Super Surround Sound

[brantlew]

Check it out.

http://www.newscientist.com/blogs/nstv/ ... sound.html

[cybereality]

Interesting concept, but its not exactly rocket science, or anything fundamentally new.

[purifier]

I would settle for a "simple" 64.4 channels Dolby Atmos, heck these days I'm envious about my friends in the USA,
they'll enjoy it with The Hobbit, no word from Dolby for Italian theaters yet

[STRZ]

The best binaural headphone experience emulating real sound sources i had with this device: http://smyth-research.com/

But it was only stereo though. The device is usually used by audio professionals to emulate their studio monitor setup on the headphones. There exist some software to do similar stuff, but it's not on par. You have to measure your ears and feed the data back into the software on that device to get it right.

If you could scale this software/device to a setup like shown in the video you'd have real Super Surround Sound on your stereo headphones!

3D sound itself is possible just with a pair of decent speakers and stereo sources from a fixed point in the room. You can hear sound from behind with your speakers beeing in front of you. But you need a good room and symmetric room/speaker setup to get it right.

Check out some qsound demos http://www.qsound.com/demos/3d-audio.htm

[EdZ]

The best binaural headphone experience emulating real sound sources i had with this device: http://smyth-research.com/

That appears to be a rebranding of a common technique known as a Head Related Transfer Function. HRTFs are built into most modern sound cards and motherboard audio (post-AC97).

[STRZ]

The best binaural headphone experience emulating real sound sources i had with this device: http://smyth-research.com/

That appears to be a rebranding of a common technique known as a Head Related Transfer Function. HRTFs are built into most modern sound cards and motherboard audio (post-AC97).

Yes and no, motherboard audio and soundcards doesn't relate to your own ears and their characteristics. But that's crucial if you want good (realistic) results. The Smyth device shows how HRTF is used in the right way.

[EdZ]

The human pinna doesn't differ that much. A precise mapping might give a slight improve, but a generic transfer function suffices for gross positioning and environment simulation.

[RoadKillGrill]

Adding some head tracking to the headphones seems like a much more practical method of getting nearly the same results. There is deadzones on the sides of the head that make it difficult to get accurate directional perception without moving the head a little.

[PalmerTech]

The best binaural headphone experience emulating real sound sources i had with this device: http://smyth-research.com/

That appears to be a rebranding of a common technique known as a Head Related Transfer Function. HRTFs are built into most modern sound cards and motherboard audio (post-AC97).

Yes and no, motherboard audio and soundcards doesn't relate to your own ears and their characteristics. But that's crucial if you want good (realistic) results. The Smyth device shows how HRTF is used in the right way.

I think that the Smyth device benefits more from the Stax electrostatic headphones they list as coming with the system than any particular optimization they do. Electrostats like that have a massive soundstage.

[STRZ]

I think that the Smyth device benefits more from the Stax electrostatic headphones they list as coming with the system than any particular optimization they do. Electrostats like that have a massive soundstage.

So you think it's snake oil hi-fi voodo don't you?

Yeah you need a good soundstage, but you can get that for less money with a AKG 701. Of course it's not a highend Stax and lacking that detail definition, but to benefit from a Stax you need top notch audio recordings, and almost all games have bad audio quality. From a normal consumer perspective you would make it worse with a Stax because you clearly hear where they messed it up.

The trick with binaural techniques is that you perceive the phantom sound sources like when you're listening to speakers, outside your head, not inside or very close like when you don't use those techniques. Without using both of your ears simultaneously, meaning that you have to feed sound from the left in your right ear and vice versa, it doesn't feel natural.

And because you're used to your ears and their characteristics in the real world, it's best to measure simulate those characteristics to include them into the HRTF algorythm.

First you need to be aware if you have variances between the hearing capabilities of your ears an feed that data back into the software wich performs the sweep measurements. Then you put a little microphone like knob into your ear to receive sweeps from your speaker system, this makes sure that the anatomy of your (outer) ear is reflected in the software.

So if you would visit a professional surround studio and calibrate your device over there, you could just take this setup including the room with you using it at home.

That's what's missing in those other devices and software i'm aware of. It's based on common technology, but the integration of those other aspects set's the smyth device a bit apart. I also don't think that they use any super algorythms, just doin stuff right.

I hope that somebody comes up with a open source software wich includes all those steps. Open source to make it easy to extract data from it to use it in virtual environments like level design, so that you put in a bunch of numbers and have the game environment calibrated to your HRTF.

I think realistic audio on a professional level is a topic wich isn't on the radar yet for most game developers, due to hardware limitations. But it could happen once VR is there. EAX and similar stuff is a big advantage over dolby surround and DTS, but still not realistic enough.

[PalmerTech]

Not snake oil, I am one of those crazy audiophile people myself. I used AudioTechnica AD700s for a long time because I loved the soundstage, then spent far too much on a pair of Sennheiser HD800s. I am sure people like you and I are convinced by proper audio (Which I hope VR will push), but the average person has no reference point for what "good" sound is, since they use something like iPod earbuds. That setup would wow them, but you would wow them almost as much with nothing but a good DAC paired to Stax.

There are a lot of people working on good audio in software, the success of Oculus has gotten me talking to some of them for fun. I am experimenting on my own in terms of VR sound hardware, which needs a wide soundstage and excellent isolation properties. The future is bright!

[TheHolyChicken]

I just want audio that sounds as good as this: http://www.youtube.com/watch?v=IUDTlvagjJA (be sure to follow the instructions it recommends)
Is that too much to ask? I never realised how good audio COULD be prior to this, and now I wonder why on earth audio hasn't been pushed further. Games would be so much more immersive if the audio "put you in the scene" as well as that video does!

[Cromfel]

It was really a revelation for me to see how poor condition the sound side is. The impression you get is that sound is "dead development" and everything is plug and play smooth and fancy. But no, it simply sucks. I hope the emergence of VR will force the audio side to take some leaps forward.

[crim3]

I just want audio that sounds as good as this: http://www.youtube.com/watch?v=IUDTlvagjJA (be sure to follow the instructions it recommends)
Is that too much to ask? I never realised how good audio COULD be prior to this, and now I wonder why on earth audio hasn't been pushed further. Games would be so much more immersive if the audio "put you in the scene" as well as that video does!

That video was a revelation for me, too. It was posted here some time ago. But it doesn't seem things are going in that direction. People likes too much the paraphernalia of a multi-speaker setup. What a sad gimmick.
Do you know this one? http://www.youtube.com/watch?v=WYdIidUIbAs
I love when you can clearly locate the matchbox somewhere at the middle of your back.

[Scog]

What we need is a system to capture a persons individual ear and head geometry (from something like the leap, or photogrammetry), and distill this into an hrtf using an advanced volumetric sound propogation simulation model. We could even subtract the hprtf (Headphone related transfer function), from a measured database of models.
To compliment this we would also want an optimised realtime sound propogation model for integration into game engines, the accuracy of which could be improved as resources permit. Features would include different sound shaders and texture mapping for the game environment, or maybe a voxel based mapping of the acoustic materials to match the structure of the sound propogation algorithim.

Tldr; physics based game sound - headphone colouration + individual hrtf = bliss

[bobv5]

Those youtube vids don't really work for me, but I still think hrtf is the way to go. (as a child I was mocked for the size of my ears, combiened with my teenaged love of very loud music might explain it...) Does anybody know if the algorithms have already been made for proper hrtf with different user ear profiles?

Would it be sane to run audio on a gpu, with openCL or some such thing?