Calling on all Synth Wizards!

[Toddskins]

Let's face it. We're the guys who take sound design to the nth level. I need your collaboration for a feat that is not faint-hearted, I think.

For 25 years, I have imagined creating 3-D sound, in a way that has not been done, at least to my knowledge.

Thank God I own a Kurzweil K-series that has numerous mathematical models that can be incorporated into the design signal, so I am fairly certain that if you guys can help me with the algorithm, I'll be able to implement it in my keyboard signal.

Ok.

Here's the scoop: Imagine a toy slinky, stretched out in front of your face, horizontally, like a pipe, moving from left to right. See uploaded image below.

For starters, I'm wishing to create a swirling sound that spins, and then moves left to right through the sound field. The variables I am pretty sure that will be varied will be 1) amplitude, 2) panning, 3) maybe a little pitch, similar to a doppler effect, and then probably some type of reverb room. From what I have read about 3-D sound, various companies have utilized phase variance (I'm not 100% sure what that would mean).

I've found this great starting point that not only gives some good evaluation, but there are a couple of great 3-D sound examples at the bottom of the page, listed under [External Links]. By all means, be sure to put your HEADPHONES on. The best example is the "Virtual Barbershop"! Listen to it now! Then come back here.

http://en.wikipedia.org/wiki/3D_audio_effect


After you have listened to that, and taken a look at the definition on that site, can you offer some input to me on how I could design the algorithm into my Kurzweil's signal path, so that when people listen with headphones, they would feel/hear the effect of 3-D sound going through and around their head?


I'm wondering, as you look at the illustration of the slinky (imagine 3 dimensional), what do those oscillating circles, along the moving slinky path, look like in 2 dimensions? I have held up a spiral notebook to the light, to try and see what that looks like in 2-D, and it seems to look like triangle waveforms. [See image below]

So, the amplitude (loudness) would be greatest at 0 degrees, and as the sound moves up the cycle to 90 degrees, the sound gets less loud, and is at its minimum loudness when it reaches 180 degree. Then, the sound begins to return toward you, as it continues moving through the cycle toward 270 degrees, and is loudest again at 360 degrees.

But to simulate 3-D sound, the speed cannot be linear. So in the 2-Dimensional triangular waveform, would not the sound seem to take longer to move along various points of the waveform?

Taking the doppler effect into account, would the pitch diminish ever so slightly as the sound moves away from you, and rise as it returns?

I will be thrilled if you guys can help me figure this thing out. Afterwards, I would love to experiment with moving the whole axis, like as if the sound were going directly through the centers of your ears. And then, move the axis vertical, right through your head from top to bottom, to simulate a tornado.

Lastly, there must be a way to also simulate height and depth positioning.

Please, let's put our engineering caps on and help me with this!

Thanks in advance!

Todd

[altitude909]

wouldn't panning it left and right repeatedly and decreasing the volume give that effect?

[Ritzy]

One word: Leslie

[Toddskins]

Originally Posted by altitude909

wouldn't panning it left and right repeatedly and decreasing the volume give that effect?

What you just described is normal stereo.

Go to the link in my post, read up on 3-D sound, listen to the example of the site, and then you will begin to understand the distinction.

[Toddskins]

Originally Posted by Ritzy

One word: Leslie

The whole idea is to get 3-D sound in a recording. When you listen to leslie in a recording, it is still in flat stereo field.

Go read the link on Wikipedia (original post), listen to the example "Virtual Barbershop".

[DAS19]

I think its called binaural recording? but I dont know what exactly you want to do with that and a synth? Maybe you could record it using binaural micing? but I dont know what that would do as for the sound. There are definitely special binaural micing techniques.

[NL5]

Originally Posted by DAS19

I think its called binaural recording? but I dont know what exactly you want to do with that and a synth? Maybe you could record it using binaural micing? but I dont know what that would do as for the sound. There are definitely special binaural micing techniques.

Binaraul recording is one way to do it. But there are psycho-acoustic proccessors that can do this, but the name escapes me right now. They can make the sound seem to come from left of the left speaker, or even from behind you.

[Toddskins]

Originally Posted by NL5

Binaraul recording is one way to do it. But there are psycho-acoustic proccessors that can do this, but the name escapes me right now. They can make the sound seem to come from left of the left speaker, or even from behind you.

Is it your opinion, then, that this could not be produced in Real Time using top-end synth workstation?


I'm thinking that whatever the end results are for the special effect of 3-D, such as recording with 2 microphones, that this could be achieved in the creation of the sound in real time, using the correct algorithm and effect (i.e. reverb with phase modulation).

But I don't really know.

I suspect that if I used the recording technique to achieve 3-D (which I do not know exactly what is involved), that that could somehow be mapped out looking at the result in 2 dimensions, and then essentially reverse engineer that so that it could be done directly from the synth in real time.

Again, I'm not sure what that 2-dimensional wave looks like that I might be able to create the same thing from the synth engine. But it only seems logical that it could be done in just this manner.

[jpmorris]

Originally Posted by NL5

Binaraul recording is one way to do it. But there are psycho-acoustic proccessors that can do this, but the name escapes me right now. They can make the sound seem to come from left of the left speaker, or even from behind you.

Aureal used to make soundcards which could do this. They rendered full 3D into a stereo pair, which was absolutely incredible for games. One of my treasured memories was playing System Shock 2 on this with a pair of headphones. It was all dark, and suddenly this voice right behind me shouted "THE MANY ARE STRONG!!" and killed my character. As you can probably imagine I almost jumped out of my skin.

Aureal held most of the patents on this until they were sued and bought out by Creative Labs, who promptly sat on the technology. As far as I know, no-one has yet equalled what Aureal were doing in realtime ten years ago.

[LemonTree]

you're thinking way too hard about this.

You want to use pan....left to right and reverb that is oscilating (more then less, more then less) for every rotation around the spring (front less reverb, rear more) and that's it

I've been doing it for years

[Toddskins]

Originally Posted by LemonTree

you're thinking way too hard about this.

You want to use pan....left to right and reverb that is oscilating (more then less, more then less) for every rotation around the spring (front less reverb, rear more) and that's it

I've been doing it for years

That sounds partially correct. Can you tell me what type of LFO wave you are using, and its rate? Does it go 'round linearly, exponentially, or logarithmically?

[Somnium7]

You may want to consider experimenting with Doppler effects and quadraphonic effects. Pick up a used Sony DPS-V55 for both of those effect types. Very interesting things happen using the Sony's quad FX blocks and routing it's 4 outputs to 4 mixer channels and playing with the pan and EQ.

Another item of interest would be the Hammond vibrato scanner. A stereo pair of these would make for some interesting 3D-like sounds. Here's some info on a scanner being used with an analog synth - really neat!
http://modularsynthesis.com/modules/...er/scanner.htm

[Erockrazor]

Don't really know what you're talking about.

But you could get 4 speaker surround system. Mix on a surround sound program. But put the rear left and rear right speakers center top and center bottom even with your stereo speakers. Is that quadraphonic? Then make the pans and everything else accordingly.

Whatever you're trying to do, I am too stupid to grasp right now. Regardless good luck, it seems like you're on a mission from the gods.

[nord]

ok i`ve listened to the barbershop:
i didn`t find it that convincing overall ididn`t know where the barber was at all times
i`ve heard really convincing effects that have been better
the bag was realistic though
and the last ten seconds; that is a really good effect with the far away close thing


the problem is using two headphones means the sound can be at any point a fixed distance from your ears ie 12 oclock is 6 oclock
then your ears work out the rest from echos i think

i`m not entirely sure what you want to do to be honest heres some links though the circular wave exists in light

nord





http://en.wikipedia.org/wiki/3D_audio_effect
http://media.putfile.com/Virtual-Barbershop
http://en.wikipedia.org/wiki/Holophonic_recording
http://en.wikipedia.org/wiki/Polarization

[IronFlippy]

Originally Posted by LemonTree

you're thinking way too hard about this.

You want to use pan....left to right and reverb that is oscilating (more then less, more then less) for every rotation around the spring (front less reverb, rear more) and that's it

I've been doing it for years

For a more convincing effect, you could modulate volume, and most importantly, high frequency information, which is the main dictator in distance localization.

[Toddskins]

I dug deeper into my reference manuals, and the KDFX effects module, in my Kurzweil, has a licenced algorithm from SRS Labs 3-D sound technology. Here's the page from the Algorithm Reference Manual:

FXAlg #965: SRS
Licensed Sound Retrieval System or SRSTM effect
Allocation Units: 1

The SRSTM algorithm has been licensed from SRS Labs, Inc. The following is from an SRS Labs press release:

SRS, the Sound Retrieval System, is based on the human hearing system. It produces a fully immersive, three dimensional sound image from any audio source with two or more standard stereo speakers. Whether the signal is
mono, stereo, surround sound or encoded with any other audio enhancement technology, SRS expands the material and creates a realistic, panoramic sound experience with no "sweet spot" or centered listening position. SRS is
single-ended, requiring no encoding or decoding, and uses no artificial signal manipulation such as time delay or phase shift to produce its natural, true-to-life sound image.

The four SRS parameters control the ambience of the image, and may have different optimal settings depending on the amount of stereo content in the inputs. To match the optimal settings specified by SRS Labs, the bass and treble gains should be set to 0 dB. This algorithm will have no effect on mono signals.

Parameters:
PAGE 1

In/Out When set to "In" the effect is active; when set to "Out" the effect is bypassed.

Out Gain The overall gain or amplitude at the output of the effect. Out Gain is not applied to the signal when the effect is bypassed.

Center The amount of "center channel" can be varied with this control.

Space The width of the image is controlled with this parameter.

Bass Gain The amount of ambience added to the Bass frequencies in the signals. A setting of 0 dB gives a best match to the optimizations of SRS Labs.

Treb Gain The amount of ambience added to the Treble frequencies in the signal. A setting of 0 dB gives a best match to the optimizations of SRS Labs.



After I experiment with this, I'll record something and upload it for all of you to hear.

[mjbphotos]

The Virtual Barbershop was interesting - but did you notice the sounds and voices never seem to come from in front of you? Kind of like surround sound 4.1 (no center front speaker). Technique was pretty simple for that - reverb and lower volume for distance, more volume and panned left or right for close up.

[Toddskins]

Originally Posted by mjbphotos

The Virtual Barbershop was interesting - but did you notice the sounds and voices never seem to come from in front of you? Kind of like surround sound 4.1 (no center front speaker). Technique was pretty simple for that - reverb and lower volume for distance, more volume and panned left or right for close up.

I hear what you and others are saying, but I think there's more to it than just this.

The definition from Wikipedia spoke about Side-Chain signal and phase inversion.

I'm sure that must be part of it to create the believable illusion.

[LemonTree]

second thoughts go download the lexicon PSP84 plug.

S

You don't need to take the engine out a car to change a tyre