Thoughts about single box stereo?

There is a very good thread by a DIYA member that was quite prolific, I can't recall the username.
I'm guessing the thread you are talking about IS the one started by member Elias about Nagaoka Tetsuo and other matrix solutions.

The recent posts on this thread by pelanj talks about the Dock E30 speaker by Orbitsound, and I think we assume that it uses a similar matrix solution.
I've fiddled around with prototypes of this kind of thing, and have found that the results are highly dependent on the listening environment, speaker placement, and listening position.

Not quite sure where adason got the idea that I have to revolutionize the world, and do it entirely on my own (if I did, why would I want to discuss it on a forum?).
 
The more I listen to that box, the more I like it. Electronic music can sound really nice with all the effects - and so does a jazz trio. Also satisfactory bass at comfortable listening levels. For the ebay price, I cannot complain.
Thanks for reporting on this!
I'm curious, have you listened to any left/right test audio? You know, where sound is alternating slowly between hard panned left and right. I'd be curious as to how much the sound would actually appear to come from left and right sides.
 
Try Michael Jacksons Thriller intro, walking foley should go across " the stereo ".

I tried the Elias SSS few years back and thought it worked nicely. Clearest memory is from listening Stan Getz Samba Triste and the saxomophone scared me, it came clearly from the side (wall) and stuck there no matter where in the room I was listening. But, it is from memory, cannot comment more accurately. I suspect any of these single speaker systems would work roughly the same, given the positioning etc. some kind of stereo field, some better some less but stereo nevertheless.
 
A measurement technique for stereo images would be intriguing. I know that I have taken impulse measurements of good, bad and different systems with mics in my ears. What could be seen in impulse measurements like that? No idea, but perhaps it's a start.
This is a tricky one...
Imaging (in the sense of being able to pinpoint where in the stereo field a sound is coming from) depends on so many things - phase shift, micro delay, relative frequency response, listening space reverb, and to a small extent relative volume. It's probably nearly impossible to measure, since it's a psychoacoustic phenomenon.

I think I would try to be as analytical as I could, while still using my ears.

I might take a wide-band mono source (music or speech, not white noise), stick it in a recording engineering software, and pan it around the stereo field, listening with my eyes closed. But not only that, I'd also put the signal through a bandpass filter, so I can determine how well I could locate the sound at different frequency ranges. Famously, of course, the sub bass is very hard to localize in any scenario.
 
This is a tricky one.
Yes it is! It's super complex and would be a huge research project. Not something to undertake lightly - tho super interesting.
Perhaps you never heard good dummy head recording
I've heard many, many. I like them, but they never fool me into thinking it isn't headphones. I'll give that YouTube a listen.
FWIW, I have, thru much had work, come up with a set of impulse responses and channel mapping that works for me on headphone. But that is for another thread.
 
An interesting third alternative to single box mono or stereo is what Apple did with their Homepod (the full size one, not the Mini).

It has a single up-firing woofer, and SEVEN tweeters that are firing in a circle, and each is controlled separately.
When calibrating (it can sense if it was picked up and moved!), it measures the response from each of the tweeters (into six microphones), and can tell weither each is firing into a nearby wall or a reverberant space. Using some heavy DSP algorithm it analyses the source stereo signal for dry sounds and ambient sounds (probably using a fancier version of Sum/Difference), and directs them either into tweeters pointing at either dry or reverberant parts of your room.

It's pretty clever. Instead of left vs right, it's dry vs wet. The dry elements of the song stays clear, since it's mostly direct signal, and the wet elements gets a sense of spaciousness, because they actually are spread around the room.

At least I THINK that's how it works.

Too bad the Homepod is discontinued. The Homepod Mini is just a simple omnidirectional speaker (down-firing, with a trumpet cone and all), but it does produce a silly amount of (non-accurate) bass for a little ball.
 
  • Like
Reactions: Pano
I'm guessing the thread you are talking about IS the one started by member Elias about Nagaoka Tetsuo and other matrix solutions.

The recent posts on this thread by pelanj talks about the Dock E30 speaker by Orbitsound, and I think we assume that it uses a similar matrix solution.
I've fiddled around with prototypes of this kind of

thing, and have found that the results are highly dependent on the listening environment, speaker placement, and listening position.

Not quite sure where adason got the idea that I have to revolutionize the world, and do it entirely on my own (if I did, why would I want to discuss it on a forum?).
It was Elias original thread, before the other threads started.
 
I suspect any of these single speaker systems would work roughly the same, given the positioning etc. some kind of stereo field, some better some less but stereo nevertheless.
I would argue that the curved refraction JBL Paragon model stands out as being a bit different from the rest. In that it both reflects the stereo field forward off the curved baffle and compensates for off-centre positioning by the opposing driver becoming direct. But thats from some blurb from the '50's.

I'm sure somewhere on Elias' thread he mentioned that traditional panning will not work, but subtle location is possible. I don't think any of these SSS solutions will ever get proper stereo location right and we shouldn't even try to make that the goal. Whats important is stereo pleasure in shitty rooms.

Struggling to find the time to get my 2nd prototype done, but this is where its at. With one of the Aurasound NS2's 2" drivers for scale. I've been theorising that the stereo effect is probably only likely to work from 1khz up and that I might want a single mid driver as well as a sub, but I'm going active, so will probably eq any problems away.
It will either revolutionise the world of audio forever, or my daughter will like it in her room.
 

Attachments

  • IMG_20220331_175336.jpg
    IMG_20220331_175336.jpg
    242.9 KB · Views: 75
  • Like
Reactions: tmuikku and Pano
Struggling to find the time to get my 2nd prototype done, but this is where its at
I'm very excited that you are trying this out!
It might be worth experimenting with the angle of the drivers, as well as with some louvers (like the Metregon C45) to control at what angle you are starting to get a peekaboo listen of direct radiation from the driver.

As for the Elias SSS (or Sum/Difference) approach, well, yeah, solid left/right projection is pretty much impossible.

Combining the signals electronically (rather than acoustically) works pretty well - it's the theory behind the fairly common Mid/Side microphone technique. It takes a little brain wrangling to wrap ones head around how a front facing cardioid mic and a side facing figure 8 mic can produce proper left/right results.

But in the unpredictable world of imperfect reflective surfaces, all you can hope for is a hint of stereo separation (but you DO get plenty of randomized spaciousness). Prototyping in my workshop I can certainly hear a test signal somewhat bouncing left and right, but it's all pretty diffuse.

You've got to wonder, in an anechoic chamber, with a pair of perfectly placed reflecting surfaces, how would it sound?
 
Elias' site has quite a lot of all kinds of info on them, how the concept plays out http://elias.altervista.org/html/SingleSpeakerStereo.html

I think the implementation could be refined some, for example directivity of the sides and front, in relation to how we hear stereo and how we'd like the phantom center, delay between the center and sides for example. Some things could probably be improved, but still it would be on the rooms mercy. If you think about it the direct sound, the center, comes first and the stereo is made by early reflections coming later. Depending on your room size and where the stuff and the listener is positioned relation of these change but direct sound would be always first and the stereo comes in later. Perhaps something to experiment on in regards of the stereo effect, DSP would allow playing with it. Directivity of the side speakers would affect how much the center would need compensation (for highs) and so on. I'm not sure what would be right things to do to achieve "better stereo", I don't know too well how we perceive stereo. I know there are the ILD and ITD stuff, phase and frequency response, delays, what not. One thing is for sure the direct left and right sound come from far greater angle (> 45 deg easily) than with traditional stereo setup (30 deg) if thinking about a sweet spot. One of the greatest advantages of SSS is that it sounds about the same almost anywhere, room sized sweetspot!

I think there could be some improvement for the system, but still it would need to be positioned exactly and room has to be particular size. Not sure if worth it, although it would be lots of fun trying to think it through 🙂 SSS is enjoyable as it is.
 
Last edited: