Any word on when Mantle + TrueAudio will launch for Thief?

[Phynaz]

The point is that the other methods are "faking" it. No one is disputing the fact that they work to provide positional audio. True audio actually uses the geometry data from the GPU to make the sound more true to life/the environment surrounding your character.

You realize that the software (game engine) already has the geometry data long before the GPU has it, right?

 

[VulgarDisplay]

You realize that the software (game engine) already has the geometry data long before the GPU has it, right?

The level designers had the geometry data in their head before the software had it.

Your point?

 

[Erenhardt]

Here is how I see the most important difference between Trueaudio and 5.1 dolby enhanced MS sound:

That, in competitive multiplayer would give unfair advantage to everyone NOT using TrueAudio over those who use it. TrueAudio users would be able to tell the vertical position of the sound source (or reflection source ) so its not that bad

 

[krumme]

Hardware acceleration isn't necessary for what they do today, which is near and far field HRTF transformations with a bit of reverb. But a higher quality HRTF takes up a lot more computation and beaming it around the environment and accounting for materials changes is definitely in the realms of ray tracing like computation overhead, its the same problem.

So while I agree what we do today doesn't require hardware acceleration we could have dramatically better sound if we had enough computation resources to do it, which we don't currently. Audio isn't done, its not a solved problem that can't improve, the very fact that a soundblaster z on 5.1 to headphones sounds is a lot more accurate for positioning than DX headphone mode tells us there is more to be done.

Microsoft didn't bring the 3D positioning into DirectX because there was no need for hardware anymore. It did it because its operating systems were getting crashed by two things - GPU drivers and sound drivers. They dramatically changed the driver interface for both, with GPUs they had no choice but to allow them in user space but minimised the interface they could have in the kernel to reduce crashing. But with sound they decided to do it all in software at a lower fidelity because it could be done on the CPU. There its stayed for years without any improvement. Before MS did that we were starting to see the emergence (from Aureal) of beamed environment bounced sound not just binaural surround sound and that was a big deal. Today with a sufficiently powerful card (I doubt TruAudio is that) we could have that and more.

I am constantly disappointed with surround sound today, its really poor quality compared to what we had a decade or more ago, a lot more can be done to produce compelling sound stage in games. A lot of it doesn't require these companies to do anything more, its the same interface as they have always been using (play a sound at position X,Y,Z with object moving at speed in x,y,z). The rest is all in the API. Clearly Microsoft isn't interested in doing this so its high time someone did, I am just disappointed its not Creative who should have been working on it for the last 10 years.

Agree. We have had the knowledge for 40 years and the technology for 10. DSP processing power have improved tremendously over the last years so i think there is good chance true audio is enough, but had creative done something about this the implementation would be here and rocking today. Now i guess we are still at least 5-7 years from any excellent use. It takes years to learn how to use it properly.

 

[krumme]

More about the 3 Tensilica dsp cores in trueaudio
http://pixelrant.com/2013/10/08/everything-you-wanted-to-know-about-amds-new-trueaudio-technology/

 

[OCGuy]

That, in competitive multiplayer would give unfair advantage to everyone NOT using TrueAudio over those who use it. TrueAudio users would be able to tell the vertical position of the sound source (or reflection source ) so its not that bad

If someone chooses to put the time into TrueAudio. Right now that list is short according to the link in the post above me. Also they are limiting which cards have access to it.

Might want to link the source for that as well...

 

[BrightCandle]

MPC: AstoundSound offers enhanced positional such as Distance and elevation/azimuth? Does it offer occlusion and reflections or wave tracing capability that Aureal 2.0 offered?
AMD: Distance and elevation are certainly within AstoundSound’s purview. For the other effects, you would have to reach out to them for clarification.

Well that is disappointing. The true HRTF is based on frequency, distance, elevation and speed - these are the 4 dimensions that need to be mapped in order to produce an accurate simulation of surround sound to our ears. Sounds like they are also missing frequency just like Microsoft did when they pulled it into DX. No they aren't doing beam forming so even more disappointment.

 

[krumme]

Well that is disappointing. The true HRTF is based on frequency, distance, elevation and speed - these are the 4 dimensions that need to be mapped in order to produce an accurate simulation of surround sound to our ears. Sounds like they are also missing frequency just like Microsoft did when they pulled it into DX. No they aren't doing beam forming so even more disappointment.

There is reflections and all sort of filters and compensation for the freqency for the look of the patens it seems
http://www.google.com.br/patents/US20100246831
Dont know if its implemented or how much power is actually available for all the stuff in amd implementation.

But another thing is just precise and usefull azimut and elevation is a gazilion times better than what we have now and will transform eg sound in bf4.

 

[motsm]

That, in competitive multiplayer would give unfair advantage to everyone NOT using TrueAudio over those who use it. TrueAudio users would be able to tell the vertical position of the sound source (or reflection source ) so its not that bad

Do you remember the source of that image, or did you edit the TrueAudio tag on it?

 

[OCGuy]

Here is how I see the most important difference between Trueaudio and 5.1 dolby enhanced MS sound:

That, in competitive multiplayer would give unfair advantage to everyone NOT using TrueAudio over those who use it. TrueAudio users would be able to tell the vertical position of the sound source (or reflection source ) so its not that bad

I really hope it wasn't you who did this......plagiarism/spreading false information/etc is not good.

http://precast.org/2011/02/quiet-please-2/

 

[motsm]

I really hope it wasn't you who did this......plagiarism/spreading false information/etc is not good.http://precast.org/2011/02/quiet-please-2/

Thanks for the info. We just had a discussion about whether or not TrueAudio actually supported early reflections in the other TrueAudio thread. There was no evidence that it did, so I was surprised to see that picture, obviously it has nothing to do with TrueAudio. Would have been cool if true though.

:Edit: This thread goes over some of our conclusions on what TrueAudio is known to be capable of. http://forums.anandtech.com/showthread.php?t=2372020

 

[blackened23]

I really hope it wasn't you who did this......plagiarism/spreading false information/etc is not good.

http://precast.org/2011/02/quiet-please-2/

Wow. Just, wow. Why anyone would do this, if not for misinformation, really is beyond me.

 

[96Firebird]

El.

Oh.

El.

 

[motsm]

Wow. Just, wow. Why anyone would do this, if not for misinformation, really is beyond me.

From what I've been seeing online, this is what most people assumed TrueAudio was going to be. A lot of people have been spreading it around like it's fact, so it's picked up steam. It's obviously not impossible that a developer could leverage the DSP's capabilities to create some sort of early reflection / ray tracing audio engine for their game, it's just not something AMD or any of it's partners has ever mentioned as a feature at this point. (AFAIK)

 

[96Firebird]

Maybe it is just me, but it seems like it would take a stupid amount of time to assign a reflection/absorption value (for sound) to each and every surface of a game.

Edit - But I guess all walls could be the same. Although not accurate, it would be a lot easier on developers.

 

[thilanliyan]

Maybe it is just me, but it seems like it would take a stupid amount of time to assign a reflection/absorption value (for sound) to each and every surface of a game.

Wouldn't whatever physics engine is being used in the game (assuming one IS being used) have those kinds of values already?

 

[wand3r3r]

I really hope it wasn't you who did this......plagiarism/spreading false information/etc is not good.

http://precast.org/2011/02/quiet-please-2/

That picture was cut so amateurishly that it's not surprising to me that it's fake.

I think the poster should explain this but that call-out and attempt to tarnish a reputation is pretty far beyond the forum rules.

 

[Keysplayr]

That picture was cut so amateurishly that it's not surprising to me that it's fake.

I think the poster should explain this but that call-out and attempt to tarnish a reputation is pretty far beyond the forum rules.

I think it's pretty easy to see that any further tarnishing is not needed after what was provided by the poster himself. Put the blame where it belongs please. I too would like an explanation (a plausible one) from the poster.

 

[motsm]

Maybe it is just me, but it seems like it would take a stupid amount of time to assign a reflection/absorption value (for sound) to each and every surface of a game.

Edit - But I guess all walls could be the same. Although not accurate, it would be a lot easier on developers.

It's not as hard as you might think. You just have to make filter presets like, wood, glass, dirt, carpet etc, each of which contains it's own EQ settings, and reflective/abortive properties. The texture artist would then simply tag his textures with one of those presets (which he is likely doing anyway so the game can properly assign matching bullet hole textures and hit sounds), then the ray tracing engine does the rest automatically.

I've built some simple impulse responses using these methods, and it's really a painless process.

 

[96Firebird]

It's not as hard as you might think. You just have to make filter presets like, wood, glass, dirt, carpet etc, each of which contains it's own EQ settings, and reflective/abortive properties. The texture artist would then simply tag his textures with one of those presets (which he is likely doing anyway so the game can properly assign matching bullet hole textures and hit sounds), then the ray tracing engine does the rest automatically.

I've built some simple impulse responses using these methods, and it's really a painless process.

Yeah I guess I didn't think about it like that. I guess it would sorta be like me assigning materials when designing parts for an assembly in CAD. All the material properties are already present, I just need to assign each part a material. Then the program can calculate an approximate weight, CG, etc...

 

[Erenhardt]

I really hope it wasn't you who did this......plagiarism/spreading false information/etc is not good.

WOW, so much fuss over the pic I pasted?

Over-reacting much? Plagiarism? Where did I exactly claim the picture is mine? There is a source on it, so it is definitely not plagiarism. Nowhere I claimed original research (which are not related to topic) as mine. Why the hostility again?

Seondly: "spreading false information" - do you have information that TrueAudio will not compute sound reflections or what?
Here is from anandtech arcticle about TrueAudio:

The advantages of utilizing the DSP are fairly straightforward. Simple audio calculations are cheap, and even simple 3D effects such as panning and precomputed reverb can be done similarly cheaply, but real-time reflections, reverb, and 3D transformations are expensive. Running the calculations to provide 3D audio over headphones and 2.1 speakers, or phantom speakers and above/below audio positioning in 5.1 setups is all very expensive. And for these reasons these effects aren’t used in current generation games. These are the kinds of effects AMD wants to bring (back) to PC gaming.

Nice ninja-edit blackened...

 

[Gloomy]

What just happened here? He clearly said "here is how I see the difference" in his post. And then he posted an MSPaint-edited picture as a visual aid. You guys are too hype right now, trying to call people out. What's up with that? :|

 

[motsm]

Seondly: "spreading false information" - do you have information that TrueAudio will not compute sound reflections or what?

AMD wouldn't specifically say what TrueAudio won't do. They have been advertising what it will do however, and they haven't mentioned ray tracing or early reflections, nor have I seen Wwise or Fmod make mention of supporting those features. I already said in this thread and the other TrueAudio thread however, that such features should be possible using the TrueAudio DSP, but they are also possible using modern CPU's, or even CUDA AFAIK. So it's just up to someone to create such an engine, perhaps TrueAudio will encourage such a thing, but it's a complete unknown at this point.

So for the record, I wasn't fussed that you used the images, but it is a bit misleading tagging them, since it makes it look they are features currently in TrueAudio.

 

[blackened23]

WOW, so much fuss over the pic I pasted?

Over-reacting much? Plagiarism? Where did I exactly claim the picture is mine? There is a source on it, so it is definitely not plagiarism. Nowhere I claimed original research (which are not related to topic) as mine. Why the hostility again?

Seondly: "spreading false information" - do you have information that TrueAudio will not compute sound reflections or what?
Here is from anandtech arcticle about TrueAudio:

.

You're attributing something to AMD true audio that AMD has never, NOT ONCE, stated in their briefings or press materials. I just gave them a once over via google. The quote you made is from an AT article stating what a DSP can potentially do. It was never stated by AMD that trueaudio will do this. It is interesting that you left the context out of the quote you took from that AT article.

But, instead, you take an internet image, modify it, and take credit for it to suggest otherwise. Then you take an AT article quote out of context to suggest that true audio does these things. When AMD has never made such a specification. And the article at AT which you quoted was merely specifying what a DSP can potentially do. When AMD describes true audio, they focus on the programmable aspect and they have never, to my knowledge, mentioned sound reflections or ray tracing. Yet here you are trying to make it sound like AMD has in fact done these things. I've looked over AMD's material for true audio and cannot find a source for this. Since you clearly manipulated internet images, i'm guessing that you can't find an AMD related source either.

It really brings past claims which were equally ridiculous into a new light. Anyway, as I said, true audio could be interesting even if i'm skeptical. Based on what i've read, it doesn't seem inspiring at all and no, I have seen zero mentions of reflections and raytracing which you're so quick to modify internet images over. Guess we'll have to wait and see when it's here. In the meantime, the games of trying to attribute something to true audio through circumstantial evidence or other sketchy means doesn't really do any favors. It is what it is. Let AMD do the talking on this. They have not made any mention of real time reflections or 3d transformstions. NOT ONCE.

Just to be clear once again: That AT article you keep quoting does NOT ATTRIBUTE real time reflections and transformations to true audio. It is merely a vague discussions of DSPs. To suggest that article attributes such things to true audio is disingenuous at best.

Let the games continue though. Whatever. It is tiring to even argue against such nonsense, not sure why I even bother.

 

[Erenhardt]

AMD wouldn't specifically say what TrueAudio won't do. They have been advertising what it will do however, and they haven't mentioned ray tracing or early reflections, nor have I seen Wwise or Fmod make mention of supporting those features. I already said in this thread and the other TrueAudio thread however, that such features should be possible using the TrueAudio DSP, but they are also possible using modern CPU's, or even CUDA AFAIK. So it's just up to someone to create such an engine, perhaps TrueAudio will encourage such a thing, but it's a complete unknown at this point.

So for the record, I wasn't fussed that you used the images, but it is a bit misleading tagging them, since it makes it look they are features currently in TrueAudio.

I don't have details about this tech, everything I know if from web articles.
I will cite anandtech artcile again:
(credits: http://www.anandtech.com/show/7370/amd-announces-trueaudio-technology-for-upcoming-gpus )

The advantages of utilizing the DSP are fairly straightforward. Simple audio calculations are cheap, and even simple 3D effects such as panning and precomputed reverb can be done similarly cheaply, but real-time reflections, reverb, and 3D transformations are expensive. Running the calculations to provide 3D audio over headphones and 2.1 speakers, or phantom speakers and above/below audio positioning in 5.1 setups is all very expensive. And for these reasons these effects aren’t used in current generation games. These are the kinds of effects AMD wants to bring (back) to PC gaming.

Additionally from http://www.anandtech.com/show/7400/the-radeon-r9-280x-review-feat-asus-xfx/4 :

Tensilica’s audio DSPs are task-specific programmable hardware, somewhere between fixed function and fully programmable in design, allowing for customized effects and processing to be done while still keeping the size and power costs low. The underlying hardware is programmable in C, while AMD for their part will be providing a TrueAudio API to access the hardware with. We don’t have a ton of details on the architecture of the DSP, but Tensilica’s product sheets imply that we’re looking at a VLIW architecture of some kind.

I guess that means the possibilities are endless, "the sky is the limit" (this quote is not mine either, credit to: unknown)
I agree that amd is not giving much details yet. There is barely anything on their web about TrueAudio. But from anadtech artciles, its clear they are aiming for some kind of soundwave physx.