How do we perceive “soundstage” and “imaging”?

[Kal Rubinson]

Nothing kills immersion for me quite like hearing a 3 foot tall singer.

Your gonna need bigger speakers.

 

[Darkweb]

Your gonna need bigger speakers.

My speakers are plenty big. Even most large speakers have tweeters at seated ear height so if they’re projecting vocals at that height that’s a big turn off for me.

 

[Snarfie]

Imaging or for me better depth layers in (stereo) sound is what i aiming at. Basically i want to hear in a live performance for instance as much the actual depth size (as possible) of the room/venue it is played in it creates for me (if recorded well) an intimate atmosphere where the band sound like a band/team.

In a studio recording it is the same but than how the sound engineer an the artist(s) are deciding how to put it on the master tape that version I want to hear not purse what i like to hear. With head phones i can't perceive any serious depth as i can with my floor standing speakers some times more than 3 meters (subjective). Room correction changed my new found sound (depending on how bad your room acoustics is) in my case quite dramatically. Basically after room correction an placing my speakers carefully my speakers aresounding almost optimal. An change for at least 70% for the better.

By the way all best whishes an a awesome sounding 2020.

 

[Thomas_A]

Don’t agree with most of this but I’m too lazy to argue.

However if “rainbow effect” is what gives vocalists a realistic height in my system then I welcome this “error” with open arms. Nothing kills immersion for me quite like hearing a 3 foot tall singer.

interestingly some electronics I’ve tried in my system seem to reduce soundstage height cues but I have no explanation why that would be.

You can also read about it here:

https://live.dirac.com/stereophonic-system-phantom-sources/

 

[thefsb]

I could not find the first one on Qobuz but I did find the other two as well as the Pat Metheny track and I understand what you are saying. These are among the most airless and claustrophobic recordings I have come across. There was some hint of ambiance on the Schlingen-Blängen track but only before the music started. I listened for several minutes and all I heard was the sound of 2 speakers with only a bit of difference between them.

As for the Nomos Alpha with Rohan de Saram, I found a couple versions on YouTube and there was one from 1966 (

), which seems to be the same as the one in your link and it, too, sounded as if there was a blanket over the cello (along with the microphones). However, there was a 2012 recording, also with Rohan de Saram and it sounds entirely different. I'll not say more about it but I suggest you give that one a listen (

) and tell us if you hear something different.

There's a wide range of musical taste among these choices but they share a claustrophobic acoustic.

Let's first cross check that we're listening to the same things.

Qobuz has only 30 seconds of the only recording of Schlingen-Blangen that matters. The others are horrible. Tidal has the whole thing.

Qobuz and Tidal both have Pop Sicle from the album High Rise II but neither have it from Live, which is on Youtube, that I originally had in mind. But the High Rise II version probably serves my purpose better.

The recording of Nomos Alpha I mean is, I believe, from 1966 and the one you linked from Youtube but that doesn't sound nice and is cut off at 8 mins when it's actually 15:30. It's on Qobuz here.

I have listened to the more recent performance you linked at the computer on the near-field monitors but not on the living room system. I won't comment until I've done that.

 

[majingotan]

My speakers are plenty big. Even most large speakers have tweeters at seated ear height so if they’re projecting vocals at that height that’s a big turn off for me.

You have to have the tweeters at ear height since the sound waves amplitude disperse much more quickly at higher frequencies.

 

[Thomas_A]

The focal point where the drivers add in phase is a design property of the speaker. It can e.g. be at tweeter axis or a point between mid and tweeter axis. The increased height of the center phantom image is caused by a different mechanism due to the stereo system, as mentioned before.

 

[Kal Rubinson]

Let's first cross check that we're listening to the same things.
Qobuz has only 30 seconds of the only recording of Schlingen-Blangen that matters. The others are horrible. Tidal has the whole thing.

Ah. I do not have Tidal any more.

Qobuz and Tidal both have Pop Sicle from the album High Rise II but neither have it from Live, which is on Youtube, that I originally had in mind. But the High Rise II version probably serves my purpose better.

The one from High Rise II is what I listened to.

The recording of Nomos Alpha I mean is, I believe, from 1966 and the one you linked from Youtube but that doesn't sound nice and is cut off at 8 mins when it's actually 15:30. It's on Qobuz here.

I will check that one.

I have listened to the more recent performance you linked at the computer on the near-field monitors but not on the living room system. I won't comment until I've done that.

OK.

 

[Kal Rubinson]

My speakers are plenty big. Even most large speakers have tweeters at seated ear height so if they’re projecting vocals at that height that’s a big turn off for me.

It was a joke but I neglected to append the smiley.

 

[Darkweb]

It was a joke but I neglected to append the smiley.

I should’ve known! It’s hard to tell around here because I got the sense we were heading toward an “imaging is snake oil” discussion.

 

[Thomas_A]

Imaging/soundstage in true stereo exists due to the phantom images. You can do tricks to make sound outside the speakers, and you have a rainbow effect with is a "pleasant" error of the stereo system. Besides the quality and type recording (one example is Cantate Domino, linked to previously), the performance of the images goes bad if the direct sound is polluted with reflections within 1 ms or so. As also discussed in the dirac link, you can imagine a stereo setup as being in a lounge with tree walls - the speaker wall is an opening to the event. In practice - mening that the wall should be acoustically invisible. The other model would be to be transferred to the event, which would require multi-channel audio.

 

[thefsb]

Imaging/soundstage in true stereo exists due to the phantom images. You can do tricks to make sound outside the speakers, and you have a rainbow effect with is a "pleasant" error of the stereo system.

I think I have experienced this and it's an effect that I dislike. It sounds weird and is distracting.

As also discussed in the dirac link, you can imagine a stereo setup as being in a lounge with tree walls - the speaker wall is an opening to the event.

I read that and I think it describes the normal stereo effect I am familiar with: position on one dimension left-to-right between the speakers.

 

[thefsb]

I should’ve known! It’s hard to tell around here because I got the sense we were heading toward an “imaging is snake oil” discussion.

If it was me that gave that impression then I apologize. I don't deny the reality of what people describe as imaging from two-channel reproduction. But I cannot relate to it. I'm curious because I suspect that, if I could, there might be interesting questions about how it relates to aesthetics and to fidelity. That curiosity is driving my listening assignments together with Kal.

 

[onion]

I imagine a thought experiment. A stereo microphone records a band with X voices/ instruments located at X different points in 3d space. The stereo mic lt and rt microphone transducers correspond to the listener's lt and rt ears if this were a live rather than recorded performance.

The recording from the left microphone plays out the left speaker. The recording from the right microphone plays out the right speaker. This playback schema does not replicate the original performance as the lt ear of the listener hears the output of both lt and rt speakers; same with the rt ear of the listener.

I'm guessing that this is a key reason why stereo playback does not reproduce the 'imaging' information present in live recordings perfectly. There are various technologies that deliver degrees of crosstalk cancellation (so the left ear doesn't hear the right speaker and vice versa) - that should (and do) improve imaging. But they are also dependent on the sound recording choices made by the producers/ engineers at the time of recording.

 

[digitalfrost]

And that's why I use ambiophonics

https://www.ambiophonics.org/true-to-life-sound-reproduction-using-recurs

 

[Thomas_A]

I think I have experienced this and it's an effect that I dislike. It sounds weird and is distracting.

Well this is preference, but I and I believe most do prefer when the soundstage appears a bit bigger than just being phantom images on a straight line between the speaker focal points. Then there are Q-sounds and such that stretch sound outside of the speakers.

 

[Kal Rubinson]

If it was me that gave that impression then I apologize. I don't deny the reality of what people describe as imaging from two-channel reproduction. But I cannot relate to it. I'm curious because I suspect that, if I could, there might be interesting questions about how it relates to aesthetics and to fidelity. That curiosity is driving my listening assignments together with Kal.

OK. I checked out the Nomos Alpha on the Qobuz link and it is, indeed, much clearer and more satisfying than the one I had heard before. Still, despite the clarity, it was quite close and airless. I could say the same for the file of Schlingen-Blängen that you sent. What characterizes these is that they are single instruments that are closely-miked. I note that the L/R meter levels track each other so closely that I suspect there is little difference between the channels and that implies that there's little stereo effect.
Let me suggest the Cantate Dominus recording that was suggested; it is simply awash with ambiance. For something that might be closer to your taste, I also suggest Magnus Lindberg's "Action - Situation - Signification" (Toimii Ensemble on Finlandia Records ‎– 1576-53372-2). Let me know.

 

[thefsb]

Let me know.

I'll report back tomorrow if I can make time to write it up.

 

[Thomas_A]

A rather good studio recording IMO is "Talking Timbuktu" (Ali Farka Toure and Ry Cooder).

 

[ajawamnet]

And that's why I use ambiophonics

https://www.ambiophonics.org/true-to-life-sound-reproduction-using-recurs

Interesting concept, tho I still find it lacking in representation of a concert hall. And as to on stage - I stood right in the viola's of the Pittsburgh Symphony Orchestra during Aaron Copeland's Rodeo. I was doing sound for one of their outdoor gigs at Point State Park (Westinghouse had built a "symphony stage" - a giant steel structure with crappy wood inside) Wow - you couldn't hear anything except brass and percussion. Amazing what the stage and hall does to the sound...

As to modern pop type music, it's so artificial to begin with - you're simultaneously inside the piano, in the middle of the drum kit, in the middle of the guitar amp, horn, etc... while also being right in front of the mouth of the singer. The artform of pop recording that pioneers like Sir George Martin, Quincy Jones, Roger Nichols and others created is that it IS an artificial aural landscape.

Sound design for film - the genesis of n-Channel systems - also have created an artificial acoustic landscape that fit the 2 dimensional visual presentation.

Game designer - now they have a real difficult task, especially those using some sort of virtual reality visual system. They need to be able to faithfully reproduce a fully 3 dimensional (with time as an integral of the 3 spacial dimensions) sound field.

I think the only way that we will ever get to a method of TRUELY realizing any acoustic environment is through a holographic method of controlling the listening environs' air molecules. I doubt that any conventional multidriver transducer type system will ever work.

Either that, or direct aural nerve stimulation bypassing any need for "transducers" That system would need to account for listener physical movement, via things like IMU feedback to the source streams for each neural interface.

Just the term "transducer" limits the accuracy of the concept. There needs to be a way - besides electomechanical means - to move air in such a way that the entire listening area has air molecules moving as they did in the original source environment. I don't believe that any recording method (again a futile attempt with multiple mechanical-electro single point source diaphragms) nor any current storage method (stereo, n-Channels) with be completely effective.

In order to be a faithful system of reproducing the original environment, the first criteria would be that if a listener moves around in the reproducing area, they will experience the same changes in arrival times/amplitude as one would moving the same way in the recording environment. That in itself is a daunting task - since any actual recording method belies the ability to capture the infinitesimal variance that occurs at each infinitesimally small unit area of the source environ. Then to reproduce that in a listening environ would probably require a direct, non-electromechanical method of controlling the air molecules, with the size of the air molecules being the quantitative limit.

One can dream, eh?


I have to add to this - a recent Bob Lefsetz (the guy Taylor Swift wrote the song "Mean" about) article talking about Jimmy Iovine. Yea, he's great and all, nice to see he's actually trying to play an instrument:
https://www.nytimes.com/2019/12/30/arts/music/jimmy-iovine-pop-decade.html

But his statement:
"I find out a lot through the artists I work with. Dre is a perfectionist of audio,
maybe one of the greatest audio producers that ever existed. And when I found
out what Dre was concerned about, that the equipment his kids were listening
to the music on — an entire generation was learning about audio through
cheap, inefficient equipment. That’s how Beats started."


got this response from me:
"What the fuck, over?

For one - Beats headphones suck - ask anyone on any of the forums like Audio Science Review and Steve Hoffman. Everytime I've seen Beats at Best Buy they're in pieces on the display... Geezz

For two - the statement "...an entire generation was learning about audio through cheap, inefficient equipment."

What the fuck over 2? Cummon - what were the Beatles listened thru? A transistor radio... or an old Magnavox with a stylus that was running at about 5 grams and made for 78 RPM's "

So I guess what I'm saying is:

"Be happy with what you got"