• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

PS4 has AMD's TrueAudio DSP - Confirmed

Chumpion

Member
I'm loving this! My NuForce HDP + Sennheiser HD600 are ready.

If this indeed means that you can get 3-D audio from stereo headphones in most games, then all you need for awesome, high-end audio for PS4 is a headphone DAC taking an optical audio signal and a good pair of headphones. For about $500 you can experience sound quality rivaling a $10k surround system!
 
I'm loving this! My NuForce HDP + Sennheiser HD600 are ready.

If this indeed means that you can get 3-D audio from stereo headphones in most games, then all you need for awesome, high-end audio for PS4 is a headphone DAC taking an optical audio signal and a good pair of headphones. For about $500 you can experience sound quality rivaling a $10k surround system!

But what if I plug my headphones in my Mixamp pro and then on the PS4? Isn't it better this way?
 

Witchfinder General

punched Wheelchair Mike
I'm loving this! My NuForce HDP + Sennheiser HD600 are ready.

If this indeed means that you can get 3-D audio from stereo headphones in most games, then all you need for awesome, high-end audio for PS4 is a headphone DAC taking an optical audio signal and a good pair of headphones. For about $500 you can experience sound quality rivaling a $10k surround system!

Eh, that's not true at all. I have Sennheiser HD800 AND a 10K+ surround system and whilst the detail and separation through the 800 are spectacular it can't for a second reproduce the scale, impact and involvement that a full-blooded speaker set-up delivers.
 
Okay, DF tweeted back to me on the TrueAudio shebang.

452gdmS.jpg


Seems like there's still some parts about the whole thing that needs future clarification from Sony/AMD.
 

nico1982

Member
Seems like there's still some parts about the whole thing that needs future clarification from Sony/AMD.
The simplest esplanation resides in SF being a launch title. If they reused the audio engine they had in place on PS3 the quickest way was moving it to CPU. That or the framework they use doesn't support TrueAudio, yet.
On top of that, how many Xbone's titles do make full use of SHAPE & Co.? Have been there deep technicals presentation about it? Forza? I wouldn't be surprised if the situation is more of less the same...
 

imt558

Banned
Okay, DF tweeted back to me on the TrueAudio shebang.

452gdmS.jpg


Seems like there's still some parts about the whole thing that needs future clarification from Sony/AMD.

Who answered from DF? Leadbetter?

Smell like this shit :

Digital Foundry: The Xbox One has dedicated audio processing hardware that is not present in PS4 or PC. How do you utilise it in Ryse?
 
Dunno who replied. At least I tried.

At least we know that Killzone : Shadow Fall doesn't utilise TrueAudio. We'll see where things go from here.
 
Okay, DF tweeted back to me on the TrueAudio shebang.

452gdmS.jpg


Seems like there's still some parts about the whole thing that needs future clarification from Sony/AMD.

This does not mean that there is no TrueAudio DSP. It is normal that some audio tasks are performed on the CPU due to the lower latency.

From B3D:

Latency-sensitive functions would go to the CPU, non-senstive to the GPU, supported functions to the hardware chip, etc

http://beyond3d.com/showpost.php?p=1806199&postcount=920

And some interesting info from the presentation.

Some points from the presentation.

* The DSP/ACP will probably be rather locked regarding custom audio DSP lines because of the latency (minimum 2 frames) and fixed functions.

* The DSP/ACP will mostly be used for codecs and filterbanks for voice/echo cancellation processing (fixed functions on the Tensilica) etc.

* Even though he writes 'many' audio DSP functions can be delegated to the GPU - he could obviously only mention the two sets of workload suitable for GPU : simple (multiply = volume, addition = Mixing) and feedforward (convolution / fir / FFT). In the audio workstation environment - only 20% is related to these functions.

* The GPU is actual not suitable for most audio workloads because of the latency related to the compute queue. Only mid to high latency resistent audio functions could be delegated to the GPU - like a convolution reverb.

* The CPU is the prefered unit for audio DSP and the ONLY solution for low latency functions.

* The Jaguar cores are plenty fast - he do compare the performance to hand writen assembly SPU code, but was the x86 code vectorized CLANG/LLVM code or handwriten intrinsics?

*The gates are now open for high-end DSP functions from the workstation environment because of x86 - this is good news for us

http://beyond3d.com/showpost.php?p=1806169&postcount=913

Also on Jaguar performance:
*On Killzone SF: 300 voices [probably a rough number to give an idea] with ton of DSP's, auxillary effects, mastering effects, all that stuff runs in less than 50% of 1 jaguar core.
*Ultra low latency
*A quarter slower than SPU's (no surprises there) at same clock, lacks fused multiply accumulate (which was present on Bulldozer, and SPU's)

http://beyond3d.com/showpost.php?p=1806179&postcount=915
 

Chumpion

Member
Eh, that's not true at all. I have Sennheiser HD800 AND a 10K+ surround system and whilst the detail and separation through the 800 are spectacular it can't for a second reproduce the scale, impact and involvement that a full-blooded speaker set-up delivers.

I dare say it is true. The only thing missing is the impact because your body can't feel the bass from headphones. Scale is produced by the 3-D algorithms, and involvement should actually be enhanced, as audio can sound as if it is emanating from above you, or very close to you, for example.

I also have a 10k sound system but I really believe a pair of headphones can rival it at a fraction of the price, WITH the help of a competent 3-D audio algorithm.
 

Perkel

Banned
I dare say it is true. The only thing missing is the impact because your body can't feel the bass from headphones. Scale is produced by the 3-D algorithms, and involvement should actually be enhanced, as audio can sound as if it is emanating from above you, or very close to you, for example.

I also have a 10k sound system but I really believe a pair of headphones can rival it at a fraction of the price, WITH the help of a competent 3-D audio algorithm.

Problem with surround is its placement. To get best 3D effect you need not only to have good system but also you need good room for that. BOSE has amazing autoconfig for their surround systems.

There is no such a thing with headphones since they will always deliver good sounding 3D effect
 

Rad-

Member
If XO and PS4 have different audio systems, doesn't that most likely just mean neither will get used by third parties to their maximum potential?
 

Kleegamefan

K. LEE GAIDEN
Okay, DF tweeted back to me on the TrueAudio shebang.

452gdmS.jpg


Seems like there's still some parts about the whole thing that needs future clarification from Sony/AMD.


If PS4 has a Sound DSP, that does't stop anyone from producing a fully CPU driven software solution!!!


GG was one of the very first PS4 developers, perhaps the TrueAudio API/Tools weren't ready in time for the development schedule KZ: Shadow Fall was on...


KZ:SF using the CPU for sound != No custom Sound DSP on PS4

Try again DF...
 
Shape seems like an extension of TrueAudio, based on 4 instead of the standard 3 DSPs, customized for their use-case (KINECT) approach.

ding! ding! ding! We have a winner.

Marginally customized off-the-shelf hardware everywhere this gen. They basically needed to make the sound chip a bit more powerful to handle the natural voice audio processing that comes hand-in-hand with having Kinect right out of the box.
 
If PS4 has a Sound DSP, that does't stop anyone from producing a fully CPU driven software solution!!!


GG was one of the very first PS4 developers, perhaps the TrueAudio API/Tools weren't ready in time for the development schedule KZ: Shadow Fall was on...


KZ:SF using the CPU for sound != No custom Sound DSP on PS4

Try again DF...

What is it with outlets like DF and Polygon jumping to conclusions, always to the detriment of accurately representing the PS4? It is very strange.
 

haihoo

Banned
Is this real an DSP-Chip?

It seems that the DSP ist CPU calculated and not a dedicated audio processor. Many PCs use CPU for DSP calculations and AMDs TrueAudio can this too.
 

seanoff

Member
Is this real an DSP-Chip?

It seems that the DSP ist CPU calculated and not a dedicated audio processor. Many PCs use CPU for DSP calculations and AMDs TrueAudio can this too.

the whole point of true audio / shape is to remove the audio from the cpu, allowing the cpu to do other things or more of what it's been doing and allow for better sound design in games.

AMD-TrueAudio-Presentation-28_575px.jpg


AMD-TrueAudio-Presentation-26_575px.jpg
 

satam55

Banned
This presentation explicitly says 1.6GHz, and it's dated November 19:
http://www.slideshare.net/DevCentralAMD/mm-4085-laurentbetbeder

Does anyone have the PDF? I want to get a closer look at the Processor on page 3 I wonder if that's the PS4 design or just something they used for the presentation?


it look a lot like this one

5148d1349996342-ts1.jpg



Edit: never mind

diagram-dataplane-n-sm.jpg

I'm guessing that it's a DPU

What's a DPU?

slide-3-1024.jpg


diagram-dataplane-n.jpg



dpu-new.jpg

Those slides are very interesting and kind of confirm that the secondary ARM processor and the DPU we see mentioned are quite the same thing basically. Documentation in those slides and from Tensilica offer quite a lot of nice tidbits.

Reading those slides made me think about sound processing on the GPU and it's presented WHY's and WHY NOT's... I seem to recall people talking about AMD's audio technology built in PS4's GPU, yet this SCEA paper does not seem to make any mention of that when discussing the pros and cons of running game audio code on CPU vs GPU vs ACP/Secondary Processor... at an AMD conference nonetheless.

Curious don't you think?

ohhh.png
Did anyone else notice from those slides & info that the PS4 has a dedicated block for voice/speech recognition, echo-cancellation, & other enhancements like SHAPE on XBox1 does? It seems like ever since it was announced that the PS4's audio chip is based on AMD's Trueaudio DSP, all those advantages that SHAPE was going to give the Xbox1 over the PS4, disappeared.
 
I'm putting this up as speculation:

The Xtensa processor in the AMD APU used as a UVD codec is a DPU not a single CPU. At design you add the Tensilica DSP blocks you want to use to the Xtensa DPU and it can perform as a audio or Video Codec as well as TrueAudio for games and more.

diagram-dataplane-xtensa.png


The custom Tensilica DSP chip in the ARM second custom chip as Southbridge will handle Audio features like MP3 and DTS audio for the blu-ray player, in fact Tensilica DSP SoCs are included in some blu-ray players. But Sony has said the DSPs in the PS4 can do 300 MP3 streams at the same time. This I believe makes the Tensilica DSP a Dataplane Processors (DPUs) which combines the strength of CPUs, DSPs, and custom logic, often with 10-to-100X the performance. DPUs can do audio and video processing, as well as gesture and audio recognition.



If the PS4 is to support being a media hub and DLNA CVP2 STB it needs a low power "Apple TV" like SoC...the combination of the ARM second custom chip as Southbridge and a DPU both with access to 256 megs of DDR3 memory makes sense.

As a 4K blu-ray player which have voluntary power limits in the 21 watt range it works, as a VidiPath (DLNA CVP2) STB it should use slightly less power and again since coming blu-ray players will be VidiPath STBs this again works. As a media hub that transcodes from h.265 to h.264 to support other DLNA STBs in the home...it works.

Xtensa Customizable Processors. Sony has been using a Xtensa LX 2 since 2007 in their CE products.

Video describing Xtensa

Xtensa LX overview
 

a.wd

Member
Does this mean that audio management is on chip rather than streamed like the original og Xbox?

So I can have my music as the soundtrack properly rather than just playing it over the top?

If so Ps4 is becoming more like the og Xbox (in terms of hardware) by the day
 

Bricky

Member
This was to be expected with the GPU of the PS4 basically being the same as a modern AMD videocard which also have this feature, but good to hear it being confirmed. For people who don't understand what it does: It's basically a modern take on EAX/DirectSound3D that was available for PC's with a dedicated soundcard back in the day. So you'll get better 3D audio positioning and it allows for more complex sound design with more soundeffects playing at the same time.

It also means a little bit less pressure on the CPU, but I wouldn't expect any significant performance improvements from that so don't get your hopes up.

The only game utilizing the technology on PC is Thief though I think? It's a bit disappointing so far (it doesn't automatically enhance every game sadly, it isn't magic), but maybe some PS4 exclusives will use it (or are already secretly using it).

This is good news for Project Morpheus too, especially with Oculus focusing on 3D positional audio in their newest public prototype. The PS4 should be able to achieve the same effect with TrueAudio and a good pair of headphones.
 

Tetranet

Member
This is good news for Project Morpheus too, especially with Oculus focusing on 3D positional audio in their newest public prototype. The PS4 should be able to achieve the same effect with TrueAudio and a good pair of headphones.

It kinds of makes you wonder if they also had this in mind when deciding on this specific part of the hardware.
 

AndyD

aka andydumi
It kinds of makes you wonder if they also had this in mind when deciding on this specific part of the hardware.

It's interesting how many things are tied together so well. Not all of them may be exploited now or even soon, but at least it seems they had a vision and a goal, and someone in the room at all the decisions who actually understood that goal and made sure it was maintained throughout the process.
 

Ramza

Banned
It kinds of makes you wonder if they also had this in mind when deciding on this specific part of the hardware.

It really wouldn't surprise me. I think a lot of the PS4 hardware points to it. I've always wondered if some future thought of theirs is why they have the PS4 camera use a different connector rather than a regular USB one.....
 
It appears that Kaveri can support h.265 and it's UVD is also using a Xtensa DPU.

So in AMD APUs the decoder is a software programmable Xtensa DPU with software configurable microcode that can support multiple A & V codecs and likely any near future codecs but the encoder is fixed hardware encoding which supports fewer options but that clearly works and is what is necessary (less latency) for DLNA and HTML5 <video> (h.264) to stream to other platforms (Skype, game streaming, support for DLNA transcoding).

Kaveri's GPU with 12 CUs has too much leakage current to use the GPU shaders for UVD post processing so Kavari's UVD 4.0 doesn't use the GPGPU, it must have a more powerful Xtensa DPU that supports h.265.

Logically, Microsoft and Sony to support multiple Codecs and near future codecs must do as AMD did and use a Software codec. Sony is supporting AMD's True Audio APIs with the same Tensilica DSPs so Sony is likely using the same Xtensia DPU family (highly customize-able but likely Xtensia LX 4? 4 CPUs in papers needed for efficient real time h.265 decoding and LX 4 has 4 Xtensa CPUs) that Kavari is using.

AMD APUs also have eyesight gesture recognition code (and now facial recognition) in a ROM that uses the "Parellel computing" in the AMD APUs. Considering that Xtensia DPUs are touted as supporting gesture recognition it's likely running on the DPU. Sony will likely do the same for gesture and face recognition as well as AR.

I'm sure the above is already common knowledge to professionals in the industry but I'm just catching on. The use cases and EU and Energy star concerns dictate the use of ARM and handheld technologies which Tensilica DPUs are designed to support. Larger GPUs have too much leakage and can't be used for Software codecs and if gesture and voice recognition is to be supported by IPTV APPS not games then likely the GPU will be off.

Logically, Microsoft and Sony to support multiple Codecs and near future codecs must do as AMD did and use a Software codec.

List of supported XB1 DLNA and likely HTML5 file types and Codecs. This is a big differentiator for a game console as a Media Hub. MOV is an Apple video format so XB1 can support MOV and play or transcode to h.264. To support all these formats requires a software codec engine right?

www.neogaf.com/forum/showthread.php?t=872846 said:
3gp audio
3gp video
3gp2
aac
adts
animated gif
asf
avi divx
avi dv
avi uncompressed
avi xvid
bmp
jpg
gif
h264 avchd
mjpeg
mkv
mov
mp3
mpeg 1 ps
mpeg 2
mpeg 2 hd
mpeg 2 ts
mpeg 4 h264 aac
mpeg 4 sp
png
tiff
wav
wma
wma lossless
wma pro
wma voice
wmv
wmv hd
HEVC or h.265 is not now a DLNA required codec nor is it a HTML5 codec but for Sony to support a streaming IPTV service or 4K Blu-ray it's required.
 
Top Bottom