[sander99]

Or, the Realiser can simulate 7.1 (for example) from a measurement with any number of speakers, including just one.

Realiser will simulate the speakers that were not available to measure.

No.
You can measure with one speaker in different positions (or with yourself in different orientations towards the one speaker in a fixed position), but the Realiser will not simulate that speaker in a completely different location where you never measured it. (Except when interpolating between the 5 points for countering head movements, that is a limited range about 30° max between the points. Extrapolation beyond the outer points is not possible.)
[Edit: I think my words could be misinterpreted, so to clarify more: you can measure a complete many-speakers system using one speaker in different positions (or with yourself in different orientations towards the one speaker in a fixed position). But if you only have one (3 or 5 point) measurement (for example only for looking straigt at it, and 30° left, and 30° right, and 30° up and 30° down for the interpolation purposes) of that speaker, the Realiser can not simulate that speaker in positions to your side, or behind you, or above you. (In theory it could use the looking 30° left measurement to create the front right speaker, and similar for the left front, but then you could not have head-tracking anymore.)]

 

[raband]

I think people are overthinking the whole thing a bit

Listen to this guy ^^

I think he's got the right idea

edit: or get an A8 - they were a lot less "confusing"

 

[Sordel]

Is it just me or are none of those examples any good at all? When it's one speaker I sort of hear the sound behind me (though I'm not sure how much that is the visual cue) but the surround sound just sounds like not very detailed stereo. I get a much better surround effect from the stereo version of Gentle Giant's Three Piece Suite.

 

[raband]

It's not meant to be good.

It's meant to let you hear how different each persons HRTF is

Each face on the screen it toggles to - that's the sound with their personal settings/recording

Each one is different, some in a big way.

 

[jgazal]

I think people are overthinking the whole thing a bit

Listen to this guy ^^

I think he's got the right idea

edit: or get an A8 - they were a lot less "confusing"

Second best post of 2017. @castleofargh won though.

 

[jgazal]

Is it just me or are none of those examples any good at all? When it's one speaker I sort of hear the sound behind me (though I'm not sure how much that is the visual cue) but the surround sound just sounds like not very detailed stereo. I get a much better surround effect from the stereo version of Gentle Giant's Three Piece Suite.

Perhaps not the best place to post this, but I can delete if you find it inappropriate.

As I see it, the idea of having just one speaker in the demo is that you don’t have the effect of ITD and ILD. So you can concentrate only in spectral cues from the pinna.

What I would object though is that the pinna filter is not the human reference.

The reference is what the real world sounds, how your brain receive those sounds filtered by the pinna in frequency and time domains and how you grow up learning and memorizing those patterns. Your own pinna changes in slow rate throughout your life, but your brain learns the difference and adjust your memories.

So having different pinna does not mean that we hear differently. It is just that some people measure in “inches” and others in “centimeters”.

That’s why convolving two impulse responses, one direct for the speakers and listening room filtering and another inverse for the headphone coupling, is such an elegant way to tackle the problem, because it solves both time and frequency domains at once and you don’t need to model each pinna effect and coupling. In other words it is how you were not wearing the headphones themselves.

@castleofargh, what do you think?

 

[Chillzone21]

Listen to this guy ^^

I think he's got the right idea

edit: or get an A8 - they were a lot less "confusing"

+1 enjoying the A8 daily. As good as Auro 3D will most likely be, I can still hear helicopters flying over my head with the 7.1 A8. 16 channels with enhance it, but I think I can live with the A8 (with a good PRIR)

Second best post of 2017. @castleofargh won though.

Looking forward to a secondhand A16 on the used market down the road.

 

[Takeanidea]

I am thinking I should have got an A8. I've never had more than a virtual 5.1 setup and I wanted something that would mimic a top stereo system above all. You live and learn

 

[castleofargh]

Perhaps not the best place to post this, but I can delete if you find it inappropriate.

As I see it, the idea of having just one speaker in the demo is that you don’t have the effect of ITD and ILD. So you can concentrate only in spectral cues from the pinna.

What I would object though is that the pinna filter is not the human reference.

The reference is what the real world sounds, how your brain receive those sounds filtered by the pinna in frequency and time domains and how you grow up learning and memorizing those patterns. Your own pinna changes in slow rate throughout your life, but your brain learns the difference and adjust your memories.

So having different pinna does not mean that we hear differently. It is just that some people measure in “inches” and others in “centimeters”.

That’s why convolving two impulse responses, one direct for the speakers and listening room filtering and another inverse for the headphone coupling, is such an elegant way to tackle the problem, because it solves both time and frequency domains at once and you don’t need to model each pinna effect and coupling. In other words it is how you were not wearing the headphones themselves.

@castleofargh, what do you think?

not sure you got the one speaker part right. ^_^ even if it's only one speaker, the demo does provide left and right channel sounds. so there are all the ITD and ILD you need. just recorded from somebody else's ears. I suspect it just feels more or less wrong for the listener depending on how close his head/torso are in shape and size compared to the 3 subjects. and then of course there is the headphone's own signature messing things up some more while we listen to the youtube demo.
maybe only one speaker is easier because it's so wrong that the brain has no problem imagining it behind us(I don't see it, it makes no sense, ok maybe it's behind? something along those lines?^_^). or maybe it's just that the masking is more even with a source behind, having almost no sound reflected by the outer ear and as such none of the specific resonance from the specific shape of the ear? IDK there are always too many possible causes and no mean to test them independently.
to me all subjects feel wrong(angle and/or height) on my headphone, but the variations from one to the other is quite obvious and it's the main purpose of the demonstration.

 

[462429]

We could also look at the shiny bright side of this: how lucky we are that as (future) Realiser owners we don't need to implement all necessary speaker-layouts with real speakers!

I agree, let's think positively. Let's take these four rooms (PRIRs) inside one preset as an additional opportunity rather than a difficulty.
Smyth will probably deliver his factory default PRIR in all four variants. And the web utility will probably be harmonized with this expanded concept.

 

[jgazal]

not sure you got the one speaker part right. ^_^ even if it's only one speaker, the demo does provide left and right channel sounds. so there are all the ITD and ILD you need.

I stand corrected. You are right. Listening with headphones this time I now notice both channels.

What do you think about acoustic crosstalk if there is only one speaker then?

 

[castleofargh]

I stand corrected. You are right. Listening with headphones this time I now notice both channels.

What do you think about acoustic crosstalk if there is only one speaker then?

sorry, I'm not sure what you mean. I'll need more context on this one, because without at least a stereo source, I can't figure out how crosstalk can exist.

 

[jgazal]

sorry, I'm not sure what you mean. I'll need more context on this one, because without at least a stereo source, I can't figure out how crosstalk can exist.

True. If sounds on that video are hard panned only one virtual speaker there is no acoustic crosstalk “recorded”*. Playing back the video with headphones also do not introduce crosstalk and then you can concentrate in HRTF differences (ILD, ITD and spectral cues).

If you play that video through real speakers as I did initially, then you have acoustic crosstalk corruption.

Am I getting this right?

*synthesized in the content itself with BRIR convolution of each person depicted in the video

 

[castleofargh]

True. If sounds on that video are hard panned only one virtual speaker there is no acoustic crosstalk recorded. Playing back the video with headphones also do not introduce crosstalk and then you can concentrate in HRTF differences (ILD, ITD and spectral cues).

If you play that video through real speakers as I did initially, then you have acoustic crosstalk corruption.

Am I getting this right?

well, playing some HRTF cues on speakers totally defeats the purpose. that's playing binaural record on a pair of speakers. but for the principle alone, then yes the left speaker will send the cues intended for the left ear alone, but the right ear will end up getting them too(modified by distance and HRTF). same for right channel. it's like the reverse of the problem we usually have with stereo albums playing on headphones. one stereo method played on the other = some result that is at the very least not accurate in position cues.

 

[sander99]

True. If sounds on that video are hard panned only one virtual speaker there is no acoustic crosstalk recorded. Playing back the video with headphones also do not introduce crosstalk and then you can concentrate in HRTF differences (ILD, ITD and spectral cues).

If you play that video through real speakers as I did initially, then you have acoustic crosstalk corruption.

Am I getting this right?

I assume the audio in the video is a binaural recording of one speaker playing in the room. Recorded (or simulated as if it were) with binaural microphones in the ears of the different persons shown. So no hard panning, just all natural binaural representation of one sound source.
Don't forget: if you listen with headphones to a emulation of for example a stereo signal played over a stereo pair of speakers with the Realiser, you are actually kind of listening to 2 "spatial rendering systems" (I don't know a better word right now) in a row, the first the principle of stereo recordings over 2 speakers (with all the flaws that come with it). The second the emulation of two speakers over headphones, the part that is done quite well by the Realiser. Because the Realiser does it quite well the net result is as if you where only listening to the first rendering part (the principle of stereo recordings over 2 speakers with all the flaws that come with it).
In the video, by using one mono sound source, the first "spatial rendering system" is taken out, one mono source is just one mono source, and you purely can concentrate on how one mono source is "spatially rendered" over headphones. Including hrtf, ILD, ITD, everything that comes with it.
That by the way is why in the demonstrations of the Realiser A16, after the measurement they first let you compare for all the speakers one by one, the real speaker with the virtual speaker. Then you can clearly hear how well the individual speakers are rendered. Also note that when listening to stereo or multichannel with more than one speaker, if the recording and speakers are good, then ideally you should not be able to point out where the speakers are located (not easily anyway), the sound should be "floating in free space". When lstening to that over headphones with the Realiser you ideally should again not be able to easily pinpoint where the speakers are.