Vertical mix placement

[Text_Edifice]

I've been doing a bit of reading lately on spatial placement in a stereo mix and think I have an ok grasp of how to get mix elements to 'sit' front to back and left to right but I haven't quite figured out how place sounds on the vertical plane.

I can hear vertical placement of sounds in mixes (either deliberate or accidental) and from listening to the LEDR tracks on Chesky Records Jazz Sampler Vol 1 it seems that it's possible to recreate the spatical cues we use to identify the relative heiight of sounds in a mix (though speaker design and placement also appear to play a part)

I think it has something to do with the high-frequency component of a sound but it's not as simple as low freq = low vertical placement and high freq = high freq placement. One of the papers I've read on psychoacoustics suggested notch filtering as a way to simulate vertical cues.

Wonder if anyone can offer some insight?

Cheers

[TimS]

What a great question - Stav talk a little about this but it would be interesting to hear more onthe subject..
Good thread Dave...

[SteveL]

Hi Dave,

I'm sorry I can't help you much with vertical mix tips - I listened to this though the other day which has the most astonishing vertical mix placements going on in it and thought it may be of interest if you've not already heard it. You need to listen on headphones to get the effect (holophonics) but the way the sound travels above your head and way down below your ears is amazing. I do recall that notch filtering plays some part in this, although the actual processing behind it escapes me. The link is below:

http://www.unoriginal.co.uk/Audio/Holophonic.mp3

Cheers - Steve

[jkhuri44]

match box around the head is the most insane thing i've ever heard! :)

the question i have is what practical uses of vertical sound placement besides atmospheric sounds, and synth effects would you want??

interesting question, however not sure for which reason i would be placing sounds above the head, etc...


i am still interested in whether this is possible or not though :P

[Text_Edifice]

There's a couple of practical applications – sound for vision / games is an obvious one and a bit of effort has gone into using head related transfer functions for gaming audio (not necessarily for vertical placement).

Even in a 'normal' mix though it's sometimes useful to be able to place elements above / below, say, a vocal. At the moment I achieve this primarily with mic choice and placement but it would be incredibly useful to be able to shift backing vocals or guitars up or down on the vertical plane.

[smash]

Easy.

Just setup your left speaker directly above your right speaker...

Sounds panned left will sound higher.

;-)


Simon "Smash" Ashby
Apologies, but it's early on a Monday morning.

[walding]

I achieved height with vocals through EQ and ambience (Lexicon 480 IR's and the real unit).

What I did is boost 18-20k on the vox anywhere between 2-6 db depending on the voice (using a daking pre/eq) and applied the 480 ambience setting with everything rolled off under 250hz.

It really does sound like the vocals are riding above the rest of the band.

[TimS]

Not AFAIA Angus..

[Futureman]

Didn't Rolands RSS (Roland Sound Space) do this? (Or try to)

[Text_Edifice]

Think there's a basic correspondence between formant regions and height (low freq sounds are perceived as lower / high freq sounds perceived higher) however in my experience vertical placement isn't as simple as (for example) eqing out the low freq component from a vocal to make it 'rise' to the top of the sound-stage. Also we can still localise sounds with significant low freq components as above us – e.g. standing under a bridge when a train / car goes over the top.

Think there's something to do with the interaction between direct / inderect sound, the shape of our ears and the effect of our head / body being in the way.

[Chris H]

I've had a go at oscillating the phase for some interesting spatial effects. A plugin called ambisone can be used for some great spatial effects also.

http://www.kvraudio.com/get/1900.html

http://www.prosoniq.com/

The specific info on all this is the work done on ambiosonics.
http://www.ambisonic.net/

I have done a bunch of recordings using a Calrec soundfeild and recordrd from the B format outputs of the Calrec encoder. I would love to have a Calrec soundfeild in my mic locker.

[wez]

QSound!

[Chris H]

QSound!

This reminded me of the old portable stereo i had in the 70's. It had an "extra wide" switch that made the sound stage of a mix sound like those Qsound examples.

[heathen]

I think it is more to do with how our ears are shaped and directional cues. Sitting if front of anything that make noise such as the TV or laptop, tilt your head backward so your chin is directed at the sound source, then tilt your head forward all the way. Both positions sound totally different right? Now work out whats different (frequency wise) and you can start to get a grasp of where to start.

When my forehead is pointed toward a source it sounds more open in the top end, when my chin is pointed toward it it sounds more congested and less open, also sounds like the mids have a notch or dip, anyway just my thoughts.

[Text_Edifice]

Hey Heath – yes think this is on the right track, as I said the research that I have read suggests notch filtering above (from memory) 2k can help provide these cues. Think there's also something to do with we hear low very freq (not so much with our ears, more with our bones)

Question is how to reproduce this effect reliably in a mix.

Surely this can't be that unusual an idea?

[Chinagraf]

I don't think it's that it's a really new idea, it's just that these sort of things have always failed in the past. They generally depend on the listener wearing headphones to accurately reproduce the spatial information. Once you play it on speakers it all becomes unstable/unpredictable as there is really only one sweet spot and outside of the monitor field you are basically hearing mono anyway.
You can do amazing things provided you can guarantee the end user listening environment/position. THat's were cinema has the advantage. What are NHK working on again... 27.1 or something?