Broad Stereo Fields

[virtua]

does anyone know how they can make stereo sounds sound like they come from outside the speakers? you know what i mean?I used to have a sterso with a button on it that you pressed and all of a sudden a sound would appear to come from somewhere to the right of the right hand speaker etc. It was like you were sitting closer to the band lol

No special reason, just always wondered how it worked.

[Chokai]

Clever panning of instruments and frequency panning really. Bass always dead center of course as its omni-directional, drums overheads are panned to either side. Stereos with these special buttons merely make a quick analysis of the sound by frequency and pan them accordingly.

I know what you mean by "outside" the speaker, but this is really just audio trickery.

[john]

hey

I hate to disagree, but I will

Difficult to get across without diagrams, so I'll start with normal stereo to try and make it clearer. With stereo, sound comes out of two sources (speakers), and is received in two locations (ears) a slight distance apart. There a few things going on beneath the surface here.

1. To place something within the stereo field, a source sound is panned within the stereo field resulting in different signal levels for that same source sound being produced by each speaker. Using a very simple model, say the source volume is 10 in mono, to produce the same overall volume in stereo each speaker will output at volume 5 when something is center panned. If you fully pan to the left, the left speaker will output 10, while the right output is zero. The opposite being the case if you pan fully to the right. But how does one volume amount coming from two speakers a set distance apart place that sound in a stereo perspective?

2. Our ears. They are not one listening point, but two a slight distance apart. It is this distance that helps our brain place sounds not only within a stereo field, but help create a 3D audio picture of the world. Lets keep it simple. Stereo. Ok, face the speakers and draw an imaginary line from the left speaker to your left ear, and an imaginary line from the left speaker to your right ear. There is a small difference in the length of those lines, resulting in a small delay in a sound reaching your ear due to the extra distance the sound wave has to travel to reach your right ear. You can do the same with the right speaker, drawing lines to each ear and noticing that the lines are of different lengths.

3. Our brains process this information along with the varying volume of the same sound. It employs bothe the delay and volume change to estimate where something actually (virtually) is. Our ear/brain combination is phase sensitive. We can detect differences in phase. We use phase and relative signal amplitudes to work out where something is in space. Something like triangulation, except using two points measuring time, volume and phase to estimate position accurately.

So how do sounds appear outside the stereo field?

Simple. We fool the brain. Remember those lines going from each speaker to your ear? Well the principle is that if we can make the lengths of those lines different, we can chage where a sound is placed. We virtually do this by varying signal amplitude and importantly changing how long it takes for a sound to reach your ear. i.e. delay.

For example: One sound, placed in the stereo field. Thinking of 1 ear, it takes time "X" for sound to reach the ear from the left speaker, and time "Y" for it to reach the ear from the right speaker. If we now delay the signal from the right speaker, say making it now 1.5Y (a crude estimate to make the point) then our brain uses processes this delay and the same thing with the sound arriving at the right ear and places the sound somewhere completely different in space. By chosing the correct amplitude and delay we can effectively fool the brain into believing that a sound originates from outwith the two speakers.

This effect can be applied at mixing stage, a process much simplified with the advent of DSP (Digital Signal Processing) units. In your old stereo when you press the button on your old stereo a DSP chip identifies where a sound is placed in the stereo field, separating the sounds, applying delay and amplitude changes to each, and outputing that to the speakers.

Et voila.

I wont start into 3D sound from a stereo source. It is basically the same principle, with the variables being calculated by some fancy DSP in real time. Damn I started .

Perhaps if bitstream or bongstuff see this post they might be tempted into an explanation.

Cheers

John

[Bitstream]

use hearing related transfer functions (hrtf) to select the azimuth and elevation of the sound. there are some sound cards that do it, and theres some free software / vst plugins too.

for more advanced spatial audio, render as many real reflections as possible. This requires huge computational power (1 hour per mono sound 2 minutes long will render 32 incedent reflections).

I currently have software to do this but the website is imminent.

There it also a myth about 3d sound; the experts quibble about the ability to determine sounds in front and behind is difficult in reality, never mind virtua. This argument is in fact a lot of crap;

http://gprime.net/flash.php/soundimmersion

Edited October 6, 2007 by Bitstream

[Lazz]

to estimate where something actually (virtually) is. Our ear/brain combination is phase sensitive. We can detect differences in phase. We use phase and relative signal amplitudes to work out where something is in space. Something like triangulation, except using two points measuring time, volume and phase to estimate position accurately.

So how do sounds appear outside the stereo field?

Simple. We fool the brain.

Years back, when still in the UK, a friend developed a special unit of electrickery called 'Frankenstein' which would work somehow exactly like that - creating sonic ambiguities which the brain/ear was forced to resolve.

Very clever - but tricky to use effectively because it was so unexpected and disturbing.

Facing the stereo speakers, it could make you believe that sounds were coming from behind you.

Wierd.

[typo]

good topic

[Thrithwaite]

What was the question again?

[redriviera]

Roland RSS (Roland Sound Space) effects processors can place things beyond the stereo field by clever use of delay and phasing. They can also simulate a sound revolving around your head in 3D.

[john]

Yeah, they probably use the head related transfer functions bitstream mentioned to work out the delay, phase and amplitude settings

[Bitstream]

Yeah, they probably use the head related transfer functions bitstream mentioned to work out the delay, phase and amplitude settings

The phase / delay of the incedent sound hitting left then right ear ( or the other way around) is called the inter-aural delay time, or IDT for short. This is what the RSS does. Using the HRTF model incorporates a highly acurate IDT so you don't need to bother calculating it. Too bad Roland did'nt do their homework.

[Prometheus]

I've been deliberating very seriously about whether to make this post, because usually I have little time for luddite philosophy, and hate carping at people like a stern old grandfather figure who's hopelessly out of touch, but anyway...

At the risk of sounding like a party pooper, personally I think spatial sound is an absurd gimmick that is more often than not pretty impractical to play back properly, especially if you have a wife who is a stickler for not having thirty feet of cable running around the living room...

On a more technical note, one thing that I would watch with all this delay and phasing going on is how your broad stereo mix sounds when it's played on little portable ghetto blasters which is essentially almost mono... A lot of people who might listen to music still do not have well set up stereo fields in their homes... Unless, of course, you intend this music to be aimed at a very specific audience...

Edited October 9, 2007 by Prometheus

[Bitstream]

I've been deliberating very seriously about whether to make this post, because usually I have little time for luddite philosophy, and hate carping at people like a stern old grandfather figure who's hopelessly out of touch, but anyway...

At the risk of sounding like a party pooper, personally I think spatial sound is an absurd gimmick that is more often than not pretty impractical to play back properly, especially if you have a wife who is a stickler for not having thirty feet of cable running around the living room...

On a more technical note, one thing that I would watch with all this delay and phasing going on is how your broad stereo mix sounds when it's played on little portable ghetto blasters which is essentially almost mono... A lot of people who might listen to music still do not have well set up stereo fields in their homes... Unless, of course, you intend this music to be aimed at a very specific audience...

Spatial sound is best played on headphones. And if you really splash out, you could get an excellent set of cans for no more that £40

[Prometheus]

Spatial sound is best played on headphones. And if you really splash out, you could get an excellent set of cans for no more that £40

I'm not really a fan of listening over cans, but I can see how spatial sound would be useful for that... I would think that would be better handled at the playback end of things though rather than committing to spatial sound at the mixing stage...

[Bitstream]

I am currently looking into generating a new set of HRTFs; 96khz 24.bit. That should produce the best spatialised sounds out there. Unfortunately I have to interpolate the old set; 44.1khz 16 bit for now. But researchers are generating better quality all the time; it's down to the mics and speakers used, and these guys are using the best.

A lot of folk use headphones with their ipods now, and ambient music has yet to climb up the ranks towards the big boys like, rock. The market is smaller, but the rewards are not diluted over thousands of other grabbing hands. Spatial sound and Ambient seems like a good solution.

otherwise, I'd be synthing it in a rock band like Floyd; just the same dream in the end.

Floyd released a track using holophonics; the proper DSP sound rendering algorithm simulates the holophonic hardware. Not just simple 3D sound.

http://www.apple2.org.za/gswv/1WSW/Audio.4D.html