Binaural Audio

[mAchiNE]

PalmerTech suggest I start a thread about this based on a longish post I made on another thread so here it is:

There seems to be a ton of information about HMDs and other VR related stuff on here but not a lot of info about sound. Often it seems sound is overlooked but after sight sound is probably the second most important thing for immersion.
This thread is an introdution to Binaural audio, the most realistic 3D surround sound you can have. This is basically 3D for your ears and only requires you to have stereo headphones, but will ONLY work when listening on headphones, on speakers it will sound like normal stereo. If you have not listened to a binaural recording i suggest you listen to this: http://www.youtube.com/watch?v=IUDTlvagjJA" onclick="window.open(this.href);return false; Binaural recording is done simply by recording sound from either a real head or dummy head with the microphones positioned where your ears are, this works the same way stereo 3D does for your eyes, it works by providing each of your ears a separate sound that they would get if you were actually there and your brain does all the hard work of combining those sounds together allowing you to pinpoint where the sound is coming from based on the slight differences in sound between left and right recordings. This give you the ability not only to tell if the sound is in front, behind, left or right of you but also above or below you. If you close your eyes it feels like actually being there. How does this apply to VR? well I am no programmer but I don't think it would be difficult (correct me if i am wrong) to emulate this in any game or virtual environment, you simply need to add a "dummy head" into the sound engine and output your left and right stereo channels from the left and right side of the virtual dummy head. I believe the Creative X-Fi CMSS-3D Headphone mode does this in games with open AL but i have not tried it for myself to confirm. This combined with a good HMD would provide a very immersive experience as far as audio visual goes. It is also possible to get a Dolby Headphone compatible sound card which simply uses an algorithm to convert a 5.1 signal into "binaural", this is far superior to so called "5.1 Headsets" that actually have 5 speakers in them, but as with depth enhanced pictures this is not as good as a proper Binaural recording (DolbyHeadphone is more like simulating the perfect 5.1 room setup on stereo headphones). You can get a great pair of sennheiser headphones HD555 for a reasonable price and if combined with one of these options will give you good 3D Audio without compramising on sound quality.

Any one who has any ideas on how we can implement Binaural audio into existing game sound engines, or any other info on this topic please post here
Also anyone who has actually used Headphone mode on a Creative X-Fi with CMSS-3D can you let us know how well it works?

[AntiCatalyst]

I think that's only useful in real world recordings. Game audio is already perfectly separated and you can basically pinpoint the direction of any sound in the virtual environment(anyone who's ever been a counter-strike nerd will verify this )

edit: ah, i spoke too soon! cmss3d apparently uses binaural cues to feed more realistic directional info to the brainz.

[seanny]

I'd look at IEMs (canalphones, etc.) for something that'll give you slightly more realistic playback of binaural recordings. Sometimes sticking things into your ears can be uncomfortable, so maybe the Koss KDE250 would be the next closest thing, being audiophile earbuds. I haven't used them though, but the principle seems to be the same in terms of bypassing the shape of your ear and pumping sound straight into the ear canal, as binaural recordings typically already involve someone's (or something's) ear to make the recording. I don't know how much this matters for games though... it depends how much their HRTF engine models the ear physically I suppose. If it doesn't, then normal headphones would be fine.

The thing that makes binaural tricky in games is the fact that there's not just one linear impulse-based convolution effect you can apply. It's the position of the ears, the pickup pattern of the ears, and the presence of the head occluding/reflecting/diffracting sound waves that makes the "HRTF" package. On top of that, everyone's head & ears are a little different, so that's another barrier to total realism. From a designer's standpoint, the best you can do is make a generalized HRTF, because the particulars of the user's head & ears and the user's model of headphones are always going to be ambiguities. If you had both perfectly mapped, maybe you could make a hyper-realistic HRTF filter for a particular person and his setup.

I've seen someone use a large set of binaural impulse responses pulled from the MIT website in order to make a Foobar2000 convolution-based plugin that allowed you to reposition virtual speakers. It was somewhat realistic (only somewhat), but made everything sound like, you know, mic recordings of music being played through speakers... a removed semi-"concert bootleg" quality. A little too much fidelity to the actual conditions of the impulse response recordings I suppose.

HRTF only relates to positioning... the other half of the realism picture is the reflections that define the acoustic space for the listener, conveying information about where the sound is, where the listener is, and where the walls are (and what material they're made of) in a particular space. A friend & I have a rough prototype of an audio engine for games that can simulate tens of thousands of individual reflections and run them through an HRTF model. As far as we know, such a game engine hasn't existed since Aureal "A3D" (which was capable of 1-bounce reflection, occlusion, and HRTF, all using a low-pass filter bank). We're working on jamming our experimental engine into (open source) Quake 3, and eventually the Unreal engine (as a plugin, if it can be done) in the future. This'll take months, if not years though.

[cybereality]

I totally agree, audio is half the experience. But you don't need any fancy emulation or drivers. Just get a pair of real 5.1 headphones. The top of the line gaming headset right now is the Psyko Carbon ( http://www.psykoaudio.com/index.php/Carbon.html" onclick="window.open(this.href);return false; ) which just came out. Basically what they have done is created a headset that supports the full 5.1 surround sound without any funky emulation or simulation. There are 5 speakers and 2 sub-woofers in the headset. Unlike some cheaper 5.1 headsets (which place all the speakers on top of each other in the ear-piece) they place the speakers on the top head-rest. Then there are tunnels that channel the audio waves to the ear-pieces. This means that the sound is coming from the proper direction, and the physical pathways are different distances too so the mind can better estimate distance. Really, this stuff is state-of-the-art, and I really believe it works. They headphones are pricey, but should be worth every penny. I am planning on buying one soon, so expect a full review shortly if you are on the fence about it.

[mAchiNE]

Yeah I saw that Psyko Carbon headset online about a year ago and the concept looked great, looks like it will actually work to provide good 5.1 surround sound unlike the zalman 5.1 headset and others like it. 2 problems with this head set though, firstly its very bulky so I don't know how you will be able to fit it on your head along with a wide FOV HMD, secondly it is still only 5.1 surround which is not true surround (only has 5 directional sound zones) I think the difference between 5.1 and binaural for your ears is like the difference between a true stereoscopic photo and a 5 level depthmap enhanced 2D Photo conversion for your eyes. But I guess untill the sound engines in the games become alot more complex this Psyko Carbon headset will be the way to go, maybe you can build a custom HMD around this headset, after all it is large enough to be the base for one
Anyway I will look foward to hearing your thoughts on it once you get your hands on one.

[crim3]

I've never experience sound localization in games like I have in real life. Only some diffuse sense of where the sound is coming from. Neither with multi-speaker setup or the creative tactic3d headphones I have from some months ago (they claim it to be really 3D, that is, the up and down dimmension is also taken into account).
This video is the most authentic experience of sound localization that I have ever had
http://www.youtube.com/watch?v=WYdIidUIbAs" onclick="window.open(this.href);return false;

I'd like to see it emulated in games.

[PalmerTech]

http://gilmore2.chem.northwestern.edu/t ... 1_tech.htm" onclick="window.open(this.href);return false;

That is a good read. Corporate sponsored, by the now defunct company Aural that seanny mentions, but still a great overview.

Cyberreality, that is a really, really interesting looking headset! I look forward to your review, I would definitely be interested in something like that. Some of the claims are dubious, though, like how they claim that the longer sound channels for distant sounds mimic the real world because the sound "takes longer" to travel to your ear; I suppose it is true on a technical level, but an extra few inches does not do anything compared to tens or hundreds of feet, considering the speed of sound. Either way, while it might work well, surround sound headsets will never rival the audio quality of normal stereo headphone; There is a reason that no audiophiles use surround sound headsets. Either way, Machine's comparison of 5.1 surround sound to a

One of the biggest problems with Binaural audio is that every person has a slightly different hearing profile. We can get really close with good recording, as shown by how easy it is to pinpoint people in Counter Strike, but it is not truly immersive. Once you hit a certain level of realism, the perception of being "in the game" skyrockets, much like a small 3D display will let you technically gauge depth, but not really feel immersed.

The Air Force, when training new pilots for combat, sets up a hearing profile manually. Basically, they put them in a huge sphere that is full of hundreds of speakers, and they play a "ping" noise through each one. The pilot then indicates the exact direction that it appeared to come from. By repeating this hundreds of times for each speaker, they can build an exact profile of how that particular pilot positions sound. They do this so that they can load that profile into the jet's computer when they fly, so that target acquisition and missile warning beeps can appear to be coming directly from the source, and allow the pilot to react more naturally. I believe NASA does the same kind of thing, but I do not know much about that.

In theory, this process would be possible to work around the HRTF differences between individual gamers and differing playback equipment, through a length calibration that each gamer would have to go through, for every single sound device they use.

Seanny, it sounds like you have a lot more experience than most of us with this. Any information about your experimental engine? I know of a few different engines out there that try to do this, but none of them do it with any particular success.

[mAchiNE]

@seanny
Thanks for your input, you suggestion of Canalphones (or in ear headphones) for better binaural could be good as it will work well with a HMD, as for the comfortability of prolonged use, you can get mould made of your ear canal at some specialist audilolgists and they can make ear plugs the exact shape of your ear canal meaning that there is no pressure on the ear canal so you can wear them pretty much indefinately without discomfort, they can mould in ear headphones into these as well so you can get a custom pair of audiophile canalphones that are comfortable for prolonged use.

@crim3
Yeah that is a Binaural recording which is why the sound localization is so good

@PalmerTech
That looks to be an interesting read i will read it properly when I have more time

One of the biggest problems with Binaural audio is that every person has a slightly different hearing profile. We can get really close with good recording, as shown by how easy it is to pinpoint people in Counter Strike, but it is not truly immersive. Once you hit a certain level of realism, the perception of being "in the game" skyrockets, much like a small 3D display will let you technically gauge depth, but not really feel immersed.

Every person has a slightly different vision profile as well but in general stereoscopic 3D works for 80% of people with or without serpeation adjustment, I think this applies to Binaural audio as well to a certian extent. As the Virtual barber shop was not recorded with my specific hearing profile but if i close my eyes then it feels like I am there when listening, even with over ear head phones as opposed to in ear, not to say that it couldn't be better if customized to each user and their setup but i think you can still get it good enough based on the average hearing profile that 80% (at a guess) of people will find it very immersive and be able to pinpoint sound directionality with good accuracy.

[cybereality]

The thing is, I think the 5.1 surround sound audio engines used in games basically virtualize what a binaural recording would be doing. Maybe some engines are better than others, sure. But they do provide something more immersive than stereo sound. I used to use a 5.1 headset, mainly for Counter-Strike, and it did provide a good spatial awareness. Granted they were a low-end model (I think I paid $60 bucks) but they easily gave a superior experience than standard stereo headphones. You really could hear where people were stepping behind you, and turn around and get a headshot. In fact, people would sometimes accuse me of hacking because the headphones were so good. The Psyko headset should be a whole lot better than these cheap ones (I hope). I am not expecting it to sound as real as real-life, just like a stereo 3D HMD is not going to fool you into thinking its real-life. But its a step in the right direction. One step at a time.

[mAchiNE]

Cyber I agree that it is a step in the right direction but we can always hope fore more.
I think 5.1 headsets are emulating 5 speakers in a room (at least the 5.1 output in the game the headset is using is assuming 5 speakers and a subwoofer in a standard Dolby layout, not specifically desinged or tweaked for a 5.1 headset, and you can't have any sort of Binaural audio on a speaker set up) The surround sound in 5.1 is only horizontal sound (correct me if I am wrong) so you may be able to tell the sound is coming from your rear left but not from above to the left or below to the left (which CMMS-3D claims to do and if nothing else allows you to easily pinpoint the enemy in FPS) Traditionally 5.1 headsets were designed to give fps gamers a competitive advantage with directional sound but were not designed for good sound quality, having awesomely accurate directional sound will not be so immersive when everything sounds like you are in a tin can. Emulation is not much better for retaining sound quality.

Here is an example of CMSS-3D in action watch with headphones:
http://www.youtube.com/watch?v=dcB9faNrz_M" onclick="window.open(this.href);return false;

But so far it seems getting good surround on a head set always compramises sound quality (weather it be a 5.1 headset or surround emulation) Hopefully the Psyko Carbon solves this problem!
One thing to note is that Dolby Headphone supports 7.1 emulation (more sound zones than a 5.1 headset) although how well it actually works is questionable
Suposedly "Official" Dolby Headphone Demo here: http://www.youtube.com/watch?v=jL1AdBng ... re=related" onclick="window.open(this.href);return false;

Cyber when do you think you will have the Psyko Carbon? I await eargerly to find out if it is as good as it looks like it will be as it is cheaper than I was expecting.

[crim3]

What it's fascinating is how 2 channels give a completely realistic experience whereas 5.1 or even 7.1 doesn't. Surround systems are a waste of resources, and only achieve a suboptimal solution.

In my experience, multi-speaker setups just give an idea of which speakers are sounding louder. Then you think things like: "well, the rear speakers are sounding louder, so I'm suppose to feel that the object represented by the sound is behind". I've never felt amazed by that, not even in theaters.
With those binaural recordings the perception process of location of the source is the same than in real life. Not which speaker is louder or quiter, only real sound localization. And, in a natural way, only needs 2 channels, as we only have 2 ears.

[tritosine5G]

Yeah, THX, 5.1 , 7.1 is bullcrap, ideally what you want is constant directivity stereo and up to 4-5 subwoofers scattered.

Loudspeaker in a room = room modes , faulty design (conventional piston loudspeakers are bad http://www.gedlee.com/downloads/directivity.pdf" onclick="window.open(this.href);return false;)
Headphones = zero room modes , and this is problematic too

headphone ppl used to have circuits that does the mixing between L and R , depending on some filter shape.

Personally I want my low-end frequencies intact ( as in, 30-200hz ) and quality high frequency, so I choose to deal with Room modes. Remember headphones are always an option.

Best way dealing with room modes is modeling the room and simulating the waves themselves but needs a lot of time or supercomputer perfomance. Maybe with new graphix cards its coming though. I'm waiting for a while now damnit.

[seanny]

I wonder how well of a bootleg surround sound headset a handful of 2" extended-range speaker drivers would make, heh. Their heavy magnets would be overkill for headset levels... but still

Seanny, it sounds like you have a lot more experience than most of us with this. Any information about your experimental engine? I know of a few different engines out there that try to do this, but none of them do it with any particular success.

We've also tried to find engines similar to ours, but yeah.

I don't want to say too much about its particulars. Our ultimate goal is to make it the default choice for game developers, so we're mulling over how exactly to release our engine. The engine itself has come a long way, but it's not ready for prime time. The current version is only capable of 2-bounce (or 2nd order) reflections for example. We're thinking of some hacks that will allow it to do hundreds of bounces for long-tail reverbs.

Audio for games hasn't really advanced since the 1990s compared to other aspects like AI and graphics, so we saw an opening to make an engine that can simulate an acoustic space using the actual game world geometry & materials. There was a wow factor to our first functional prototype, even though it was full of glitches and had no HRTF model (the ears were like omni mics, so there was no directionality). You could walk around a room with a sound source playing and tell how far you were from the walls. Move to a corner and hear the bass boost like mad. Move to a cramped space and hear a "boxy" sound. Place a wall between you and the source and hear only the "room sound". It was way more impressive than the Aureal A3D stuff because we were dealing with several thousands of reflections.

It's a vast departure from the current mode of game audio, where each room/environment is a preset on a virtual mixing console of sorts with effects/reverb sends. The only part that's ever simulated is the positional stuff, but the rest is stuck in a paradigm of sound mixing from half a century ago. Our engine would be a whole new paradigm, because suddenly the acoustic space matters beyond which reverb effect preset to load.

Anyway, as far as HRTF/binaural is concerned, we implemented a human ear pickup pattern for the two virtual ears and it sounded pretty decent. Because our model included simplistic filtering for making a sound brighter or duller (based on the angle of the source to the ear) and not just attenuation, it managed to convey a sense of sounds being "above" or "below". Theoretically you could convey a sense of up/down in a 5.1 setup as well, but part of me wishes Ambisonics took off so the true nerds can have like spherical arrays of speakers for today's games & movies, heh. Maybe our engine will support arbitrary setups someday.

Theoretically you could import human head geometry into our engine for hardcore HRTF, but that's a whole can of worms we won't open until way later. At its best, that sort of HRTF might be realism overkill for anyone not using in-ear monitors. Many games (and other artistic products like movies, music) aim for heightened/enhanced/idealized realism rather than actual realism. When we first imported the human ear pickup pattern for example, we found that it was too detailed and specific. It made sense to someone's ears but not ours, so we averaged it out a bit to make it more general, and reduced its power so that it could be used with headphones & speakers alike.

[mAchiNE]

@seanny
That sounds promising and that would remove the need for bulky 5.1 headsets, you could always put in a "headphone mode" where the power is not reduced as much. Good luck to you and I hope one day to see this in some up and coming games!

[cybereality]

I still think that a proper 5.1 headset will have a technical advantage over a simple stereo headset. I admit sometimes the sound quality is not as clean as stereo sounds, but it does seem to provide a more immersive experience. If you have the proper processing of the sound on the software side, I think there is more potential with more speakers. Maybe no one has done it yet, but that doesn't mean it can't be done.

[mAchiNE]

@Palmer
I read that 3-D Audio Primer by Aureal Corporation link you posted, very interesting read.

@ Cyber Sometimes more isn't always better, if you haven't already, I suggest you read the document PlamerTech posted a link to further up this thread, it should explain why we only need 2 speakers to achieve 3D surround sound, after all we only have 2 ears. Sure 5.1 is much better/more immersive than standard stereo left right format (which is only one dimensional) but this does not limit stereo headsets to non 3d sound. What I am talking about is Binaural not Stereo, and no games that I am aware of implement a Binaural Headphone mode, the best thing you can do for comparison is listen to the Virtual barber shop link i put in the first post on this thread with a stereo pair of headphones (listen with your eyes closed) you will see what true surround is and this is possible to do in games and virtual environments with the proper processing of the sound on the software side. I dont think its necessary to have extra speakers in a headset to complicate things (and make them more expensive) when 2 will do the job, sure if you are talking speakers in a room the more speakers you have the better the surround, but in a head set it is another story because the speakers are in a fixed position relative to your ears and if the sound engine is complex enough it can emulate binaural sound.

But that being said at this point in time the game sound engines are not complex enough and the attempts to emulate surround (Dolby headphone) or binaural (Creative CMSS-3D) in these engines seem to destroy the sound quality so 5.1 is the best option at this time especially if the Psyko Carbon has good sound quality as well as sound localisation, at least untill such time as seanny gets his sound engine perfected and into the mainstream

[AntiCatalyst]

that virtual barber shop sounds just like regular stereo to me. sure i can hear him walking around, but I'm never sure if he's in front of me or behind. I'm using regular in-ears.


I'm sure it's very good for a recording, but compared to rendered stereo in a game, it's not groundbreaking or anything, apart from the obvious advantage of it being, well, real so you get different sound "moods" as the sound scatters differently from different places in the room


i have to say i think 5.1 headsets are better(never tried though ), because they'll fit anyone's ears. Instead of emulating a direction based on a general hearing profile, they're physically sending the sound wave from that direction and letting the ears do the rest.

[crim3]

that virtual barber shop sounds just like regular stereo to me. sure i can hear him walking around, but I'm never sure if he's in front of me or behind. I'm using regular in-ears.

Something like that happened to me the first time I heard that video. I was at my job place. It turned out that the crappy realtek sound of that computer was outputting only the left channel through both left and right channels. When I fixed the problem and heard it properly it was such a shock that now I'm a believer forever more.

[AntiCatalyst]

That's not it i should clarify.. i can pinpoint his direction, but that direction could either be in front of me or behind me.

[cybereality]

Yes, I have heard that barbershop clip from a while back (but I listened again just to refresh my memory). Yes, it is pretty amazing. And the matchbox clip that was also posted was good too. That one actually had sound from above and below, which I don't think is possible with stereo (at least I've never heard it before). But what I don't understand is why game developers aren't doing this already. That Aureal document was from the 90's. They've known about this for quite some time. Is there just not a big enough market? Do people not care about audio (or have a "good enough" approach). Or is this the case where they just hate anything with 3D in the name? 3D audio!?! Must be a gimmick!

Either way, I still have high hopes for that Psyko headset. Might just order it when I get paid tomorrow. If it sucks, whatever. I'll just sell it on eBay. I have to imagine its better than the $50 stereo headphones I'm using now. I am talking about out-of-box with existing PC games. Not including whatever theoretical quality you could get if there was a binaural audio driver for gaming (of if developers supported it natively).

[PalmerTech]

People do not care/understand it, so the market is not that big. Furthermore, of the pool of people who DO care, very few of them are qualified to actually implement it. I don't think there are enough people with the right skills in the right places. And aside from that, the entire sound industry has focused on speakers, from stereo setups to surround sound. Music, movies, and often videogames are almost always mixed with speakers in mind, not headphones.

I mean, imagine how easy it would be to record movies with binaural audio, using a dummy head. If they cannot be motivated to do something that basic, then how can we expect a much harder to accomplish software solution?

[mAchiNE]

I think people don't care because they don't know any better, most people have not experienced high quality sound (especially in headsets) most people probably assume their iPod headphones are good quality.

Cyber if you give the Psyko Carbon the thumbs up I might get one and build a DIY HMD around it using the HV056WX1 LCD screens for a High FOV HD HMD with surround sound. Just trying to decide weather or not to get the parts to do this or a 3D projector first

[PalmerTech]

If 3D is important to you, I think a projector might make more sense. These DIY HMD projects never go as planned, so if you are counting on it being your main 3D solution, I think going for something more reliable makes sense.

In the meanwhile, you can watch all of us spin our wheels figuring this out!

[cybereality]

Dude, get the projector. All this DIY stuff is more like a hobby than anything. It shouldn't be your main setup.

Also, I just bought the Carbon headset. I hope its all that its cracked up to be. It certainly better be for that price, OMG!

And, yeah, most people probably just don't know what they are missing. Same with stereo 3D visuals, I guess.

[mAchiNE]

http://www.head-fi.org/forum/thread/500 ... dio-system" onclick="window.open(this.href);return false;

Conclusion

So what are we left with? To be frank, I was really looking forward to this headset, and was hoping to be blown away. The theory seemed interesting and sound, and Psyko definitely fulfilled on the directional aspect of this headset, like I said, it is an improvement over the conventional 5.1 headsets in my opinion directionality wise. And looking at the previews of the headset, I thought “Hey, with the drivers being in the headband, this can solve the…stigma 5.1 headsets have had with the low sound quality from cramming smaller drivers into the ear cups, since you could fit larger drivers into the headband, but then Psyko has taken a turn into a stigma of their own with lower sound quality due to the distortion the sound gets when it passes through those wave guides. So as I was saying, with all of this said, what do we have? As it currently stands, I would not choose this headset over the other ones currently in my arsenal. I just don’t think the sound positioning compensates enough for the loss of quality when compared to normal stereo headsets and Dolby Headphone, especially when factoring in the prices. At 300 dollars, I wanted the Psyko 5.1 to deliver better audio quality and better directional positioning over conventional 5.1 headsets, and I only found it delivered one of those. However, I do believe that the concept Psyko has created is an idea worth investing in. The directional aspect is down, but in future products, I think Psyko just needs to do some research on the distortion of sound when it’s going through those waveguides, and perhaps find a way to distort the sound right out of the gate so when that distorted sound goes through the waveguides, it sort of gets distorted…back into a viable sound…if that makes any sense. It does recreate a 5.1 system quite well…but it’s not the Polk Audio 5.1 quality I was hoping for. But give the company and the concept time to grow, and hopefully we can expect a better performing implementation for a more consumer friendly price.

I'm always interested in your reviews Steggy. I read this but didn't have the chance to truly reply until now.

I was on a massive quest to find headphones that sounded like real surround, and I have tested so many that I lost count. Out of all the ones I tested/own, the K701 (with Dolby Headphone) is easily the ones that blew me away the most.

As for true 5.1 headphones, I had the Tritton AX Pros, and was in awe at how... mediocre they were, even in giving you that surround sound feel. I bought them after reading so many reviews about how insane their surround sound positioning is, when in fact, I find the good old KSC75 to pull surround sound off better, lol.

I was interested in the Psyko, just out of morbid curiosity, so thanks for your thoughts on them. At least I know I wasn't really missing anything.

Try out the K701 or K702 with something like the Mixamp. You'll be in awe out how well they pull off surround sound.

Quotes from the above link which is a review of the original Psyko 5.1 Headphones over on the Head-fi fourms, sounds like they have good directional sound but bad soud quality, sorry Cyber this info is bit late for you but hopefully the new model Psyko Carbon you got addresses this sound quality issue the original set had. Otherwise it seems a good set of stereo headphones and dolby headphone surround emulation is the way to go on this for now.

[mAchiNE]

http://www.head-fi.org/forum/thread/534 ... -headphone" onclick="window.open(this.href);return false;

Here is an interesting write up (again on head-fi) comparing a bunch of stereo headphones using dolby headphone 5.1 emulation and the relative sound/directional quality.

[PalmerTech]

I owned a K701, and I currently own UR40s, AD700s, RE0s, and while I do not own the KSC75, I do own Koss PortaPros, which use the exact same driver. I usually use my AD700s for gaming because, like he said, they have a HUGE soundstage. They lack in bass, yes, but mine are pretty heavily modded to add a bit of thump to them. That, along with the fact that I run them out of a FiiO E5 amplifier with hardware bass boost enabled, and I could not ask for a better gaming headset. They are large, but all that goes to being VERY comfortable, and they are actually very lightweight. Here are some pics of mine, done some costmetic upgrades as well:




You can get them at Amazon:
http://www.amazon.com/gp/product/B000CM ... B000CMS0XU" onclick="window.open(this.href);return false;

But eBay is probably a better bet. Go for a used pair, because some of the new ones are actually fakes. The K701 is nice, but I do not really like the soundstage as much, the AD700 is the soundstage match for an HMD with 120 FOV. Anyhow, that is my little rant on headphones.

[cybereality]

@mAchiNE: Well, I've been using the Psyko Carbon headset for about a week now. Its not as good as I hoped it would be, but it still works nice. In the games that support it, the surround sound is a clear advantage and really adds to the immersion. So far the best game I've played with it is Metro 2033, it really sounds great. Left4Dead also works well. Its pretty easy to tell what direction sounds are coming from, and you can quickly turn a shoot. So that part does work as advertised, though its not always as immersive as I would hope. The amp also provides a nice level of bass, which makes the gunshots feel more real. However, the overall sound quality is not that great. Even a cheap pair of $40 stereo headphones will probably sound better in general. Sounds on the Carbon are very hollow and tinny at times, and voices can sound muffled. At first this was really annoying, but I have gotten used to it. So I still like them, and I think they do enhance the gaming experience. I'm just not sure they were worth $200, especially since they are crap for anything but 5.1 gaming. Maybe I will writea more detailed review when I get a chance, but thats the gist of it.

[JMX]

Has anyone tried the Aureal A3D cards? I used it with Half-Life 1 and mods and it was pretty amazing. With headphones you can tell where a sound is coming from in full 360 3d, above or below. I got pretty good at lobbing molotovs and stuff over walls/onto roofs at someone moving around. They were always surprised at how accurately they could be heard. When I built my last PC I was trying to get another A3D card but Aureal apparently went bankrupt from lawsuits by Creative and the drivers aren't supported anymore. I haven't heard of (or heard) anything as good as the A3D tech, except for that barbershop video.

There's a bit more on them here:
http://en.wikipedia.org/wiki/Aureal_Semiconductor

Hopefully seanny's work pans out! I guess beware of Creative and be legally prepared for anything. Maybe some kind of open source 3d audio engine could be made.

[EdZ]

Rather than festooning your head with an array of tiny speakers, it seems that in-ear monitors and a HRTF are more effective, possibly with some additional user tuning to the HRTF to deal with minor variations in the pinnae. Binaural audio does not work (or at least, barely works) with over-the-ear headphones, as you're trying to pass audio recorded inside virtual/actual pinnate through your own pinnate again, resulting in unpredictable weirdness occurring to the sound.

Most sound cards and on-motherboard sound chips (AC97 and HD-Audio/Azalia), and every one I've encountered in the last half-decade or so, have implemented an in-driver HRTF filter, that takes multichannel sound as an input and gives you a very good facsimile of binaural audio. Many games have a 'headphone' mode in the sound settings that often (but unfortunately not always) means HRTF. Games using OpenAL can also be used to provide true 3D HRTF (IIRC, includes Unity, Unreal 3, and ID Tech 4).

[Chriky]

I came across an open source library to render binaural stereo. I made a post on the R&D board - http://www.mtbs3d.com/phpBB/viewtopic.php?f=138&t=15307