Jump to content

'Height' as perceived by some: Fact or placebo delusions?


britishcomposers
 Share

Recommended Posts

I have been thinking about this (dangerous thing) but it occurred to me that when mixing an album there are pan controls on each channel.  These have the affect of moving the signal to the left or right for stereo placement in the mix.  Depth comes from increasing and decreasing the volume of a channel (if it is more silent it appears further back in the mix).  There is no "height" control and no way to raise or lower the height of a channel.  Even with a binaural recording, again it is only left or right).   As such "height" is a perception (if we hear a helicopter the brain tells us it is at ear level or up high so we imagine height?).  However, as I said before that perception of overall height is down to how well the loudspeaker projects the sound into the listening area.  So I hark back to what I said earlier .. the first Martin Logan hybrids to me sounded like all the sound was down at floor level.  I never had the chance to raise them on stands to see if that affect went away.  

So I guess logic would say if there is no control to achieve the affect (no pan for up and down) then technically there cannot be height but I guess if you can record something so that most of the sound is coming from the tweeter at the highest point in the loudspeaker and play with the depth you may be able to create a better illusion of height in the recording mix (e.g. the triangle in an orchestra or tubular bells being higher than the violins or such like.   

The other interesting thing of course is where there is a visual aid .. we put up with pretty crap TV sound but the brain is a clever thing when we can actually see the people play it becomes easier to pick out instruments and get an illusion of space and height (e.g. watching a dvd where a helicopter is doing airobatics or other affeccts like in the last Die Hard where they bring a helipcopter down with a car ... 

So on the basis that there is no pan control for up and down - but only for left and right methinks that perhaps by altering frequencies and volumes it may be possible to achieve a perception of height (are there any record producer/engineers who are wammers who could perhaps enlighten us?).  In general though a loudspeaker if it does not have height (my ML example) is audibly noticeable :)  whereas a loudspeaker with height just sounds normal (i.e. it does not draw attention to something missing) ... 

Link to comment
Share on other sites

I don't think I've ever been able to determine height with my stereo system but if you've got a good pair of headphones, binaural recordings can certainly do it. Check out "Dr Chesky's Ultimate Headphone Demo Disc". It's quite an experience.

Link to comment
Share on other sites

2 hours ago, uzzy said:

I have been thinking about this (dangerous thing) but it occurred to me that when mixing an album there are pan controls on each channel.  These have the affect of moving the signal to the left or right for stereo placement in the mix.  Depth comes from increasing and decreasing the volume of a channel (if it is more silent it appears further back in the mix).  There is no "height" control and no way to raise or lower the height of a channel.  Even with a binaural recording, again it is only left or right).   As such "height" is a perception (if we hear a helicopter the brain tells us it is at ear level or up high so we imagine height?).  However, as I said before that perception of overall height is down to how well the loudspeaker projects the sound into the listening area.  So I hark back to what I said earlier .. the first Martin Logan hybrids to me sounded like all the sound was down at floor level.  I never had the chance to raise them on stands to see if that affect went away.  

So I guess logic would say if there is no control to achieve the affect (no pan for up and down) then technically there cannot be height but I guess if you can record something so that most of the sound is coming from the tweeter at the highest point in the loudspeaker and play with the depth you may be able to create a better illusion of height in the recording mix (e.g. the triangle in an orchestra or tubular bells being higher than the violins or such like.   

The other interesting thing of course is where there is a visual aid .. we put up with pretty crap TV sound but the brain is a clever thing when we can actually see the people play it becomes easier to pick out instruments and get an illusion of space and height (e.g. watching a dvd where a helicopter is doing airobatics or other affeccts like in the last Die Hard where they bring a helipcopter down with a car ... 

So on the basis that there is no pan control for up and down - but only for left and right methinks that perhaps by altering frequencies and volumes it may be possible to achieve a perception of height (are there any record producer/engineers who are wammers who could perhaps enlighten us?).  In general though a loudspeaker if it does not have height (my ML example) is audibly noticeable :)  whereas a loudspeaker with height just sounds normal (i.e. it does not draw attention to something missing) ... 

Not only that. If it's just level, then it's just quieter or louder. A true impression of depth comes if the HF is very slightly rolled-off as the volume is reduced.This simulates the increased absorption of higher frequencies by air. I can't remember the actual figures, but there is a known amount by which HF gets attenuated by distance for than low frequencies. That's why the rumble of thunder is audible over very long distances, the crack of lightning over much shorter distances.

Recordings of an orchestra done by a simple cross-pair cardiods, coincident figure-of-eight or spaced omnis will naturally have more distance between the front rang of musicians and the back ranks, so will naturally record the front-back cues as well as the normal Left-Right.

The problem with panning, is that this is normally done with multi-microphones close-miking the instruments, then panned into position. As all instruments are then a similar distance from their own microphone, the front-back cues due to reducing HF are missing. 

. I note the comments made about Ambisonics, but sadly, it's never worked for me, I only ever hear the sound in my head, not externally as it's supposed to. The BBC proms and various demo CDs have all failed to work for me, and yet, I hear height in certain recordings, so I'm assuming it's me. I will be having a play with LEDR later, as I'm intrigued as to how height would work. 

S. 

  • Upvote 1
Link to comment
Share on other sites

4 hours ago, britishcomposers said:

Can you explain this further?  I also wonder about the recorded medium and any expectation placed where re-creating this can be improved or hampered.

The repeated observation of variation in height presentation on different speakers comes around from using familiar music to evaluate different speakers and just as importantly, to confirm correct speaker placement within the room - to focus or enhance the soundstage.

Medium is 'always' CD or digital - test tracks vary from folk to dance.  Listening in the dark and/or with eyes closed where possible.  

I use certain points in certain tracks where I know there is a great variation in sound placement and then position the speakers to improve these and focus them as much as possible (whilst also ensuring a tight centre focus).   I'm really anal about speaker placement so I use multiple methods, from measurments to laser pointers to listening tests.

Anyway, using a track on the same speakers (converted from passive to active linear phase) and placed on the same stands, in the same position.  A sound effect that I know intimately, (which appears to move upward quite fast and then shifts up and down) once played on a linear phase system, the sound shifts upward quite fast as before, but then stays fixed to the spot (rather than then moving up and down).  The only thing that changed was the phase of the frequencies being played as I recreated the passive crossover points and slopes in the active crossover.   Looking at the music (which you can do when it's digital), I could identify the frequencies prominent during these effects, comparing the speaker's phase from passive to active, I could see the passive set-up had a decent but gradual change in phase across a good section of these frequencies - caused by either the passive crossover or the natural phase variation of the drivers.   I therefore came to the conclusion that the removal of the up and down motion corresponded to the removal of the phase shifts in the speaker.

I can't think what else would or could have caused such a subtle difference (in overall impact) but at the same time as being such an obvious change.

Link to comment
Share on other sites

  • Moderator
1 hour ago, SergeAuckland said:

 I will be having a play with LEDR later, as I'm intrigued as to how height would work. 

I believe it is done by taking advantage of people associating HF with height. So a sound can be made tall by accentuating the HF and lower by HF roll off, although I think the bass is accentuated. It is not an easy effect to utilise.

For depth, I believe the following is taken advantage of:

  1. Louder sounds closer
  2. Brighter sounds closer
  3. Less reverb sounds closer

Have fun with the LEDR.

Doug Jones invented the test and there is an AES paper on it but as a non member I can't get a copy without $33.

Edited by George 47
doug jones comment
Link to comment
Share on other sites

Ross, I completely agree with this.

My Open Baffle speakers are about 5' tall with ribbon tweeters on the top baffle, angled to the listening chair and all drivers phase & time aligned. The soundstage has great depth & reasonable width, sometimes beyond the speakers (depending on the original recording of course). Image clarity is pin-sharp vertically between the tweeters & midrange drive units, about 4' up from the floor.

Here's the interesting bit - it's quite easy for those of us with an active system to change the time alignment of any set of drivers & this will blur the imaging of instruments/voice but it can also affect the impression of height. I just did it!

Delaying all drivers except the tweeters by 8ms, then instruments with a lot of high frequency information seem to come from a higher point, confused by the blend of other frequencies produced. Imaging is less precise & fluctuates. Delaying by 15ms, the system sounds uncomfortable.

Therefore it seems logical to me that with less than perfect time/phase alignment, some of the effect we 'hear' is caused by this. The rest may be expectation bias which we 'audiophiles' all seem to possess in abundance!

Link to comment
Share on other sites

14 hours ago, rv295 said:

I'm also 100% convinced that phase affects height placement.

I'm not sure on this. When correcting phase only with FIR (passive or active) the height of performers doesn't change. 

However early reflections could have an impact on imaging which would affect perceived size of sound stage (height included). 

Link to comment
Share on other sites

I'm thinking along more simple lines. More volume, more noticeable ceiling reflection, which with the directivity of higher frequencies will act much like a first reflection and trick the mind as to thinking the sound is half-way between the tweeter and the ceiling.

Link to comment
Share on other sites

4 hours ago, uzzy said:

I have been thinking about this (dangerous thing) but it occurred to me that when mixing an album there are pan controls on each channel.  These have the affect of moving the signal to the left or right for stereo placement in the mix.  Depth comes from increasing and decreasing the volume of a channel (if it is more silent it appears further back in the mix).  There is no "height" control and no way to raise or lower the height of a channel.  Even with a binaural recording, again it is only left or right).   As such "height" is a perception (if we hear a helicopter the brain tells us it is at ear level or up high so we imagine height?).  However, as I said before that perception of overall height is down to how well the loudspeaker projects the sound into the listening area.  So I hark back to what I said earlier .. the first Martin Logan hybrids to me sounded like all the sound was down at floor level.  I never had the chance to raise them on stands to see if that affect went away.  

So I guess logic would say if there is no control to achieve the affect (no pan for up and down) then technically there cannot be height but I guess if you can record something so that most of the sound is coming from the tweeter at the highest point in the loudspeaker and play with the depth you may be able to create a better illusion of height in the recording mix (e.g. the triangle in an orchestra or tubular bells being higher than the violins or such like.   

The other interesting thing of course is where there is a visual aid .. we put up with pretty crap TV sound but the brain is a clever thing when we can actually see the people play it becomes easier to pick out instruments and get an illusion of space and height (e.g. watching a dvd where a helicopter is doing airobatics or other affeccts like in the last Die Hard where they bring a helipcopter down with a car ... 

So on the basis that there is no pan control for up and down - but only for left and right methinks that perhaps by altering frequencies and volumes it may be possible to achieve a perception of height (are there any record producer/engineers who are wammers who could perhaps enlighten us?).  In general though a loudspeaker if it does not have height (my ML example) is audibly noticeable :)  whereas a loudspeaker with height just sounds normal (i.e. it does not draw attention to something missing) ... 

I actually think mixers have come on a bit from the stock 'references' widely accepted and known, and depth is surely re-created by certain phase techniques that can recede a given subject, or conversely project in-front of the plane of a loudspeaker pair, as required by the balance engineer who simply 'dials-it-up' on an editing/mixing suite.  Radio 3's Late Junction can often bring-up some interesting models of music, (often electronic), that demonstrate such facets.  

In fact, many chart records these past 20-years tend to have what I call the 'psychedelic' 360-degree effect in their mixes that lend themselves well to a matrix or Hafler sum & difference set-up, though they also work well with certain loudspeakers:  QUAD ESL's and Tannoy dual-concentrics are very convincing in this regard when sat on-axis.  

The same can be said for hall ambience on classical recordings by using certain microphone specifications and positioning, plus these days, a degree of mixing-manipulation with 'blend' cues that can add further dimension to a given recording where a dryness of balance is evident.  The BBC do this for the Barbican and their own Concert Hall at BH.  Done badly though and one ends up with something akin to that dreadful 1970's/80's Chandos sound of wishy-washy bathroom acoustic where the subject matter gets drowned.  

  • Like 1
Link to comment
Share on other sites

I may be being overly simplistic but I have always assumed that the basic question of height would match that of laterality, such that:

1) To record height information you'd need one microphone positioned above another.

2) To play back height information you'd need one speaker positioned above another.

Since we have neither I've always assumed that perceived variations in height were entirely due to drivers being at different heights and different sound sources covering a different frequency band.

  • Like 1
Link to comment
Share on other sites

13 minutes ago, MartinC said:

I may be being overly simplistic but I have always assumed that the basic question of height would match that of laterality, such that:

1) To record height information you'd need one microphone positioned above another.

2) To play back height information you'd need one speaker positioned above another.

Since we have neither I've always assumed that perceived variations in height were entirely due to drivers being at different heights and different sound sources covering a different frequency band.

That's why the  Soundfield microphone (originally Calrec Soundfield microphone) has four capsules in a tetrahedral arrangement. This results in four signals that can be manipulated to provide L-R stereo, horizontal surround sound with no height element, or if the loudspeakers are tetrahedrally arranged, with height information too.  Unfortunately, to get height out of four loudspeakers means a very large room or a very restricted listening area, as phantom images have to be created between loudspeakers.  The soundfield microphone was also used to good effect to create ambisonic recordings that provide surround over headphones, for those for whom it worked. It was also used to create UHJ matrix surround recordings. It was a very versatile instrument, but didn't catch on much outside of specialist classical recordings, Nimbus being the most popular as far as I recall.

Height with two loudspeakers must always be very variable, dependant on recordings, room acoustics and loudspeaker performance as it's not based on anything reproducible for a mass audience, unlike two-channel stereo which works pretty much for everyone.

S.

  • Like 1
Link to comment
Share on other sites

1 minute ago, SergeAuckland said:

Height with two loudspeakers must always be very variable, dependant on recordings, room acoustics and loudspeaker performance as it's not based on anything reproducible for a mass audience, unlike two-channel stereo which works pretty much for everyone.

FWIW Dolby Atmos ceiling speakers are starting to become more popular amongst the home theatre crowd.

Link to comment
Share on other sites

50 minutes ago, MartinC said:

I may be being overly simplistic but I have always assumed that the basic question of height would match that of laterality, such that:

1) To record height information you'd need one microphone positioned above another.

2) To play back height information you'd need one speaker positioned above another.

Since we have neither I've always assumed that perceived variations in height were entirely due to drivers being at different heights and different sound sources covering a different frequency band.

My rationale is that, hi-fi reproduction aside, one must look to our own hearing faculty to determine what it is that gives us an ability to localise the given height of a subject.  In order to assess this, two people would be required:  one blindfolded and another holding two identical objects that are sources for sound;  one held in each hand and in-front of the listener by about three-feet and central to the blindfolded individual, - but with one item (let's say a wine-glass being 'dinged' by a finger whilst holding the stem) held higher than the other.  Now sound interactions with surrounding reflections, (carpet, plastered ceiling and walls being primary 'fixed' sources for these), will, to a certain extent, quite probably influence the test to a correctly assessed outcome with identifying the two positions.  But I wonder how well this same experiment would be if conducted in an anechoic sound chamber or even outdoors, stood on a lawn and away from building walls?

In short, I think it is our ability to locate sounds from our hunter/gatherer DNA through reflection cues that assist in the identification/positioning process, and when certain phasing models as cleverly applied knowingly or otherwise in sound capture, recreation and reproduction models, this enables a further sense of 'perception' in much the same way;   one that is otherwise completely non-existent in any factual form:  real life or reproduced.   

Edited by britishcomposers
Link to comment
Share on other sites

14 minutes ago, SergeAuckland said:

It was a very versatile instrument, but didn't catch on much outside of specialist classical recordings, Nimbus being the most popular as far as I recall.

Or garage door tests!

  • Upvote 1
Link to comment
Share on other sites

7 minutes ago, britishcomposers said:

My rationale is that, hi-fi reproduction aside, one must look to our own hearing faculty to determine what it is that gives us an ability to localise the given height of a subject.  In order to assess this, two people would be required:  one blindfolded and another holding two identical objects that are sources for sound;  one held in each hand and in-front of the listener by about three-feet and central to the blindfolded individual, - but with one item (let's say a wine-glass being 'dinged' by a finger whilst holding the stem) held higher than the other.  Now sound interactions with surrounding reflections, (carpet, plastered ceiling and walls being primary 'fixed' sources for these), will, to a certain extent, quite probably influence the outcome of a correctly assessed outcome with identifying the two positions.  But I wonder how well this same experiment would be if conducted in an anechoic sound chamber or even outdoors, stood on a lawn and away from building walls?

In short, I think it is our ability to locate sounds from our hunter/gatherer DNA through reflection cues that assist in the identification/positioning process, and when certain phasing models as cleverly applied knowingly or otherwise in sound capture, recreation and reproduction models, this enables a further sense of 'perception' in much the same way;   one that is otherwise completely non-existent in any factual form:  real life or reproduced.   

You've made me realise that one thing I'd overlooked in my simplistic analysis is that we don't have ears on the top and bottom of our heads :D.

Edited by MartinC
  • Upvote 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...