[jgazal]

I understand what you meant. It's just a terminology issue. HRTF is the first process that converts the multichannel source to binauralized signals (which are not yet sound in the air). HRTF can be performed by the Realiser or by the Yarra. Beamforming is the subsequent process in the Yarra that converts the binauralized signals to 12 separate signals (each one specific to each speaker driver) and thus projects sound to each ear. They are two separate processes, although the marketing literature conflates them sometimes.

Just to add to your excellent post: if you have stereo content you may choose only beaforming/crosstalk cancellation rather than two virtual speakers over a generic HRTF (BRIR), depending how that two channel content was recorded and/or mixed. With multichannel content, mainly with height channels, I believe you are going to need a PRIR to get the full performance.

 

[sander99]

I understand what you meant. It's just a terminology issue. HRTF is the first process that converts the multichannel source to binauralized signals (which are not yet sound in the air). HRTF can be performed by the Realiser or by the Yarra. Beamforming is the subsequent process in the Yarra that converts the binauralized signals to 12 separate signals (each one specific to each speaker driver) and thus projects sound to each ear. They are two separate processes, although the marketing literature conflates them sometimes.

Yes, I see I was a bit sloppy in my formulation: I should have said "passes on a certain channel to the next stage in the device, responsible for creating the beams, more specifically to be beamed to one certain ear unchanged (without actually applying any filtering normally associated with HRTFs)".
It's just that it is difficult to incorporate all the details without making it a page-long story.

 

[Erik Garci]

Just to add to your excellent post: if you have stereo content you may choose only beaforming/crosstalk cancellation rather than two virtual speakers over a generic HRTF (BRIR), depending how that two channel content was recorded and/or mixed.

The Yarra itself does not let you choose, at least not in its current design. That is, when the Yarra receives a 2-channel source, it only does beamforming, without applying an HRTF first, which is great if that is what you want. However, if you have a 2-channel source that was not already binauralized, and you want to binauralize it, you would use the Realiser or a similar device to binauralize it, and then feed the 2-channel binauralized output to the Yarra. You might even want to upmix it from 2-channel to multichannel before binauralizing it.

Choices:
1. beamforming only
2. binauralizing then beamforming
3. upmixing then binauralizing then beamforming

With multichannel content, mainly with height channels, I believe you are going to need a PRIR to get the full performance.

I'm planning to try "Dolby Atmos for Headphones" on the Xbox One with the Yarra. I have already tried it with my crosstalk-free PRIR over headphones, which gives me an idea of how it will sound with the Yarra. It seems that the current version of "Dolby Atmos for Headphones" does not fully support height channels, but I suppose that it will eventually.

Dolby Atmos material ==> Dolby Atmos for Headphones ==> Realiser crosstalk-free PRIR ==> headphones

Dolby Atmos material ==> Dolby Atmos for Headphones (binauralizing) ==> Yarra 3DX (beamforming)

 

[VandyMan]

Given that the the Realiser provides a very convincing multi-channel experience with only two channels, can you describe what you believe the Yarra would add? The conversion to binaural?

 

[sander99]

Given that the the Realiser provides a very convincing multi-channel experience with only two channels, can you describe what you believe the Yarra would add? The conversion to binaural?

If you talk about use in combination with the Realiser: no, the Realiser does the conversion to binaural, the Yarra only beams the two "headphone-channels" to your ears effectively replacing the headphones. So what does it add? Having a similar experience as using the realiser with headphones, but now without using headphones. (And what would be the benefit of this above listening to the real speakers? -have only one small speaker: the Yarra itself -create less noise to the environment -having the acoustics of the simulated speaker/room combination)

 

[VandyMan]

Thanks for the explanation. Sounds interesting.

 

[Erik Garci]

So what does it add? Having a similar experience as using the realiser with headphones, but now without using headphones.

Also, the Yarra reduces the need for head tracking because the front-center sounds will still seem to come from the location of the Yarra when you turn your head.

 

[sander99]

I got a reply from Mike Smyth to my "workaround" proposal. He thinks it is an excellent idea but it can't be done now and the reason why can be a bit of a shock:

Hi Sander,
The idea is excellent, but unfortunately it isn't possible with the A16 at this moment.
The problem is the current HDMI interface board from MDS that we are using - it doesn't allow more than 2 channels to be output. The previous version of the board allowed 8 channels to be output, and the next version of the board (using a different HDMI transceiver) will also allow 8 channels to be output - but the version we are using only allows 2 channels.
But it's certainly something that we should consider in the future - for this type of application.
Very best regards,
Mike
​

So out of the window are my plans to use one of my current 7.1 av receivers (without analog 7.1 input, only HDMI LPCM 7.1 input) to amplify 7 of the channels for use with real speakers.
So I asked more specific questions about this:

Hello Mike,


Oh, I guess this also means that I can not use my legacy 7.1 receiver without analog 7.1 input but with 7.1 LPCM input over HDMI to amplify 7 of the output channels in AV Amp mode? (And 7 more channels with my older receiver that has 7.1 analog input).
But it is not a disaster, I can buy another dated used 7.1 amp with analog 7.1 input.
I assume the AV Amp mode stil wil be implemented (using the analog outputs)?
And in the HDMI pass-through mode? If LPCM 7.1 was coming in to the Realiser HDMI input would it be passed through? And all bitstream formats?


Regards,
Sander
​

And his reply was:

Hello Sander,
The limitation I described is when the audio bitstream is being decoded and the video bitstream is being looped thru the unit. Essentially you can't inject more than two channels of 'generated' audio into the looped-thru video. It's a physical issue in the design of the HDMI board - MDS switched to using Silicon Image devices in order to accommodate HDR but the downside was this audio restriction in the output channels.

However in stand-by mode, where you want to by-pass the A16, then the audio and video bitstreams (including 7.1 pcm) should both loop-thru the A16 without any alteration. In this case you could use your current AV to receive this signal, and decode it as normal. This is a sort of 'active' standby (or pass-thru) since the HDMI board is powered up.

Yes, the analogue outputs will provide a standard AV mode of operation.

Hope this helps.

Mike
​

 

[jgazal]

Also, the Yarra reduces the need for head tracking because the front-center sounds will still seem to come from the location of the Yarra when you turn your head.

Wisely chosen word. Perhaps using the headtracking without the headphone may improve height and back channels localisation. I really don't know. Please review as soon as you test it!

 

[castleofargh]

while I completely get why the A16 can and should work fine, limited only by the various transducers and ambient noise in the room when calibrating, I'm very skeptical about the Yarra. in some ways it's easier, in other ways ....
from how I imagine it working, it must be a comb effect festival to be able to cancel out effectively some signal not destined to a given position. or the tweeters really are super focused and if so what happens when we move on our chair a little?
don't get me wrong, I'm sure a perceived 3D can work, that's not what I doubt here. but beyond that I'm still in my default skeptic mod about any concept of fidelity, consistency, or practicality. on the other hand I'd love to be dead wrong, because then it would be a pretty cool toy to have on a desk.

 

[Sordel]

while I completely get why the A16 can and should work fine, limited only by the various transducers and ambient noise in the room when calibrating, I'm very skeptical about the Yarra.

I looked at the Yarra Kickstarter off the back of this thread and it seems to me to be trying too much. Every manufacturer of a soundbar claims to be able to give convincing surround, and none of them do so. I'd buy one if they work, but this time I'll be waiting for a proven product before ordering.

 

[sander99]

what happens when we move on our chair a little?

I have similar thoughts. When you move your head one head-width one ear will be in the wrong beam, and the other at best hears (some of) the same beam.
I just considered backing the Yarra project because
-now it is (near) half price
-I already invested in the A16 so it is a relatively small extra investment
-with the A16 you get the most out of the Yarra, without a Realiser I don't know

I looked at the Yarra Kickstarter off the back of this thread and it seems to me to be trying too much. Every manufacturer of a soundbar claims to be able to give convincing surround, and none of them do so. I'd buy one if they work, but this time I'll be waiting for a proven product before ordering.

But I think the Yarra is truly something else than all those other sound bars.

 

[Sordel]

But I think the Yarra is truly something else than all those other sound bars.

Even at $1000, which is more than the MSP that they're currently indicating and three times the price for which they're offering this on Kickstarter, this would be suspiciously cheap for a soundbar with a subwoofer and a technological advance of the scale that they're promising. I'm concerned about the nearfield use for gaming and the lights as well: the lights are a gimmick and the near field makes me worry that the surround effect doesn't survive any distance. I'm also suspicious that the drivers seem too small, so some of the non-directional reinforcement from the sub probably falls into the audible range and messes up the effect. Then there's the fact that the only decoding format listed is Dolby Digital, which is lossy. They may be decoding other formats once certified but now we're asking quite a lot for it to be a good decoder on top of all its other tricks. It also bothers me that their video testimonials have so many cuts in them, like they had to remove a lot of qualifications to create the impression of very positive listeners.

If a product sounds too good to be true ... it usually is.

 

[sander99]

the lights are a gimmick

Indeed, and I sure hope they can be switched off

I'm also suspicious that the drivers seem too small, so some of the non-directional reinforcement from the sub probably falls into the audible range and messes up the effect.

Because of the beaming, a lot less power is needed to achieve a certain sound pressure level at the ear. The rest of the room doesn't have to be filled with sound. The crossover point is at 150 Hz (they said at the first "kickstarter live event)". (But it could be a shallow sloped crossover, indeed far into the "localisable range".) With respect to sound pressure in the low range, I guess the idea is that you put the sub very close to the listener(s).

In general: their goal was to see what they can do in the affordable price range. The fact that beaming is energy efficient and requires less potent drivers is a very fortunate aspect in this context.

But you should not think about the Yarra as a normal soundbar or traditional speaker system in terms of its usage. With a traditional system everyone in the room can listen, also walk around the room etc. (ok, not hearing a proper soundstage everywhere but also not hearing nothing [Edit: actually only the sub] or very strange things)(in the "normal stereo" mode of the Yarra your concerns of the small drivers probably are right). Enjoying the Yarra's qualities (assuming that it works) will require some dedication of the listener(s) (not moving around, also not turn your head too much - if you turn 90 degrees of course at least one of the beams can never reach the proper ear). That is part of "the price you pay" for a "too good to be true" experience (again assuming that it works).

 

[Erik Garci]

And of course it would have been nicer if the Yarra could simply use 3 different inputs for the 3 users.

They showed a demo of this towards the end of the live Q&A yesterday. The 3 sources and beams were mono, and it was on a lab system, but maybe it would also work for stereo/binaural.