3D Audio (crosstalk development from Princeton)

silent_riot · Postby **silent_riot** » January 29th, 2011, 11:11 am

http://www.princeton.edu/engineering/video/player/?id=3990

Professor Edgar Choueiri @ Princeton has developed a DSP to obtain 3D audio from Stereo Speakers.
Essentially, he has eliminated crosstalk at the listener's position via processing.

Take a look at the link and listen to the video, there are demos of the effects (even the closing music).
Do not use headphones, use your desktop/laptop speakers.
Previous demos (those who remember the stuff from SQS), have only worked with headphones.

Note, that you can physically achieve this effect (without DSP) by putting a wall between the speakers, up to your seat.

Jonathan

silent_riot · Postby **silent_riot** » January 29th, 2011, 11:25 am

3stagevtec · Postby **3stagevtec** » January 30th, 2011, 3:46 am

Nice find! Even on my netbook, I can pick up the strong 3D image..

Will have to play that video over later on my desktop, would love to see that technology go mainstream! no more need for expensive multi channel speaker systems..

silent_riot · Postby **silent_riot** » January 30th, 2011, 9:01 am

The whitepaper for those who have the balls...

http://www.princeton.edu/3D3A/Publications/BACCHPaperV4d.pdf

southside connections · Postby **southside connections** » January 30th, 2011, 9:46 am

i had the balls to read it

then i saw the math

i got an understanding of it, but not a working , fully understood grasp

hmmm, will get it right soon

silent_riot · Postby **silent_riot** » January 30th, 2011, 10:23 am

Thanks for the laugh, southside.

but the math is irrelevant.

The abstract, introduction and summary are in simple language and are the only things I would expect anyone on this forum to understand (including the postgrads).

For those who read the paper, focus on those three sections and let me know your thoughts.

The importance of this technology is the simplicity of its application (an addition of a filter to a stereo system). This would mean the end of 5.1 systems and the dawn of 3D audio from 2.1 systems.

It could mean a new platform for entertainment and communications technology (as seen in the videos).

IMO, this is revolutionary, and should become mainstream once the technology is adapted to a multiple listener environment.

Sully · Postby **Sully** » January 30th, 2011, 11:24 am

We already have processors that can auto tune a system for tonal accuracy, this would help with imaging immensely.

silent_riot · Postby **silent_riot** » January 30th, 2011, 11:33 am

What you mean, Sulls? For the correction of tonal accuracy as a concern in the whitepaper?

Sully · Postby **Sully** » January 30th, 2011, 7:16 pm

Just like the newer processors have microphones that help auto tune for tonal accuracy, this technology can be used to assist with auto staging (I'm assuming that the algorithm is going to assist with image placement).

silent_riot · Postby **silent_riot** » January 30th, 2011, 9:11 pm

^^Actually, this is an entirely different approach, but as mentioned in the whitepaper tonal accuracy is an initial concern which would need to be fixed after the crosstalk was removed.

southside connections can probably explain the algorithm to you.

southside connections · Postby **southside connections** » January 31st, 2011, 3:07 pm

alluh reallll firetruck up in here unno

i tired tell people, yall will regret this soon

Sully · Postby **Sully** » January 31st, 2011, 4:34 pm

southside connections wrote:alluh reallll firetruck up in here unno
i tired tell people, yall will regret this soon

Regret what?

Postby SR » February 1st, 2011, 9:04 am

yah regret what

i like the concept

but how does it affect tonal accuracy

as well as in a normal listening environment your ears are acustom to hearing "crosstalk" no matter the source

silent_riot · Postby **silent_riot** » February 1st, 2011, 10:02 am

Well the premise of this technology is to recreate the listening environment at your ear lobes by reducing the effects brought on by stereo loudspeakers. i.e. controlling crosstalk.

Sound reproduction via stereo speakers inherently introduce too much crosstalk. Natural sound (i.e. not sound reproduction) does not contain any crosstalk because there are no stereo sources to begin with, that's the aim of this technology.

With regards to tonal accuracy, there is a big difference, you can even tell from the videos things sound tonally a bit worse than normal stereo reproduction- this is a concern he admits and addresses in the whitepaper (Introduction- section B). All in all, this is very promising technology.

Those who remember the "Virtual haircut" from SQS, this is similar idea but different implementation.
Virtual haircut introduces crosstalk into headphones (which have zero crosstalk to begin with) to recreate the environment.

Postby SR » February 1st, 2011, 10:17 am

keep in mind all "stereo" amplifiers will introduce some form of xtalk from the signal path on the electronics

thats why some purists belive in using dedicated left and right amplfiers

or some amplifiers are actually 2 separate amps in one chassis with separate power supplies

Lance · Postby **Lance** » February 1st, 2011, 10:06 pm

The practical application for this technology is limitless..

silent_riot · Postby **silent_riot** » February 2nd, 2011, 6:28 am

http://www.princeton.edu/3D3A/Publications/Pure_Stereo.pdf

Some easy reading material.
Note the requirement for high directivity loudspeakers (think horns) to eliminate the listening room's effect on the sound field.
Also the confirmation that tonal accuracy is being preserved in the processor.

The only disadvantage I can see is that the sweet spot will only accommodate one listener, and that custom measurements must be taken in the listening room.