Height in soundstage

Unfortunately the Mana boys were at it again on this thread and I was in the process of deleting the last two posts but due to the unfortunate default behaviour of the post deletion mechanism I accidentally deleted the entire thread :inferno:

Unfortunately - I don't think it's recoverable. Sorry guys

Michael.

 

SteveC - can you repost the link about how humans detect sound elevation? It was probably the most interesting bit of the thread.

Cheers,
Michael.

 

Height cues are a result of outer ear-borne reflections interacting with the standard frequency response of the ear. Such interaction manifests itself as a series of notch filters - with the position of the notch (in frequency terms) being responsible for the shift in upwards or downwards perception.

Several books and websites outline the following in more detail, but the phenomena is replicable with a notch-filtered noise signal through headphones. Many pieces of software will achieve such a signal, and allow variable notching (in the region of 6kHz - 10kHz). Starting at the lower frequency and working upwards, carefully selected notches will produce a sound that appears lower than it's source, moving upwards above the source at the higher end of the frequecy range.

The following (154K pdf) link cover the basics of localisation and HRTF:

www.geocities.com/holaedward/sensationandperception.pdf

 

Groan. It's a geocitiies site, sometimes these bandwidth limit access and "disappear".

Can you post the title/author and any other useful information (like what journal this article appeared in..)

 

Google has it cached in HTML format:

http://216.239.59.104/search?q=cach...rd/sensationandperception.pdf+&hl=en&ie=UTF-8

Looks like it was written by a guy called "Edward Hess" as a Psycology 6014A term paper

Didn't bother reading it but "test tone" has summed up the basics (which were covered nicely in the article that Steve C posted on the original thread).

The point I was trying to make in the original thread is that unless you have things like a human outer ear around a microphone to cause these frequency notches, height information is not even recorded, let alone able to be reproduced by a normal stereo system. Even if you did record it (eg in a binaural recording) you'd have to listen to it with headphones to get the effect.

Michael.

 

it seems to me that soundstage height, as we are dicussing it here where instruments and voices are at different heights, is a property of room / speaker interaction according to the material reproduced and therefore is a wholey artificial phenomenon.
however there is also the phenomenon of an elevated soundstage where the sound is projected above the speakers giving rise to the 12 foot tall gutarist syndrome that i think of when someone uses those immortal words hi-fi.
the last phenomenom involving soundstage height is a dark one and has caused many threads to be lost in the nether regions of the web and so shall not be mentioned here
cheers


julian

 

Surely any kind of soundstage is a 'wholly artificial phenomenon'?

Paul

 

What do you mean? S'pose it depends on your definition of "artificial"...

Dunc

 

OK here you are psychoacoustics: elevation cues

 

We hear sounds coming from places other than the loudspeakers. This is obviously an illusion.

The only mechanism for actually recreating a soundfield at your ears as it might have been at the original event is to use a dummy head binaural technique. The soundstage from a multi-miked recording is obviously an artifact of the mixing. The way one can record something in a room, replay it in another room, and still hear artifacts of the original room is quite a remarkable illusion.

My point is that hearing the illusion of height is no more artificial than the illusion of depth or width. It seems to me that discounting the importance of just one aspect of the illusion is rather arbitrary.

Paul

 

This thread completely confuses me

If there is no sound stage then why doesn't the music coming from the loudspeakers appear as a point source exactly equidistant from both speakers in line with the mid point of the speaker cones, and wouldn't this be called mono?

The idea of stereo is that, for instance, a guitar when playing a note will be mixed more to one channel than the other, and the bass, keyboards etc. Wouldn't this produce some kind of sound stage due to constructive interference of the soundwave produced? I agree with Paul this is artificial, but there should still be one.

 

Apologies for the duff link, Geocities and their ilk make poor hosts.

Height cues come from frequency modification as a result of interaction with the pinna. They can be experienced first hand, and they can be simulated via headphone playback. When suitably recorded, such modifications can be replayed on a standard stereo system. The (correct) assumption, though, would be that the modified information would undergo further modification as a result of impacting the pinna.

Bearing in mind that this impact occurs on the horizontal plane, the modification is not a replica of that which caused the height cue modification in the first instance. The question must, then, be as to the nature of the second modification, and whether height information is lost completely.

 

Hi Lee.

I agree with you.

But what about soundstage HEIGHT? do you like to feel some of it is above and below you, or all on the same vertical plane?


Chris

 

Bringing this thread back to life as I found a recording last night with a fair bit of height in the soundstage.

It's a track by the Waterboys recorded for/by the BBC in their Golders Green studio which is apparently an old converted theatre. Although the original album track (Don't bang the drum) is a hugely powerful number filled with layer upon layer of overdriven electric guitars, this is an acoustic, more atmospheric version.

Anyway, they stuck Mike Scott (vocals, piano) and Steve Wickham (the finest non-classical violinist you're ever likely to hear) on the stage but put Anthony Thistlethwaite (sax) and Roddy Lorimer (trumpet) up in what were the old theatre boxes. Not sure how they miked it up but I would guess it was perhaps just a couple of mikes, letting the acoustics of the venue do the rest.

What you get is a real feeling of the venue (a bit like the Cowboy Junkies' Trinity Sessions) with the piano, vocals and violin very much in the centre stage. However, the piercing trumpet certainly sounds higher up in the soundstage (though not quite as high as it probably sounded in the venue) as does the saxophone.

As you can tell by my enthusiasm, it's a cracking track and the fact that it was recorded live in one take tells you all you need to know about the skills of the musicians involved and the near-telepathic understanding between them.

I must say that I got quite caught up emotionally in the music/lyrics - although I noticed the height in the soundstage, I wasn't listing to the hifi equipment, I was listening to the artists' work and it was great.

Matt.

 

I think the distinction is that soundstage width and depth information, although if you like generated aftificially, is a recreation of the true spatial distribution of the sounds at the recording session (or of the spatial distribution created by the sound engineer in the mixing process). This is possible since the information we need to produce this effect on playback is recorded and stored on the CD/record. Judging from SteveC's link the same is not true for soundstage height, since standard microphones do not have an equivalent to the external ear to create the required data. Thus any apparent soundstage height seems to me to be completely artificial, being purely a result of the speaker/room arrangement. I'm not saying this is definitely right, but the impression I've got from what I've read in this thread.

Might this not just be because the tweeter of your speakers is presumeably higher than the other drivers? Does anyone have an example of a piece of music where a HF sound is clearly very low in the soundstage?

 

Well, the fiddle was playing similar (if not higher) frequencies than the trumpet and yet it sounded level with the speakers (at tweeter level) whereas the trumpet and sax sound a couple of feet higher up.

Actually, I might give that track a listen through headphones to see if they give an impression of height.

Matt.

 

Yeah, that would be interesting.

 

Well, I gave it a go (through my HD600's) and not a great deal of height I'm afraid - perhaps a little but more of a distant sound to the trumpet and sax.

I also couldn't resist a quick blast of Roger Waters' "Sexual Revolution" from "The Pro's and Con's of Hitchhiking" album. There's a short burst of thunder on that track just before Clapton launches into a rather tasty guitar solo.

Now, if you've never heard this through headphones, give it a go and you will be amazed - at a fair volume it genuinely sounds like real thunder outside the house because the sound is completely above you.

So there's total height information when replayed through headphones but replay it through your normal speakers and... nothing, it's just there in the background.

Matt.