Smyth Virtual Surround System

[Hopstretch]

I think a few people heard this at CanJam last year? Anyway, it's now shipping.

Obviously the focus is home theater, but I wonder if it can be applied to "simulate" out-of-head stereo listening?

EDIT: I just RTFA more closely, and it can indeed do stereo.

[Dreadhead]

Very cool. Seems pricey at 3k though.

edit: well it includes a Stax system and head tracker so maybe not.

[padam]

I guess it comes with the SRS-2020. Pricey, but it's probably the best, period.

[blubliss]

I heard this at the last SoCal meet and it was unbelievable. Literally can't tell if you are listening to the headphones or the speakers.

I spoke with the rep today and am on the list for a unit. They are somewhat flexible with knowledgeable headphone people to sell the unit without the Stax setup. They just don't want some fool with Ratshack headphones trying to use the system and saying it sounds like crap.

[Dreadhead]

I guess it comes with the SRS-2020. Pricey, but it's probably the best, period.

It's actually pretty straightforward mathematically and I'm not to sure I love their approach to head tracking and the treatment of the delays but it is good. (details can be found in their patent if you wish)

It's essentially linear so distortion will not come through so if you try and model a SET tube amp setup you may be less impressed but it is very cool technology.

[manaox2]

I heard this at the last SoCal meet and it was unbelievable. Literally can't tell if you are listening to the headphones or the speakers.

I spoke with the rep today and am on the list for a unit. They are somewhat flexible with knowledgeable headphone people to sell the unit without the Stax setup. They just don't want some fool with Ratshack headphones trying to use the system and saying it sounds like crap.

Amen, I wouldn't want to pay extra for the headphones. I wouldn't mind knowing what they expect for people with their own STAX, this is definitely interesting. I'd have to try it first and save for a while, but reports are pretty awesome so far.

[Dreadhead]

Also it sort of makes your expensive dac useless because whatever you do the SVS ADC and DAC are being used to output to both the speakers and to the headphones.

[tyrion]

This was the highlight in terms of gear at CanJam '08. It was nothing short of amazing. No money for them at the moment so this will have to wait for the end of the year.

[Augsburger]

This was the highlight in terms of gear at CanJam '08. It was nothing short of amazing. No money for them at the moment so this will have to wait for the end of the year.

X2! One of the last scenes in Saving Private Ryan took on such realism with the bullets and tracers whizzzing all around and the reverb of the tank's treads was chilling. Then there was the live concert of the Eagles (I think), really did sound like you were tenth row center as if I was sitting right in the middle of the audience. I would love to get the DVD of the Trinity Sessions if it existed, and demo the Smyth system again.

[morphsci]

.... I would love to get the DVD of the Trinity Sessions if it existed, and demo the Smyth system again.

There is a DVD for Trinity Revisited but I do not think there is one for the original Trinity Sessions.

[Augsburger]

There is a DVD for Trinity Revisited but I do not think there is one for the original Trinity Sessions.

I haven't been able to find the original Trinity Sessions on DVD yet but you just reminded me of my Sessions at West 54th Street DVD. I bought it after I heard it on a really nice surround system and forgot about it until now. I would like to hear it again on the Smyth System to determine if it is as good as I thought. The DVD's acoustics from what I remember were really good and the fact that it was an "unplugged" recording session made it a must have.

[Dusty Chalk]

I haven't been able to find the original Trinity Sessions on DVD yet...

As was stated before, I'm pretty sure this does not exist.

[HeadphoneAddict]

This is worth keeping my eye on for the future. My friend does home theater installations, so if my Jamo or Velodyne 5.1 isn't good enough, I could take the rig over to one of his setups and set it up on a better system.

[bozebuttons]

I also heard this at CanJam08,Truly impressive.I will own one at some point its that good.

[jp11801]

This was absolutely amazing I finally got to check it out at the LA meet after missing it at CanJam. It was groundbreaking for DVD entertainment. I am not so sure I'd want to do alot with 2 channel audio but movies and music videos are presented at a new level.

They have a pad that the listeners chair sits on that to me really cool. It seemed to provide sub woofer info plus a visceral physical effect. You feel the tanks in Private Ryan it was mind blowing.

[atothex]

I usually get hyped up pretty easily, but I just can't buy into this thing. So everyone says its great... but is it better than $3000 worth of speaker stuff in a decent room? That'd be a clincher for me.

[Fungi]

For a good number of people, the decent room part is unattainable/impractical.

[dvse]

This is indeed a great development for headphone listening, the idea has been round for quite some time but was never really commercialized.

Of course, the physical set up is entirely unnecessary - it can all be done in software once you have a head tracker. What would be really interesting (and dramatically increase market appeal) is to find a way to calibrate the system without an external reference.

[dvse]

Checking out the patent has got me thinking about ways to make this usable as a piece of PC software. The biggest obstacle would be getting personalised HRTF measurements.

One way around it would be to construct some sort of parametrised HRTF from existing measurements (some research groups seem to have datasets available). A simple approach would be to assume that HRTFs lie on a smooth manifold and create a 2D embedding using an algorithm like isomap. Can then let the user to move a couple of sliders to get the best effect for some test sounds. This can be repeated for different virtual source locations.

To those who have some experience with DSP, does the above sound reasonable?

[Dreadhead]

Checking out the patent has got me thinking about ways to make this usable as a piece of PC software. The biggest obstacle would be getting personalised HRTF measurements.

One way around it would be to construct some sort of parametrised HRTF from existing measurements (some research groups seem to have datasets available). A simple approach would be to assume that HRTFs lie on a smooth manifold and create a 2D embedding using an algorithm like isomap. Can then let the user to move a couple of sliders to get the best effect for some test sounds. This can be repeated for different virtual source locations.

To those who have some experience with DSP, does the above sound reasonable?

In short yup. I have other ideas though.

There is nothing particularly earth shattering about this particular box or approach but I do give them props for commercializing after the false start with Yamaha. There is a lot of similar software that is available already, beyer even has a system and DSP box with head tracking etc. What makes their patent is the in-situ calibration not anything else which there is already prior art for.

I emailed back and forth with one of their engineers but when I pointed out some of the prior art he mysteriously stopped replying to email....

I'm thinking of getting some in ear mics and some other things to do some of this myelf since I've already had quite a bit of success with just building mono DSPs that essentially make one driver sound like another. The strength of the in-situ approach is that when you do it that way your error in your absolute measurement is entirely moot because all you care about is the difference.

As I said before thought it won't model distortion.

[blubliss]

Also it sort of makes your expensive dac useless because whatever you do the SVS ADC and DAC are being used to output to both the speakers and to the headphones.

There is an optical out on the unit so i guess you can run it through your dac later. I don't know what running through multiple dacs will do to the sound. I guess i'll find out.

[Dreadhead]

There is an optical out on the unit so i guess you can run it through your dac later. I don't know what running through multiple dacs will do to the sound. I guess i'll find out.

Yup. Well you are at least using the (far less than $3000) ADC in the Smyth box on whatever whatever decoder you had to get it to 8 channels in the first place.

[dvse]

In short yup. I have other ideas though.

There is nothing particularly earth shattering about this particular box or approach but I do give them props for commercializing after the false start with Yamaha. There is a lot of similar software that is available already, beyer even has a system and DSP box with head tracking etc. What makes their patent is the in-situ calibration not anything else which there is already prior art for.

Indeed there is nothing new, it's the most basic way to get the HRTF and has probably been done by early 90s. I had a brief look at more recent literature and people are trying to do stuff like building a 3d model of the ear and then simulating sound propagation (admittedly with with mixed success), interpolating HRTFs based on comparing new listener's ear shape to those previously measured etc.

I emailed back and forth with one of their engineers but when I pointed out some of the prior art he mysteriously stopped replying to email....

I'm thinking of getting some in ear mics and some other things to do some of this myelf since I've already had quite a bit of success with just building mono DSPs that essentially make one driver sound like another. The strength of the in-situ approach is that when you do it that way your error in your absolute measurement is entirely moot because all you care about is the difference.

As I said before thought it won't model distortion.

The need for such calibration with a high quality surround system makes the appeal somewhat limited (unless of course you can get yours done ). This would be great as a VLC plugin though.

[blubliss]

So I finally got around to plugging the Smyth into my reference 2-channel system. I had to get a cheap RCA > XLR box (to make the levels correct) since the Realiser is unbalanced.

Short story is, wow, this is pretty amazing. I am only using my measurement done in a 5.1 studio and the SDS/R10, but I have high hopes when I measure a reference 2.0 room it is gonna be awesome. I think the headphones do impart some sonic signature but I am primarily listening to the room where the measurement was done. Truly amazing technology.

For anyone stuck with headphones, this may be a viable solution for recreating a speaker environment.

For movies and headphones it is unbeatable.

[deepak]

So I finally got around to plugging the Smyth into my reference 2-channel system. I had to get a cheap RCA > XLR box (to make the levels correct) since the Realiser is unbalanced.

Short story is, wow, this is pretty amazing. I am only using my measurement done in a 5.1 studio and the SDS/R10, but I have high hopes when I measure a reference 2.0 room it is gonna be awesome. I think the headphones do impart some sonic signature but I am primarily listening to the room where the measurement was done. Truly amazing technology.

For anyone stuck with headphones, this may be a viable solution for recreating a speaker environment.

For movies and headphones it is unbeatable.

If you have an Omega 2 could you try the Smyth with it? As I understand the small soundstage of the O2 shouldn't matter much with the Smyth setup?