and spatial sound much better but more natural than Qsound with only two speakers.

    As mentioned at the top I still get the most amazing experience using mostly 5.1 SACDs like Amused to Death and a few other seminal works like Dark Side of the Moon, In the court of the Crimson King(a DVD 5.1 lossless 24/96), Wish you were here, Tubular Bells… Even solo artists like Nora Jones with Come away with me – close your eyes and she’s in the room with just her piano and not my room but hers as recorded. There’s no pianos flying around the room just a new great sounding room with her and the piano centre stage, wonderful.

    Do I still listen to just the stereo pair, very rarely but sometimes as the recording is not quite good enough as the system is probably too revealing. Am I still interested in improving the main stereo pair, most certainly, and very interesting in the new PS Audio AN3, or perhaps needs to be a just a bit smaller for me – an inch or 3 off the width would make this more ‘home’ friendly.

    1. I have to agree with Alan completely. I started my surround sound journey in the late 60’s. Back then it was called quadraphonic sound. It was plagued by many competing formats SQ, QS, CD-4 and a few others. SQ ended up the dominant format with the Sansui QS vario matrix system a distant second. These were matrix quad formats with the Lr and Rr folded into the front channels, (I’m being very simplistic technically), and then decoded into 4 channels with the appropriate decoder. The early decoders were terrible, little or no separation, and should not have been released.

      Then SQ came out with their full logic decoder and Sansui with vario matrix. These decoders were capable of around 20 dB of channel separation, quite acceptable. Unfortunately by the mid to late 70’s quad was dying a slow death. But then came the Tate Directional Enhancement System for SQ quad. This system was capable of 40 to 50 dB of channel separation. Again, unfortunately this didn’t save quad.

      Audionics of Oregon developed the Space and Image Composer which utilized the Tate chips. This was a remarkable unit (and still is the standard) but just to late and the company went under. Fosgate then came out with their own Tate decoder which also was a great unit. Both these units were also very excellent surround synthesizers.

      Anyway, to get to my point. There is a company in Australia, Involve Audio, which has developed a product called the Surround Master v.2. This unit is capable of decoding SQ and QS formats better than the Tate units. It is also capable of surround synthesis, which it is mostly touted for, that will blow you away, of course depending on source material, some better than others. It also does 2 channel only synthesis.

      I happen to have several hundred SQ and QS albums, because I stated my collection way back when. They sound great on this unit. Stereo synthesis is truly remarkable. I also use the new Dolby and DTS upconverters. To me surround is just better than plain stereo. If you are interested in remarkable surround synthesis or you are an old timer with some SQ and/or QS records you need to check this out:
      http://involveaudio.com/product/surround-master-2-involve-decoder/

  5. In the early 80’s, i was in the audio biz, and we sold Carver including the Holograph. I got some training on how to set it up and the speakers had to be very carefully chosen for it to work 100%. We ended up using something from Infinity called mini-monitors or something like that. they had a metal grill and were pretty small. In addition, your head had to be perfectly positioned as Paul said. Even an inch or two could make a difference. Anyway, to this day i have never forgotten what happened and i have heard Amused to Death as well. We played a cut from I Robot by Alan Parsons, and when a chorus came on, they were behind me. THERE WERE NO BACK SPEAKERS AT ALL. Yet i heard the chorus coming from behind my head. It was truly amazing.

  6. Did anyone ever have a pair of Carver Amazing Loudspeakers? I bought a pair of those monsters with my first paycheck out of college, before I moved out. My Mom almost messed herself when I brought them home. They were the most power hungry speakers I have ever owned, but I loved them at the time. I also owned 2 1.0T amplifiers running in mono to drive them. It was really fun times going to our local dealer that carried Carver stuff back then.

    My favorite sound was the local store that had Magnepan Tympani speakers and Audio Research gear. The Carver stuff brought really cool technology and sound at a lower budget…which I needed at the time.

  7. Like Alan-ADT ,I have been enjoying 5.1 SACD’s. I will have to check out Q sound. Paul I know you are a big fan of Welcome to the Machine by Pink Floyd,do yourself a favor and listen to it in 5.1. The Doors Riders on the Storm is another delight as the storm is all around you.

  8. The goals and methods people devise to create greater envelopment, more concert like sound is what interests me most in audio. This was my one invention based on my own analysis and I like to see how others have seen and tackled the problem. There are a number of approaches and methods and this appears to fall into one category I’d call acoustic cross-channel phase cancellation.

    The one I’m most familiar with is Ralph Glassgal’s Ambiophonic sound. It seems to be the genesis of this approach. Other cross channel phase cancellation systems were purely electronic and Ralph appears to have invented one of those too, his separation control invented when he worked for Fisher Radio. The concept is used to increase channel separation and it or something like it seems to be incorporated into every home theater receiver in the “wide” mode. This improves sound stage width to get the maximum out of the three front center channel, left channel, and right channel arrangement.

    Electronic phase cancellation is nothing new. I first encountered it studying matrix method of creating stereo FM channel demultiplexing when I was about 13 years old and actually built one from a Lafayette Radio Kit. The Zenith method adopted by the FCC an still used today in the US and around the world had to be compatible with monophonic FM radio. In addition to the 50 to 15 khz L+R mono FM signal, an L – R signal is transmitted at an upshifted outband frequency on the carrier from 23 to 53 khz. The L-R signal is separated at the FM tuner detector output and added in phase to get the L signal and out of phase to get the R signal. Proportions are adjusted for maximum channel separation.

    Cross channel stereo separation in this class of audio enhancement device is performed with sound fields using the same principle. When I first heard about it I was surprised that it worked at all. All things considered it works fairly well in a limited place in space where the effect is maximized. Its goal is to get your left ear to hear only what comes from the left speaker and your right ear to hear only what comes from the right speaker. To do this a time delayed (phase shifted and inverted) signal from the right channel is sent to the left speaker and a time delayed (phase shifted and inverted) signal from the left channel is sent to the right speaker. The time delay is necessary because the distance of travel and therefore time of arrival is different for the two speakers. It’s in the microsecond range. Originally before this idea, Ralph actually put a wall between the speakers extending out to the listener and you had to put you head against the wall.

    Ralph’s time delay circuit uses a recursive arrangement where the delayed signal is fed back through a loop. He wouldn’t or couldn’t tell me why but he said it worked better. The time delay, gain of the recursive loop (rate of attenuation) and overall signal gain are adjustable to optimize performance. These systems work best where there are no room reflections to degrade the effect. I listened to this system on two occasions at Ralph’s house. The effect was interesting. What I heard was sound over a very wide lateral angle, almost 180 degrees located between me and the speakers and having pinpoint localization. It was kind of like listening to a cross between headphones and speakers. The sound had remarkable clarity which I attribute to the monster Soundlabs electrostatic speakers. Ralph also installed a horizontal surround system using commercial electronic reverberation which seemed to have little effect despite the number of additional Soundlabs speakers including two additional “ambiopoles” directly behind the listener mirroring those in front. The amount of equipment in this no holds barred system must be over a million dollars or at least close to it.

    Edgar Chouiri uses a similar idea using only two small stand mount speakers in photographs of his anechoic chamber in his lab at Princeton University. His recordings are binaural. His variant uses a low cut filter because he recognizes that when frequency gets higher and consequently wavelength gets shorter, the area where cancellation will occur will diminish to the point where it no longer works and will instead of adding out of phase, they will add in phase creating a high frequency peak. The filter is designed to nullify this peak to get system frequency response flat. I haven’t heard it.

    There are other systems taking different approaches. Ambisonic sound uses four microphones or three microphone pairs in three axes X,Y, and Z and a fourth omnidirectional microphone. that is used as a loudness reference only to set gain for the playback system. These microphone pairs are L minus R, Front minus rear, and up minus down. In principle these signals can be added algebraically to arrive at sound from any arriving angle. In practice, “higher order” systems use additional microphone arrays to achieve more angles.

    There are other multichannel approaches that are similar to each other. They appear to work best in the large space of theaters.

    The one that comes closest to my own is Wave Field Synthesis invented at Delft University in 1988. This is by far mathematically the most conceptually complicated of all of them. It appears to arrive at a similar or same conclusion as I have staring at an entirely different point and through a very different path.
    https://en.wikipedia.org/wiki/Wave_field_synthesis
    If both systems were developed to their full potential, an impossibly daunting task even today, they should be identical. But in practical demonstrations the implementation has some similarities but many important differences.
    Among the limitations of WFS is that it relies exclusively on anechoic recordings, has not been adapted for use with commercial use, in current prototypes need many more channels and its measurement methods to establish the relationship between different sound field components is difficult and may be fatally flawed. The difficulty of solving this last problem almost caused me to throw in the towel on the whole idea. The crux of the problem is separating time of arrival and loudness and spectral content of individual reflected sounds. This is why an impulse test signal can’t work. The solution revolved around a simple mathematical trick. Normally a waveform is either in the time domain or the frequency domain. My method uses both at the same time. The time of arrival of reflections is kept in the time domain while their amplitude and spectral deviation from the first arriving sound is in the frequency domain. This is how I fully account for every single reflection arriving at a point or through an imaginary closed surface, preferably a sphere. The relationship is mapped one frequency at a time.

  9. I love QSound. I’ve used QSound recordings for decades as references. If your speakers or your room setup produce a less than amazing soundstage on QSound recordings, some important things are going very wrong — and I’ve heard some expensive systems get it wrong.

    I also like using as references some old-fashioned analog recordings where phasing and flanging were used to create extraordinary spatial effects. Certainly “Whole Lotta Love” is tops among them, but Jimi Hendrix “Still Raining, Still Dreaming” and Santana “Singing Winds, Crying Beasts” are pretty close. I’d love to know what other favorites people here have.

    1. Just listened to the tracks you mention and they are indeed great examples of Sound. Guess I’m looking for even more suggestions should you (or anyone else) have more to share. Thanks!

  10. I have to admit to having never heard this term until this post. I went to Amazon and put in Qsound and got one result which was a group named Qsound. How do I know if I have one of those recordings or are these just the way they are mixed on the representative albums?

    1. The Bose Guy Speaks :

      I liked Qsound especially effective on Pink Floyd’ P.U.L.S.E. I transferred to 96/24 WAV’s before I sold it 3 years ago (I bought in ’95 along with the cassette – the CD is useless without the bonus tracks only on cassette and LP). With Qsound and me smack dab in my sweet spot. My 901’s really show off the “surround” effect. Now when I listen the “normally” recorded Stereo and sit in my sweetspot, I don’t get the nasty stereo sprayed everywhere effect. Bass is precise, mids are easy on the ears, and highs are sweet. The stereo imaging does have that box disappearing thing going for it (of course that true for ALL loudspeakers when you close your eyes…LOL).

      I still have my ’92 CD of Amused to Death. While I’m more in the Gilmour camp than the Waters (Rattle That Lock – great album), it has it’s good moments for a Waters political as usual charge album (yeah Floyd did it on both Momentary Lapse of Reason & Division Bell – but a lot more subtly).

      I too have been a discrete channel 5.1 fan since DVD-Audio & SACD (glad that SACD won but why are there still so many new DVD-A releases). A properly mixed BluRay/DVD concert in 5.1 is always a treat.

      1. rajugsw. You should try some BluRay audio only albums like Alan Parsons Project “Tales of Mystery and Imagination” “Beatles No.1” King Crimson etc. Much fun.

  11. Bob Carver’s Hologram Generator preceded the Polk speaker. Bob was the visionary conceptually, Polk implemented Bob’s epiphany physically instead of electronically.

  12. I would like to add an additional device to this discussion; The Trinaural Processor
    which is still a production device. It was one of the last products designed by James
    Bongiorno (he passed away in 2013) If his name is not familiar to you, please google it.
    It generates (3) channels from a stereo source, I have used this device for years and
    recommend it. It works very well on the majority titles in my collection, a very few not
    so well. I feel it is a worthwhile contribution in providing something closer to a musical
    event in my home and would not consider parting with it. I own the Audionics, Carver
    Sonic Hologram and Fosgate devices, the Trinaural is the best so far. Every once
    in awhile, I will drag out of of these devices out for a comparison.
    At one time, 3 channel reproduction of music in a domestic environment seemed to be
    “the way”. Paul Klipsch was an advocate of it and I remember seeing an old audio ad for
    Ampex 3 channel open reel recorders (they made professional models) that had Frank
    Sinatra in his home (supposedly) with a 3 channel setup. I think Magnepan still offers a
    3 channel configuration on special order. I played my copy of “Amused to Death” and
    was surprised with the spatial reproduction, fortunately I am still here to write this.
    I am really looking forward to a production device based on some of the recent articles
    I have read in the AES journal. I understand Mr. McGowan has done his own research
    into these “waters”, perhaps such a solution will be a future production device from PS Audio?

  13. All of these systems are FAKE sound and work because of the “break-in” of your brain to fake 2 channel stereo. When you listen to 2 channel pan pot stereo for hundreds to thousands of hours, your brain latches on to faint effect of level differences, the diffraction/lobing artifacts of your speakers and the time and space filtering effects of your room reflections and associates it with a spatial image – because that is what you sub-consciously expect to find.

    This process is not just “tuning” or “acclimating” of your brain’s aural processors, you actually grow neurons and connect them in neural circuits to represent the fake images so you hear them again and again. The effect is stronger from speakers with similar diffraction patterns. If your first real HiFi was Magnepans or quads, you will hear the best imaging from planar speakers. If they were mini-monitors, then the BBC 5″ portable design license will be the epitome of soundstaging.

    If you conditioned your ears speakers with large radius baffles like Thiels, or wide baffles like the old Snells as a teenager, then they will be your fake stereo reference for life. These, at least, will offer a better image from near coincident pair recordings, which have the maximum amount of real stereo information.

    Systems which depend on phase shifting in attempts to re-create a real sound field will inherently distort the real stereo temporal information and signal waveforms in ncp recordings. Cancellation will also “blow up”, requiring excessive levels and exciting other non-linearities in speakers like Doppler inter-modulation, exceeding DuMax, etc. I have heard pretty convincing demonstrations with 360ft2 of SoundLab radiating surface, but the mini-monitor version was very constrained in effective dynamic range and fatiguing.

    My solution for commercial releases which are mixed and panned mono sources with artificial reverb: L+R summed mono. There is zero real stereo information so it will sound better if you don’t engage the fakery.

    A good experiment to show this effect is the Stereophile test tracks where John Atkinson walks around a room talking. Everybody hears hours of acoustic speech a day during their formative years, so your brain computer grows circuits to interpret REAL stereo for speech. This is also why the “one ear test” works better for speech than music.

    If you compare these experiment between speech and music recordings, you will get different results because speech and music are processed in different parts of the brain. Your speech processor was mostly acoustically trained, and your music processor was most likely trained by loudspeakers or headphones – and this neuroplastic process happens between birth and teenage years.

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301
1-800-PSAUDIO

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram