Smyth Research's Realiser 16 - Up to 16 Channel Surround Over Stereo Headphones

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
The advent of crap formats like MP3 and other compressed monstrosities have expanded the portable sound market. Also Dr Dre managed to market a $50 set of headphones at a $400 price tag and made it cool to use headphones (with pretty colours!). As a result the general price of headphones are now way up for the same cost.....this means profit!

Headphone are now becoming a major market.
All too true.
 
Hi Otto

The advent of crap formats like MP3 and other compressed monstrosities have expanded the portable sound market. Also Dr Dre managed to market a $50 set of headphones at a $400 price tag and made it cool to use headphones (with pretty colours!). As a result the general price of headphones are now way up for the same cost.....this means profit!

Headphone are now becoming a major market.

Regards

Chucky
Chucky with all that mobile gadgets like phones/tabs it is logical but i'm very doubtful about commercial success and phones with
price tag nearing 2 grants could flood the streets. don't know sales data but i do suspect previous serie Realiser 8 sold at best maybe
hundred worldwide. as you pointed out, for such market, the phones need to be fancy colored with price tag do not exit price of the
gadgets, to which they connects.
 
Chucky with all that mobile gadgets like phones/tabs it is logical but i'm very doubtful about commercial success and phones with
price tag nearing 2 grants could flood the streets. don't know sales data but i do suspect previous serie Realiser 8 sold at best maybe
hundred worldwide. as you pointed out, for such market, the phones need to be fancy colored with price tag do not exit price of the
gadgets, to which they connects.
I do not believe that the general headphone market has much, or anything, to do with the Smyth market.
 
Hi Kal

Actually IF YOU SIT IN THE SWEET SPOT and use really good speakers such as electrostatics you get superb object positioning way better than the center channel lump under the TV.

Um, nope. Or actually, maybe yep, if your center channel is a 'lump under the TV'. Mine is a freestanding model identical to my left/right , at the same height and distance as they are. I guarantee you my 'object positioning' is as good as yours in the sweet spot, and far better than yours when there is any movement of my head or body away from the sweet spot.

That's not boasting, it's science. It's how audio/acoustics works. A two channel L/R can never be as accurate as a 3-channel LCR setup (that goes all the way back to Bell Labs research in 1930s, when it was determined that 3 channel was minimum for good depth reproduction, but in the end, 2 became the standard, because, 'wife acceptance factor' existed even then) See slides 32-35 here: http://www.aes-media.org/sections/pnw/ppt/jj/soundfields_vs_human_hearing_edited.ppt

As far as delivery channels, Tomlinson Holman (who knows a thing or two about the topic) seems to think 10.2 improves on 5.1 (which he helped develop). If nothing else, it adds height information that is surely not accurately conveyed in 5.1.
But this is 'true' surround, not surround synthesis, which is what Smyth seems to be doing , i.e., a revved up version of DPL II.

https://en.wikipedia.org/wiki/10.2_surround_sound
 
I do not believe that the general headphone market has much, or anything, to do with the Smyth market.
Kal, that was my initial message, curiosity about motivation of someones to jump into such field, which absolutely do not guarantee any return on investments.
isn't would be more sense to direct their knowledge, experience, funds and effort into something more practical and affordable for "general hardware market"?
but anyway, what do i know. rich folks have their own bad habbits :p
 
Kal, that was my initial message, curiosity about motivation of someones to jump into such field, which absolutely do not guarantee any return on investments.
isn't would be more sense to direct their knowledge, experience, funds and effort into something more practical and affordable for "general hardware market"?
but anyway, what do i know. rich folks have their own bad habbits :p

The Smyth brothers, ex-dts, seem not interested in commercializing their inventions. The original Realiser has been available for many years and I know that there have been inquiries about applying/licensing the technology but to no avail.

FWIW, I was greatly impressed with the Realiser: http://www.stereophile.com/content/music-round-45#F4IpZlWlKViZl2K0.97
 
We have to recognise that any soundfield has to be capable of description as just two time varying instantaneous pressure values - that is all our two ears have to go on. However many arbitrary channels you chose to record, transmit and reproduce over loudspeakers that is ‘all’ (!) you are actually trying to replicate. The race to ever more channels will therefore I’m sure eventually prove to have been fundamentally flawed. What is the most accurate sound localisation technique? – binaural stereo. Why does it work so well? Because it delivers the with the greatest fidelity the sound pressures from a pair of microphones in a dummy head to your eardrums. Yes, I know that obviously doesn’t work over loudspeakers because of the crosstalk between the channels that occurs in the room (despite a few failed attempts at binaural crosstalk cancellation - didn’t JVC market such a thing many years ago?). However I can’t help but feel that there are clues in the way that binaural works that will someday be used to transform the way that ‘surround sound’ is delivered. Could you extract a pseudo binaural headphone signal from conventional stereo or surround recording? I have a gut feeling that there must be enough information there to be able to do so (and wonder if that is your approach Chucky?) although the HRTF of the dummy head itself seems to be pretty crucial in the effectiveness of binaural and this may prove difficult to emulate. Cracking the technique over in the room loudspeakers may take a little longer!
 
I've been puzzling over surround via headphones for a while, the artificial head/binaural technique is good but not good enough. I've always found it to have poor front-back discrimination. One of the things I was thinking about was that the ear, its shape, and the shape of the sound path to the eardrum must come into play. If you move a sound source that is at a distance from you, but directly in line with your nose, then move it in an arc from of constant radius from in front to behind you - so the time difference to either ear is the same. You can distinguish quite accurately whether the sound is coming from in front, above, or behind, which leads me to think that simple time differences are not good enough and you have to model the ear to get a good surround representation.
 
I've been puzzling over surround via headphones for a while, the artificial head/binaural technique is good but not good enough. I've always found it to have poor front-back discrimination. One of the things I was thinking about was that the ear, its shape, and the shape of the sound path to the eardrum must come into play. If you move a sound source that is at a distance from you, but directly in line with your nose, then move it in an arc from of constant radius from in front to behind you - so the time difference to either ear is the same. You can distinguish quite accurately whether the sound is coming from in front, above, or behind, which leads me to think that simple time differences are not good enough and you have to model the ear to get a good surround representation.

Yes, that's very true - I've also generally found the front / back discrimination of binaural to be poor whereas left to right is frighteningly realistic. And you're right, the trans-head time difference certainly isn't the key parameter as shown by those binaural recordings that use a simple Perspex baffle between the microphones - much less effective than those that use a fully modelled head (especially with anatomically correct ears). Clearly the frequency /amplitude characteristics of the inner and outer ear are of great significance, unfortunately they are unique to every individual! That's obviously why reportedly the most convincing binaural recordings are those made specifically for an individual with microphones placed in their ear canals - sadly not a mass marketable solution!
 
Yes, that's very true - I've also generally found the front / back discrimination of binaural to be poor whereas left to right is frighteningly realistic. And you're right, the trans-head time difference certainly isn't the key parameter as shown by those binaural recordings that use a simple Perspex baffle between the microphones - much less effective than those that use a fully modelled head (especially with anatomically correct ears). Clearly the frequency /amplitude characteristics of the inner and outer ear are of great significance, unfortunately they are unique to every individual! That's obviously why reportedly the most convincing binaural recordings are those made specifically for an individual with microphones placed in their ear canals - sadly not a mass marketable solution!
Yes and, of course, that is exactly what Smyth does to create its transfer functions, albeit from a playback system, not from the real performance. Mebbe we should get the Smyth guys on this.
 
Dear All

May I digress temporarily from headphones.

Really good comments. In regard to the center speaker being the only way to create depth and enhance positioning (Bell labs in 1930's), yes but not really. Turns out that a number of years ago we at Involve did a study to examine why a good testing speaker can sound bad and a bad testing speaker can sound good. My initial theory was that it was going to be something related to the impulse decay response but I was wrong. We tried to correlate all the standard test parameters to perceived sound quality....you know, frequency response, thd, imd, tone burst, polar response. And in reality none correlated (except polar in a sense).

"Overture" and I were both stunned and confused by the result, we both sat down (on the floor actually) and then it struck us that the prime factors may have been more obvious. When we looked physically at a "good" sounding speaker vs a really good testing bad sounding speaker it became more obvious.

I will let you guys on a little secret: The differences were regarding what I call concentricity and the front back polar response. Turns out that 99% of the speakers we buy are WRONG! We constructed test jigs with test audiences on blind trials and found when all drivers were moved to be concentrically placed as opposed to the usual up/ down tweeter/ mid/ woofer placement the audience response was dramatic with ALL positively preferring the concentric placement- it was not subtle. We tested time alignment- and that is bullshit.

Next we constructed jigs to test radiation patterns comparing monopole (front firing)/ bipole/ dipole and again the audience preference was dramatic with almost a universal preference to DIPOPE, the least preferred was monopole. Again it was not the usual audiofool subtle stuff, it was in your face.

The real difference using concentric dipoles was a greatly enhanced sense of depth and vastly improved imaging and separation of sonic objects. All tests were conducted with conventional cone speakers. This was an interesting result as 99% of all speakers sold are non concentric monopoles!!!!! This was even more interesting as we are and were manufacturers of electrostatic loudspeakers (see Nakamichi dragon electrostatic and our kit partner www.eraudio.com.au and http://www.hifizine.com/2013/03/the...ew-with-charlie-van-dongen-and-rob-mackinlay/).

Guess what- electrostatics are concentric dipoles- the answer was staring us in the face all the time!

This is what I was getting at in my previous post that the best solid stereo images with depth I have experienced have been with STEREO electrostatic loudspeakers, that was the reason I got interested in them in the mid 1990's. Imaging and depth with 3 speaker solutions have always been vastly inferior.

So part of my point is the current "pissing contest" on more and more channels is based on faulty logic and test results of say the 1930's conducted on speakers that were frankly wrong for serious imaging. If you ask me later (very nicely) I could explain why concentric bipoles produce thus superior imaging.

This is my point with headphone - and certainly no disrespect to Smyth who I believe produce a truly excellent product. Claiming 16 channels is essentially meaningless. Consider the results you get from (inner ear) personal recordings on dummy heads and played back inner ear.....its TWO CHANNEL.


Getting back to headphones. As I stated previously the problem with headphone surround is removing the in head sound and creating an effective frontal image, back sides is actually much easier. The real trick is to do it with all heads torso shapes and ear shapes (Smyth do the multi user thing well apparently). Localisation of frontal image is a complicated thing as there is no left/ right time delay discrepancy to help in working it out. It is more concerned with that frontal flap of skin blocking your ear hole and the relative micro time delays from direct/ reflected paths around this flap plus the frequency contouring of the frontal image. As a note, tests have shown that in an anechoic chamber users have difficulty in determining frontal distances and essentially the perceived distance is shortened considerably. This implies frontal imaging is heavily reliant on reflections and the relative time arrivals of direct to reflected sound and their sequence.

Any way "Overture" or as we know him as "Dave the Bitch" is currently working on such a headphone surround device coupled to Involve surround decode, so you get 2 channels in surround headphones out. Cannot guarantee if it will work (and on all heads) but we are getting encouraging results. The product will be pitched at a lower price than Smyth buy not at the 5- 16 channel market.

Wish us luck, we might succeeded.

Regards

Chucky
 
Last edited:
Some very key observations there Chucky.

On the three channel stereo issue I’m also a non-believer. I‘ve seen the arguments going back decades about how it should improve stability and localisation etc. – but for me it never has delivered on that promise and I actually dislike the effect of ‘pinching in’ of the two channel soundstage that it almost invariably causes (not ameliorated for me by having all three speakers identical) . The third channel may have some value as dialogue channel in home cinema, but then it’s only really acting as an ugly sticking plaster for all kinds of other problems. I don’t think it works at all well for music reproduction (particularly orchestral).

Done well, two channel stereo can (and thankfully does) produce wide and deep soundstages that extend beyond the boundaries of the loudspeakers and that are not limited to a very small sweet spot. But ‘doing it well’, as you have identified is absolutely critically dependent on loudspeaker performance (no other component, including the number of channels, has remotely as much influence on fidelity). In my experience, speaker complexity and dispersion patterns wildly effect imaging. Some (if not most) large, complex, expensive multi drive unit speakers are truly awful at imaging whereas many small, simpler and cheaper speakers can be surprisingly good.

After buying more loudspeakers over the years than I’d be prepared to admit to, I have found that the ones that work best always have the same two common characteristics. They have a small number of drive units (often only one full range or coaxial / concentric) and are very narrow. I’ve often thought that this must equate to getting as close as you can to having a point source. Certainly the dispersion characteristics of multi-drive unit speakers can be a nightmare (just listening to the tonal colour and ‘phasiness’ of pink noise changing as you sit down , stand up or move about can be very shocking when you’ve just spent a fortune on some so called ‘time aligned’ beast!). It doesn’t surprise me therefore that your concentric dipole panel approach works so well (as Peter Walker also found out of course!) In audio less is so often more.

On the headphone issue I do indeed wish you luck! Even the best binaural recordings I’ve heard still don’t image well front to back and placing them outside of the head is clearly very difficult indeed. Doing so from non-binaural material would be one hell of an achievement!:brew
 
Dear All

May I digress temporarily from headphones.
Thanks. This is very interesting and not entirely surprising.

This is my point with headphone - and certainly no disrespect to Smyth who I believe produce a truly excellent product. Claiming 16 channels is essentially meaningless. Consider the results you get from (inner ear) personal recordings on dummy heads and played back inner ear.....its TWO CHANNEL.
Let me suggest that your terminology is not proper. No one (except neurophysiologists) records from the inner ear and no one (except users of cochlear implants) plays back anything in the inner ear. You are really talking about recording and playing back with transducers in the external auditory meatus, a portion of the outer ear.

Resume.
 
On the three channel stereo issue I’m also a non-believer. I‘ve seen the arguments going back decades about how it should improve stability and localisation etc. – but for me it never has delivered on that promise and I actually dislike the effect of ‘pinching in’ of the two channel soundstage that it almost invariably causes (not ameliorated for me by having all three speakers identical) . The third channel may have some value as dialogue channel in home cinema, but then it’s only really acting as an ugly sticking plaster for all kinds of other problems. I don’t think it works at all well for music reproduction (particularly orchestral).

Done well, two channel stereo can (and thankfully does) produce wide and deep soundstages that extend beyond the boundaries of the loudspeakers and that are not limited to a very small sweet spot. But ‘doing it well’, as you have identified is absolutely critically dependent on loudspeaker performance (no other component, including the number of channels, has remotely as much influence on fidelity).
I do not know exactly how you did your tests and what program materials you used but my results are different. On a number of occasions, I have asked groups of listeners to compare what they hear from a 2channel recording With the identical performance played from a 3channel recording, using identical speakers across the front. In every case, the listeners predicted that center imaging would be improved and clarified. In every case, however, they reported that the major difference was a much wider soundstage with better localization/resolution at the lateral extremes. Source material was from RCA Living Stereo SACDs and my hypothesis is that the output from the L/R speakers was improved because the central source material was not being conflated with the L/R signals.
 
Thanks Kal and Soundfield!

My observations on the actual benefit of the center channel is to "band aid" the essential problem of surround sound and one of the main reasons Quad died in the 70's (I was there!!) and that is to stabilise the central vocalist image when you sit off center. Interestingly whist large electrostatics (in my experience) have the best imaging in the sweet spot they actually have the worst imaging off center. The major down side of the center channel in a typical installation where it is placed under the TV is that there is a discrepancy of positioning of the visual voice origin and the apparent image location (the voice is under the TV). This has always disturbed me.

In regard to Kal's observation that in his experience 3 channels typically image better than 2, this is entirely possible with cone speakers that are not in some way compliant to being a concentric dipole. As observed by Soundfield often smaller 2 way speaker image fairly well, this is primarily due to the smaller "woofer" being in close proximity to the woofer. Also having a narrow frontal width can widen the polar pattern.

The concentricity thing can be thought of as traffic lights of red, green, blue. When viewed close they are clearly coloured. At large distance if all are on they may appear as white. It is related in physics by the "Raleigh Criterion" which in practice defines at what distance you can resolve an object as separate individual objects (pity 4K TV's typically void this at normal sitting distances whilst consuming 3 times the power- oh well can't argue sales). When the different drivers are concentric the brain has an easier job of putting the whole sound together and actually reduces fatigue. I am so spoilt by years of electrostatics that normal cone speaker give me a headache in 5 minutes. This is not to say you cannot use cones, our test jigs were all cone based and ended up imaging superbly. I note it has become common practice to put the tweeter between two mid ranges, this is better and partially mimics some aspects of a concentric driver, right idea and I suspect most designers do not know why!

When the sound source is a dipole (bipole is second best) the sound is equal front to back, this creates more frontal reflections around the room and assists the ear/ brain decode a frontal point in space. Think about it for a second - how many sounds in nature only fire frontally?

I know this is slightly off topic but it is related.

As only one member of QQ (Hi Dwight) has visited our factory down under and experienced this system I understand why QQ members are supportive of the center channel band aid. Hey Rustyandi pay us a visit!!!!!!!! We have used a patented technique using the most directional speakers (electrostatic) with a bunch of tricks to eliminate the need for a center channel and allow proper central imaging in all seats of the house with either 2 or 4 surround (yay 4!) speakers. It works with all other types of speakers but electrostatics are best. Yeah yeah -I know its not possible....blah...blah ...blah. Check out out Y4 system.

I feel like a voice crying out in the wilderness on my war against multi multi channels! In short in our home theatre set up in Melbourne we set up the best 7 channel Dolby based system the shop could offer (around $40K) and we switch it A/B to our 4 channel cone based Involve system with the sweet spot correction running from a STEREO source. So far in around 50 test monkeys not one wants it switched to 7.1, have not got an Atmos setup yet to compare.

Re headphones my reference to recording from the inner ear canal is not a practical thing but the best dummy head trials doing just that have reported quite good frontal imaging along with up, left, right and rear. Problem most dummy head recordings are based on a generic plastic head and typically microphones located outside the ear canal and played back with over the ear headphones- clearly a wrong setup. The fact that inner ear dummy head technique experiments produced superior imaging is strong proof that multi multi channels is a false concept.

The headphone problem is a very difficult one and at the moment the Smyth by all reports is by far the most realistic system. Lets see if we can improve matters, cannot guarantee it though!

Regards

Chucky
 
In regard to Kal's observation that in his experience 3 channels typically image better than 2, this is entirely possible with cone speakers that are not in some way compliant to being a concentric dipole. As observed by Soundfield often smaller 2 way speaker image fairly well, this is primarily due to the smaller "woofer" being in close proximity to the woofer. Also having a narrow frontal width can widen the polar pattern.
It is not just that 3 channels image better than 2 but that their primary improvement was not the stabilization of the central vocalist but the expansion and clarification of the lateral soundstage which, again, imho, had to do with the removal of the center signals mixed into the L/R. Would it have been different with different speakers? Perhaps but I suspect that the nature of the observation has to do with the signal mix and not the transducers.

Re headphones my reference to recording from the inner ear canal is not a practical thing but the best dummy head trials doing just that have reported quite good frontal imaging along with up, left, right and rear. Problem most dummy head recordings are based on a generic plastic head and typically microphones located outside the ear canal and played back with over the ear headphones- clearly a wrong setup.
Again, you are not using correct terms. No one is putting anything in the inner ear without surgery. You are probably referring to outer ear parts including the pinna and the external auditory canal.
 
my hypothesis is that the output from the L/R speakers was improved because the central source material was not being conflated with the L/R signals.
In the spirit of "Let me suggest that your terminology is not proper" could you please explain 'conflation' of signals for those of us simple engineers who have not come across the term before?


PS - Kudos to Chucky for sneeking in a reference to Raleigh distance, it doesn't come up much on this forum for some reason!
 
Dear All

May I digress temporarily from headphones.
y


The inherent deficits of 2ch loudspeaker setups vs 3 channel (and more) is not based only on the Bell Labs work from the 1930s. Again, see, links I posted, for starters.

As for the experiments you describe, were results ever published?
 
Back
Top