• 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.

3D Scan your gaming area for enhancing gameplay

DrAspirino

Member
Nov 19, 2018
293
337
395
Chile
Most good audio setups have been doing this for many, many years.
Except that most of them aren't connected to their manufacturer server 24/7 and explicitly tells you what the connection to the network is used for. You can monitor their activity on your router (if you have a good one) and see that it doesn't "phone home", unlike what a console does.
 

baphomet

Member
Oct 29, 2011
6,038
1,073
755
Denver, CO
Except that most of them aren't connected to their manufacturer server 24/7 and explicitly tells you what the connection to the network is used for. You can monitor their activity on your router (if you have a good one) and see that it doesn't "phone home", unlike what a console does.
Do you understand the miniscule amount of information an audio/mic scan could tell anyone? You wouldn't even know the size of the room, let alone the shape of the room.

Yet people here seem to think Sony is somehow getting a fucking high res picture of your house.
 

DrAspirino

Member
Nov 19, 2018
293
337
395
Chile
Do you understand the miniscule amount of information an audio/mic scan could tell anyone? You wouldn't even know the size of the room, let alone the shape of the room.

Yet people here seem to think Sony is somehow getting a fucking high res picture of your house.
That's not the issue here. The issue is granting Sony access to the microphone. Sony, much like Microsoft and Nintendo, constantly monitor what players do on their consoles. There's enough of monitoring on their part while we're gaming for them to even listen to us without our permission, even when we're not gaming and the console is on "stand-by" mode.

Call me paranoid, but I already have a cover for the webcam in my laptop and tape in the microphones. I leave my cellphone in my room while I go dining with my family (and we all do that) so we don't risk family conversations being overheard by anyone.
 
Jun 28, 2013
1,471
1,063
780
The video isnt related. This is done with a camera to check your eyes.

This isnt new. Most high end systems already do this without even telling you. All they need is a microphone. They measure how long takes sound to reach your mic and at what strength. To make a audio model of your space.

What sony wants to do with headphones, they might need headphones that have some kind of scanner?
 

Ozzy Onya A2Z

Member
Apr 16, 2012
7,821
536
800
Melbourne, Australia
Pfft, take a look into WiFi 3D mapping. Shit is going to take off big time over the next 5-10 years. Literally mapping interiors based on WiFi signals. Fairly accurate already too e.g. width from wall to table is 1.5m away. Gaming will inherit the tech once it's accessible via routers/phones/consoles. Some weird AR/VR stuff incoming I'm sure.
 

onesvenus

Member
Aug 26, 2004
355
41
1,440
I think I'm missing something.

Foveated rendering works in VR because the display is so close to your eyes that some resolution is being "wasted" by being displayed on the peripheral vision area where humans can't see really well. In that case it makes sense to drop resolution in that area so power is not wasted.

Playing on a TV from your sofa is a completely different matter though. In normal conditions the TV will be in the zone gazed by the fovea 90% of the time. If you use foveated rendering on something that's completely in the fovea region, where everything you see is sharp, you are decreasing the quality of the image you are seeing and not taking advantage of how the eye works.

When reading text, the action where the fovea is more focused, gives a perception bound of 20 degrees. This means that changes outside those degrees are not perceived and everything there can be lower res without you noticing. If you are 2m away from your TV set, that gives us a region of 27 inches where everything needs to be at the highest resolution possible. This may give the idea of doing foveated rendering as something that can be done but perception of shape is 60 degrees from the center so that gives us 68 inches. This means that you are perceiving shape changes in all TV sets below 68 inches when looking at it from a 2m distance. Color and motion changes are perceived even on larger angles.

Another thing to take into consideration is the resolution of the position that could be determined by the array of mics. Array of mics are very good at detecting sound directionality but not really good at pinpointing locations. The best papers I found say that they can give you a ~30 degree estimation.

TLDR: it's almost impossible to do foveated rendering as the OP is suggesting without noticing loses in image quality
 
Last edited:
  • Like
Reactions: Zannegan

rofif

Member
Sep 13, 2019
2,446
3,123
490
SEriously, if there was a camera like trackir on top ov every tv or monitor, games could be so much more 3d
 

onQ123

Report me for starting troll threads. Ban warning.
May 1, 2010
15,913
4,251
1,275
Using direction of arrival with unique audio signature for object location detection

Abstract
Multiple independently movable devices are located upon a platform. On the platform is a directional microphone array. Each of the movable devices is assigned a unique audio signature, which may be a pulse train of chirps or a pseudo random signal or some other audio sequence that is unique to the respective device. Each movable device announces itself with its unique audio signature, and the platform's directional microphone system determines the location from which the unique audio signature comes from. For range, a difference between the time of flight (TOF) of the light relative to the sound signature is used. As another technique a unique audio signature may be an audio water mark concealed in normal audio which may be emitted by the movable device. The watermark provides the identifying information of each movable device.
What is claimed is:

1. A device comprising: at least one processor configured with instructions to: receive, via at least one sound sensor, respective sound signals emitted by respective movable objects in an environment, each sound signal establishing a respective audio signature unique to the respective moveable object; identify each of the movable objects based at least in part on at least one unique characteristic in the respective sound signal; and determine at least a bearing to each of the movable objects based at least in part on the respective sound signals, wherein at least one of the sound signals comprises a pseudo random sonic signal that is unique to the respective device.