[GeorgeA]

According to your findings, double binauralization compromises audio quality, while conversion from one format to another one is a lossless process, the audio fidelity is unchanged.

 

[SierraMadre]

According to your findings, double binauralization compromises audio quality, while conversion from one format to another one is a lossless process, the audio fidelity is unchanged.

Not exactly. At the end of the day, binaural-whatever’s end format is still 2 channel stereo. The Xbox is not converting the 2 channel binaural PCM in to Dolby Digital 5.1, it’s converting it into Dolby Digital stereo but it’s still binaural just like the original binaural stereo PCM was. Double binauralisation occurs in both scenarios described once the Yarra’s surround mode is engaged but audio quality is only compromised in one.

I’m simply saying that the Yarra’s binaural processing, its surround mode, doesn’t appear to significantly compromise audio quality (if at all) when applied to Dolby’d binaural input rather than original PCM. The difference is simply whether the Yarra’s surround processing is applied to binaural PCM or instead binaural PCM that has been converted to Dolby digital for input to the Yarra.

 

[GeorgeA]

I’ve recently purchased the "Alessandro Quarta plays Astor Piazzolla" blu-ray audio disc. Owing to mShuttle technology, I could download 96 kHz/24bit flac and binaural audio files. By playing these files through my Yarra in surround mode, I couldn’t notice any significant difference with respect to audio quality between the flac and binaural files.

 

[SierraMadre]

I’ve recently purchased the "Alessandro Quarta plays Astor Piazzolla" blu-ray audio disc. Owing to mShuttle technology, I could download 96 kHz/24bit flac and binaural audio files. By playing these files through my Yarra in surround mode, I couldn’t notice any significant difference with respect to audio quality between the flac and binaural files.

Ok but I wasn’t saying that flac vs binaural PCM or whatever made a difference to me. I was saying that the Yarra’s surround mode, when applied to binaural 2 channel PCM, ruined the audio quality for me and that that didn’t happen when applying he Yarra’s surround mode to that same binaural 2 channel PCM if it had been transcoded into Dolby digital stereo beforehand. It’s no surprise that the Yarra’s 9 HRTF models are only selectable when fed with Dolby digital.

Also, I could be wrong about this but I don’t think your Yarra is playing raw flac files per se, I don’t think it has the capability to decode them. It will instead either be flac converted into PCM by your PC / Mac, flac converted into AptX by your DAP or PC if using Bluetooth or, if feeding the Yarra via 3.5mm line-in then just plain analogue stereo.

 

[GeorgeA]

I haven’t written my previous post for the sake of contradicting your findings. As I haven’t got any xbox console or other device that can transcode pcm binaural format into into Dolby digital, I had to reconsider my tests.

I’m aware that HRTF models are only selectable when fed with Dolby digital. However, what may happen when Yarra is fed via usb port and the sounds were felt somewhere around my ears? This sensation was for both flac and binaural files. Of course, flac and binaural files were played by using a software player inside my laptop. I think that the binauralizer works based on some generic HRTF model. When Dolby digital signal is present, one can chose among several HRTFs.

Moreover, based on what I’ve read at Mark Waldrep's blog, I fed my Yarra via 3.5mm line-in thinking that that way the binauralizer was by-passed. The result was that I could also hear the sounds somewhere around my ears similar to Yarra fed via usb port. If the binauralizer had been by-passed, the sounds derived from flac and binaural files would’ve been different. The immersive effect should’ve been manifest only for the binaural files. Is my judgement reasonable?

Anyhow, today I’ve written a message to the Comhear Customer Support asking them to explain the issue of double binauralization.

 

[SierraMadre]

I haven’t written my previous post for the sake of contradicting your findings. As I haven’t got any xbox console or other device that can transcode pcm binaural format into into Dolby digital, I had to reconsider my tests.

I’m aware that HRTF models are only selectable when fed with Dolby digital. However, what may happen when Yarra is fed via usb port and the sounds were felt somewhere around my ears? This sensation was for both flac and binaural files. Of course, flac and binaural files were played by using a software player inside my laptop. I think that the binauralizer works based on some generic HRTF model. When Dolby digital signal is present, one can chose among several HRTFs.

Moreover, based on what I’ve read at Mark Waldrep blog, I fed my Yarra via 3.5mm line-in thinking that that way the binauralizer was by-passed. The result was that I could also hear the sounds somewhere around my ears similar to Yarra fed via usb port. If the binauralizer had been by-passed, the sounds derived from flac and binaural files would’ve been different. The immersive effect should’ve been manifest only for the binaural files. Is my judgement reasonable?

Anyhow, today I’ve written a message to the Comhear Customer Support asking them to explain the issue of double binauralization.

I didn’t say I couldn’t get 3D sound without Dolby, just that I couldn’t get pleasant 3D sound.

The way to bypass Yarra’s layer of binauralisation is simple (and it has nothing to do with inputs connections): don’t activate surround mode and keep it set to stereo. Surround mode on the Yarra (whatever content it is applied to, PCM, Dolby Digital etc.) is a form of binauralisation. The Yarra cannot extrapolate binaurally recorded or binaurally mixed / pre-binauralised content with correct 3D positioning unless the surround mode is activated. You can test this by running the binaural virtual barber shop demo on YouTube. Compare positioning of cues in stereo vs surround mode. Cues will be correctly positioned for surround mode but will not be for stereo. Double binauralisation appears to have always been the intent. At the end of the day, we’re talking about a front firing driver array several feet in front of us. I’m not sure how/why anybody expected content binauralised for headphones strapped directly to one’s head to render accurately in full 3D on the Yarra in stereo mode without additional processing.

 

[Erik Garci]

I’m not sure how/why anybody expected content binauralised for headphones strapped directly to one’s head to render accurately in full 3D on the Yarra in stereo mode without additional processing.

It was claimed that the Yarra's beamforming would be equivalent to 30 dB of crosstalk cancellation, but when I use my SPL meter to measure the ILD, it only gets up to 10 dB in Surround mode. If it was 20 dB or more, it would work much better for binaural content.

As far as Stereo mode is concerned, I think it doesn't use beamforming at all, thus it provides no benefit for binaural content.

 

[Stefanthebass]

I think we have to sort out things. Regarding the 3D surround sound I see two different aspects:
1) multi channel encoding
2) sound delivery

1) multi channel encoding: we all know, the Yarra is capable to deliver 2 channels only. Stereo, binaural, transaural, it always describes a 2 channel signal that somehow includes 3d positioning information. So you could say that this is the native format of the Yarra. Everything else has to be converted to transaural 2.0 using the hrtf. This transformation is the first function of the Yarra. It can be separated from:
2) sound delivery: Yarra offers two modes: stereo and surround. In my understanding stereo is just that. Playing a 2 channel audio like any other speaker. No sound beams, no advanced processing. Surround is were the magic happens with two channels delivered right to your ears with high separation between the channels, with low crosstalk. This works quite good in my opinion.

So if binaural 2.0 recordings sound good for you, then 2) is ok. If 5.1 doesn't, then 1) is not good. If it is the implementation of the conversion or just the wrong hrtf, it can be both.

So try stereo/binaural material using surround mode first, setup your equipment for that and of that sounds ok, then go on for step 2, which is the multichannel function. If stereo is good and multichannel not, maybe try a transcoder.

Ok, there is one more thing regarding the stereo. As they like to control crosstalk (Peter says something in the demo video) they perform a trick that I don't really like. They take a bit of the left channel signal (around 20%) and subtract it from the right and vice versa. They do this, to get a more clearly separated signal, but in fact it has a negative effect depending on your distance from the speaker. It leads to over emphasised side channels (instruments on the fast left and right are louder) and some reverb effects are boosted, too.

 

[Erik Garci]

Ok, there is one more thing regarding the stereo. As they like to control crosstalk (Peter says something in the demo video) they perform a trick that I don't really like. They take a bit of the left channel signal (around 20%) and subtract it from the right and vice versa. They do this, to get a more clearly separated signal, but in fact it has a negative effect depending on your distance from the speaker. It leads to over emphasised side channels (instruments on the fast left and right are louder) and some reverb effects are boosted, too.

There are two ways to look at this effect. One way is that it emphasizes side audio. Another way is that it is diminishes middle audio, especially voices which tend to be in the middle. Either way, the result is the same, because middle and side are relative to each other.

I would prefer this effect to be configurable by the user, as opposed to being fixed at 20% or whatever. I have another speaker system that offers such a setting, which is called Voice, and it can kept at 0% (no effect) or incrementally adjusted all the way to one extreme or the other.

 

[Stefanthebass]

I would prefer this effect to be configurable by the user, as opposed to being fixed at 20% or whatever.

That's exactly, what I wrote them at least two times. It would be great if some of you could also contact them!

You are absolutely right about losing the center channel when the sides are made louder! If you for example listen to "The Violence" from "Rise Against" it's hard to understand the vocals using the Yarra. Notheless it sounds great

Maybe not so well known is the effect of reverb, reverb coming from digital reverb processors, not natural reverb:
in fact there are a lot of reverb processors used in music production, that are NOT mono compatible. Listen to "9 Million Bicycles" from Katie Melua. The vocals are produced using such an effect. Listen to that using the Yarra and compare the mix to "normal" loudspeakers. Much more reverb!

What do these reverb processor do? Well, they calculate a mono reverb signal and add it to the right and the left channel, but with inverted polarity (left=positive (1) and right=negative (-1)). If you mono the music the reverb is gone. With the Yarra the reverb gets much stronger!
Here are some formulas to explain it:
Original mix, reverb: left: 1, right: -1
Mono: left = right = left+right, so: right = left = 1+-(1) = 0
Yarra: left = left-0.2*right = 1-0.2*(-1)= 1.2, right = -1.2

For a signal in the center (left=right=1):
Yarra: left = left-0.2*right = 1-0.2*(1)= 0.8, right = 0.8

So you end up with the side at 1.2 and the center at 0.8! That's almost 40% or 3.5dB more reverb!

As I wrote in some other forum, using the "Eqalizer APO" for Windows I am able to control the effect as the Equalizer offers the possibility to "Copy channels". I have tested some configurations and the ones that add between 15% and 20% of left into right and vice versa sound best.
If the Yarra gets these signals and it processes the channels the signal going to the speakers is almost the original left and right channel.
To my ears it sounds much better and more natural. But unfortunately it works only with my Windows PC and the USB connection.
Again, some formulas proving the effect, for example using 20% in APO:
APO EQ: left = left+0.2*right = 0.8, right = -0.8
Yarra: left = left-0.2*right, using the EQ signal: left = 0.8-0.2*(-0.8)= 0.96, right = -0.96, so approx. 1 and -1, the original signals!

I am sure that a bit of the effect used by Yarra is ok to improve the channel separation and to control the crosstalk. I also believe that the optimal amount depends of the listeners distance from the spreaker.
And at least for the computer speaker use case with a very short distance the currect amount is too high.

I am not sure at all, if that algorithm is used for stereo sources only. I also have not tested it on "far field" setups. If it is part of the beamforming (if I were the designer I would put it here), then it could explain the problem of the low volume center channel for multi channel.setups.
But anyway, I would also think that a tool to set the individual volumes of the channels for a .multi-channel configuration would also be helpful.

 

[AppleheadMay]

I have been using my YARRA 3DX with almost no problems for over two weeks. Yesterday, the sound would cut out for fractions of a second several times per minute. I had made no recent changes to any settings. I tried a number of things that did not work. A full re-install of the firmware fixed the problem.

How does one do a full re-install of the firmware?
I have initiated a firmware update on a Mac and it has been running for over an hour now.
So I think it's safe to say it's stuck. Pull out the plug anyway?

 

[ajax1]

I used the Comhear Windows Yarra 3DX Control Panel for installation.
It offers a QUICK and a FULL install.

Looking at the screen you posted, I believe the Mac updates do a Full install. The QUICK install is very fast. The FULL install is about 12 minutes and is multi-step.

To do a full re-install, I just did the FULL install again.

 

[AppleheadMay]

I used the Comhear Windows Yarra 3DX Control Panel for installation.
It offers a QUICK and a FULL install.

Looking at the screen you posted, I believe the Mac updates do a Full install. The QUICK install is very fast. The FULL install is about 12 minutes and is multi-step.

To do a full re-install, I just did the FULL install again.

Thanks!
And thank Bill for Windows.
I just pulled it off the Mac anyway, hooked it up to my PC, got a FW version number of all -------, tried Quick Install, got an error.
Tried a full install and here it is, brand and shiny new!

 

[SierraMadre]

Ok, there is one more thing regarding the stereo. As they like to control crosstalk (Peter says something in the demo video) they perform a trick that I don't really like. They take a bit of the left channel signal (around 20%) and subtract it from the right and vice versa. They do this, to get a more clearly separated signal, but in fact it has a negative effect depending on your distance from the speaker. It leads to over emphasised side channels (instruments on the fast left and right are louder) and some reverb effects are boosted, too.

That’s how pretty much any VSS I’ve tried works though. If you get some headphones and try a different VSS solution and then compare the sound from one cup only to the other only, it’ll almost certainly do the same thing you describe.

 

[SierraMadre]

I haven’t written my previous post for the sake of contradicting your findings. As I haven’t got any xbox console or other device that can transcode pcm binaural format into into Dolby digital, I had to reconsider my tests.

I’m aware that HRTF models are only selectable when fed with Dolby digital. However, what may happen when Yarra is fed via usb port and the sounds were felt somewhere around my ears? This sensation was for both flac and binaural files. Of course, flac and binaural files were played by using a software player inside my laptop. I think that the binauralizer works based on some generic HRTF model. When Dolby digital signal is present, one can chose among several HRTFs.

Moreover, based on what I’ve read at Mark Waldrep's blog, I fed my Yarra via 3.5mm line-in thinking that that way the binauralizer was by-passed. The result was that I could also hear the sounds somewhere around my ears similar to Yarra fed via usb port. If the binauralizer had been by-passed, the sounds derived from flac and binaural files would’ve been different. The immersive effect should’ve been manifest only for the binaural files. Is my judgement reasonable?

Anyhow, today I’ve written a message to the Comhear Customer Support asking them to explain the issue of double binauralization.

Hi.

Saw your recent posts on the Kickstarter page forum. Couldn't respond there as I'm not a KS backer rather Indiegogo.

Just a correction re Dolby Atmos for Headphone via PC / Xbox to Yarra. The connection from PC/Xbox to the Yarra in this context would not need to be done via 3.5mm headphone out. Indeed, that would be sub-optimal when optical or HDMI are available as connection methods and Atmos for Headphone can be applied to those connections for output to the Yarra.