PDA

View Full Version : Re: What causes wobble of center voice?


Stig Erik Tangen
September 5th 03, 05:01 PM
Franco Del Principe wrote:
> Is there someone sharing the same experiences? Are there more
> convincing, i.e. experimentally proven explanations for this effect?
> And what is the remdy? Is there at all such a thing like an audio gear
> producing rock-solid imaging? Or is it so much depending on the
> listening room?

From my experience its quite obvious where the problems is; the room.
Its impossible to get a stable image unless you get control over the
reflection pattern within the room. If the lateral reflections are not
perfectly symmetric between left and right, the image *will* shift. Its
most noticable on mono signals (ie. voice in the center) because the
signal from both speakers must be exactly the same to create the
illusion of a dead center between the speakers. Most often, the summed
output (speaker + room) of each speaker are not the same at the
listening position because of non-symmetric reflection pattern.

The cure is quite simple; treat the primary lateral reflections with
absorbtion or diffusion. You will be amazed how much this improves the
imaging! These reflections happen at the side walls in your room, at the
spots where you would be able to see the speakers if you put up a mirror
on the wall. Also concider the rear wall of the room, especially if you
sit close to it. Or... if you're as crazy as myself, you could build a
LEDE room (se reference below).

Recommended reading; The Master Handbook of Acoutics, F.A.Everest, ISBN
0071360972

Cheers,
Stig Erik Tangen

Harry Lavo
September 5th 03, 06:15 PM
"Franco Del Principe" > wrote in message
news:b8U5b.266601$cF.84154@rwcrnsc53...
> Dear members of RAHE,
>
> Here is a question I asked myself over and over again, but I
> never read a convincing answer in the audio press nor in
> this newsgroup:
>
> Why is a singer's voice not always in the center and why is
> it sometimes moving?
>
> I would call myself a dedicated audiophile in the sense that
> I listen consciously to my favorite music. That is, I am
> alone in the very quiet room, the lights are dimmed, my eyes
> closed and I let myself drag away by the music. Of course I
> sit in the sweet spot, my head rests on the chair's back and
> I pay very much attention to every single instrument in the
> 3D space extending in front of me and in particular to the
> lips of the singer... and there is where the trouble starts.
>
> As I understand, there is a general rule in the
> recording/mixing procedure that the singer's voice is
> exactly in the center. Sometimes the frequency band between
> 2 and 3 kHz is boosted to make the singer stand out in front
> of the band. However, I do have CDs where this is not the
> case, i.e. the singer is clearly left or right of the center
> or embedded within the band. But this is not related to my
> question.
>
> What if an exactly centered voice moves during the playback
> within one and the same song? It is not much, but clearly
> perceivable, about 1-2 feet left or right, forward or
> backward, sometimes even up or down.
>
> I encountered some explanations in the audio press like
> - pre-amp, amp, CD player are badly designed
> - the speakers are flawed
> - the recording is flawed
> - the singer is moving in front of the microphone
> - modulating interference between direct and reflected sound
> in my listening room
>
> What puzzles me even more is that when I listen to test CDs
> (with music, not tones or noise) every single test performes
> excellently. That is, left/right, center, front back are
> sharply defined, and stable. (Uhm, yes, it is just 2-channel
> stereo, no surround).
>
> Is there someone sharing the same experiences? Are there
> more convincing, i.e. experimentally proven explanations for
> this effect? And what is the remdy? Is there at all such a
> thing like an audio gear producing rock-solid imaging? Or is
> it so much depending on the listening room?
>
> Thanks for your help.
>
> Cheers,
> Franco
>

Singers are ordinarily "mono-mic'd" just for this purpose...to keep the
voice from "moving". However, if the singers you mention have been recorded
using x-y stereo mic'ng, then you will get the effect you notice. Not
saying this is what you are experiencing, just a possibility.

BEAR
September 6th 03, 05:42 AM
Franco Del Principe wrote:

> Dear members of RAHE,
>
>

<snip>

> Why is a singer's voice not always in the center and why is
> it sometimes moving?

Sometimes it is the recording.

There was a cut on a Dorian Sampler of a duet, male/female with
a quartet backing... the voices seemed to move oddly.

I know the recording engineer, and mastering engineer. So I asked.
The recording was made with stereo microphones. The singer moved
as she sang. You can hear it. :- )

So, it depends upon the recording. MOST pop recordings are made with
MONO (single) microphones for vocals and the mix is a "pan potted"
affair where mono signals are moved left/right electronically. The
"stereo" is synthetic. So the sound of the singer or anything else done
this way will be stable and not move at all.

<snip>

>
> What if an exactly centered voice moves during the playback
> within one and the same song? It is not much, but clearly
> perceivable, about 1-2 feet left or right, forward or
> backward, sometimes even up or down.

Up/down is almost certainly a phase shift related phenomenon where
you are possibly getting reflections off the ceiling at certain frequencies
that cause a shift in the apparent location of a given part of the mix - it
tends to happen with respect to frequency, and is likely to be an artifact
of the polar response of the speakers and your room interacting with that.
(this includes nasty edge refractions)

Up/down is *not* information that is inherent in the recording, nor is it something
that two speakers *can* do under normal circumstances, since there is
nothing above or below physically to generate a sound, and there is
no "data" in the recording that corresponds to this position. Normally,
speakers can produce left/right - front/back. Not up/down.

>
>
> I encountered some explanations in the audio press like
> - pre-amp, amp, CD player are badly designed
> - the speakers are flawed
> - the recording is flawed
> - the singer is moving in front of the microphone
> - modulating interference between direct and reflected sound
> in my listening room

Most likely one of the the latter three... the speakers or electronics
are unlikley to cause the singer's position to move.

The *only* chance that the speakers are causing a *left/right* movement
is if the left and right speakers have enough variation in the frequency
response so that the relative amplitudes at any give frequency between
the left and right speaker are different. Almost like having two completely
different speakers for your left and right... but if you have recordings where
the image is rock solid, then you can completely forget about this possibility
AND the possibility that there is some problem in the signal chain (which
is *extremely* low probability).

_-_-bear

>
>
>

<snip>

>
>
> Cheers,
> Franco

September 6th 03, 10:23 AM
Franco Del Principe > wrote in message news:<b8U5b.266601$cF.84154@rwcrnsc53>...
> What if an exactly centered voice moves during the playback
> within one and the same song? It is not much, but clearly
> perceivable, about 1-2 feet left or right, forward or
> backward, sometimes even up or down.

Imagine an acoustic recording made with only 2 microphones spaced
about 7 inches apart, about 15' back from the foremost performer
(probably the singer). The microphones will capture the phase
difference of the signals to get L/R balance and also will capture
room effects to get some amount of front/rear balance.

This arrangement would capture every move the singer makes, left to
right or front to back. Even if the singer stands still, it might be
audible even if he just moves his head.

Not many recordings are made with this minimalist mic arrangement, but
some are.

Drew Eckhardt
September 7th 03, 03:45 AM
In article <b8U5b.266601$cF.84154@rwcrnsc53>,
Franco Del Principe > wrote:
>What if an exactly centered voice moves during the playback
>within one and the same song?

Asymetric speaker placement w.r.t. the room and differences in the
adsorbtion/diffusion of the side walls both create frequency dependant
left/right image shifts. This was real bad in my office where there's
a book shelf along one wall, the speakers needed to be close to the side walls,
and I ended up using different bass/treble settings on the two channels
to maintain a stable image.

We may have similar vertical effects because carpet/pad adsorbtion decreases
with frequency and the relative phase of reflections when they reach your ears
(resulting in boost/attenuation) is also frequency dependant.

Controlled dispersion speakers (my personal experience is with dipoles and
MTM arrays; although horns/cardoids should work in similar ways) suffer
much less from these effects.

The changes may also be on the recording.

>Is there someone sharing the same experiences? Are there
>more convincing, i.e. experimentally proven explanations for
> this effect? And what is the remdy? Is there at all such a
>thing like an audio gear producing rock-solid imaging? Or is
>it so much depending on the listening room?

Controlled dispersion with adsorbtion at the first points of
reflection is probably a good place to start.

--
<a href="http://www.poohsticks.org/drew/">Home Page</a>
The Congress shall assemble at least once in every Year, and such Meeting
shall be on the first Monday in December, unless they shall by Law appoint a
different Day.

Nousaine
September 8th 03, 04:57 PM
(Drew Eckhardt) wrote:



>In article <b8U5b.266601$cF.84154@rwcrnsc53>,
>Franco Del Principe > wrote:
>>What if an exactly centered voice moves during the playback
>>within one and the same song?
>
>Asymetric speaker placement w.r.t. the room and differences in the
>adsorbtion/diffusion of the side walls both create frequency dependant
>left/right image shifts. This was real bad in my office where there's
>a book shelf along one wall, the speakers needed to be close to the side
>walls,
>and I ended up using different bass/treble settings on the two channels
>to maintain a stable image.
>
>We may have similar vertical effects because carpet/pad adsorbtion decreases
>with frequency and the relative phase of reflections when they reach your
>ears
>(resulting in boost/attenuation) is also frequency dependant.
>
>Controlled dispersion speakers (my personal experience is with dipoles and
>MTM arrays; although horns/cardoids should work in similar ways) suffer
>much less from these effects.
>
>The changes may also be on the recording.
>
>>Is there someone sharing the same experiences? Are there
>>more convincing, i.e. experimentally proven explanations for
>> this effect? And what is the remdy? Is there at all such a
>>thing like an audio gear producing rock-solid imaging? Or is
>>it so much depending on the listening room?
>
>Controlled dispersion with adsorbtion at the first points of
>reflection is probably a good place to start.

IMO the most usual explanation is the lack of a dedicated center channel and a
suitable decoder.

Franco Del Principe
September 10th 03, 05:41 PM
Dear Stig, Stewart, Harry, BEAR, mrclem, Drew, Nousaine,

Thank you all for your input! I really learned something!

It's good to hear that I can exclude my rig *per se* as a
cause of unstable imaging.

Since I am not familiar with recording/mixing procedures the
notion that a singer's movement in front of the microphone
might be the cause was rather surprising to me. I'll keep it
in my mind, though, when comparing different CDs next time.

My strongest suspicion was indeed on the interference of
direct and reflected sound. I actually bought a copy of F.
Alton Everest's "The master handbook of acoustics" a few
weeks ago to learn more about this subject, but I made it
only to the first couple of chapters yet... I will read more
and consider lateral reflection damping. However, this might
be an issue for my wife ;-)

Cheers,
Franco

Stewart Pinkerton
September 10th 03, 09:03 PM
On Wed, 10 Sep 2003 16:41:42 GMT, Franco Del Principe
> wrote:

>My strongest suspicion was indeed on the interference of
>direct and reflected sound. I actually bought a copy of F.
>Alton Everest's "The master handbook of acoustics" a few
>weeks ago to learn more about this subject, but I made it
>only to the first couple of chapters yet... I will read more
>and consider lateral reflection damping. However, this might
>be an issue for my wife ;-)

Try some *really* expensive Turkish or Persian rugs, hung on battens a
couple of inches from the wall. She'll love 'em, and you'll love the
sound......................... :-)
--

Stewart Pinkerton | Music is Art - Audio is Engineering

All Ears
September 12th 03, 08:52 PM
I just read this in a review of the Meadowlark Kestrel. It seems to be
relevant to this thread.

"Vocalists magically appear right in front of you. Strings are never harsh
or bright unless the recording is. Even then it won't drive you from the
room as other, more analytical transducers may. I do not hear any metallic
overlay of subliminal ringing that seems to plague many metal domes like my
old Paradigm Reference Studio 60 v2s. Individual images do not waver, wander
or flicker side to side as they do elsewhere - including the very slight but
still audible image drift on B&W's Nautilus 802. Listen to a solo violin,
perhaps fellow Canuck James Ehnes playing Bach's Sonatas and Partitas
[Analekta FL 2 3147-8]. The Kestrels never once shift Ehnes' violin as
though it were comprised of different segments in different locations. On my
Paradigms, the violin slides back and forth to make following the music
difficult
"Franco Del Principe" "

The full review can be seen at
http://www.6moons.com/audioreviews/meadowlark3/kestrel2.html

KE

> wrote in message
news:avI7b.408088$o%2.185699@sccrnsc02...
> Dear Stig, Stewart, Harry, BEAR, mrclem, Drew, Nousaine,
>
> Thank you all for your input! I really learned something!
>
> It's good to hear that I can exclude my rig *per se* as a
> cause of unstable imaging.
>
> Since I am not familiar with recording/mixing procedures the
> notion that a singer's movement in front of the microphone
> might be the cause was rather surprising to me. I'll keep it
> in my mind, though, when comparing different CDs next time.
>
> My strongest suspicion was indeed on the interference of
> direct and reflected sound. I actually bought a copy of F.
> Alton Everest's "The master handbook of acoustics" a few
> weeks ago to learn more about this subject, but I made it
> only to the first couple of chapters yet... I will read more
> and consider lateral reflection damping. However, this might
> be an issue for my wife ;-)
>
> Cheers,
> Franco
>

All Ears
September 13th 03, 06:44 PM
snip

> So, let's see if we understand this: Mr. Ehnes, when he plays,
> keeps his violing in precisely one spot, never moving it here or
> there, not up nor down, narry left and certainly not to the
> right. It is kept rigidly in place, locked in position, and
> never slides back and forth.
>
> And this reviewer noted this fact.
>
> A couple of questions, if I may:
>
> 1. How does the reviewer KNOW that this is the way Mr. Ehnes
> plays?
>
> 2. Why does Mr. Ehnes NOT play like every other violinist I have
> ever seen, where the violin, at the hands of the supposed
> master in chargem moves about?
>
> 3. How, possibly, could the reviewer EVER follow the music
> played by a violinist who, through inconsideration it would
> seem, moves the violin around, making it difficult to follow
> the music?
>
> 4. How does the reviewer account for the fact that, possibly
> with the exception of the violin played by Mr. Ehnes, an
> apparently extraordinary instrument, the radiation pattern
> of violins changes quite dramatically with frequency as
> different vibrational modes of the violin are excited.
> This, combined with the unpredictability of the recording
> or performance venue, leads to a chaotic "image" at best
> of where the violin "is."
>
> 5. Does this reviewer simply have too much spare time and not
> enough productive work to occupy it?
>
> When I sit down at my harpsichord, "where" the music seems to
> come from shifts all over the place, if a position could be
> identified at all, and not from a fixed point or even in some
> orderly fashion like from left-to-right as one progresses
> chromatically up the keyboard. The apparent source shifts even
> as the note decays as some partials decay at rates different
> than others, phase relationships change in a time variant
> fashion and as I myself move about.
>
> Now consider a review published some years ago in the Absolute
> Sound, as I recall, where the reviewer claimed that some product
> had such precise imaging that the notes played on the lower
> keyboard and upper keyboard could easily be placed in two
> definite planes, one lower and one higher.
>
> Which is ironic, since the strings, the places where they are
> plucked, the places where the cross the SAME bridge on the
> soundboard are the same and, indeed, they are radiated
> acoustically by the same soundboard (multiple keyboard
> harpsichords have but one soundboard).
>
> Actually, it's not ironic, it's foolish and stupid, because the
> reviewer, who apparently had NEVER heard a live harpsichord and
> quite apparently had never seen one, simply assumed that if they
> had two keyboards, the sound MUST be emitted from two different
> places, and then prejudiced his review on that incorrect
> assumption.
>
> I would encourage people who are complaining about why the image
> on their speakers wander, or are imprecise, or who are suffering
> from the dreaded "unfocused soundstage" to GO to a live concert
> and LISTEN WITH THEIR EYES CLOSED and see exactly how difficult
> it is to by sound alone pinpoint the source of the music.
>
> And, if like this reviewer, you have difficulty following the
> music because the violin shifts left and right, then, like this
> reviewer, you made the wrong choice for a hobby, suz real music
> played in the real world by real musicians on real instruments
> don't sound like this reviewers audio Parnassus.

I only submitted this because it seemed relevant to this thread, I did not
do this review, so I am probably the wrong guy to stand up for these claims.

Fortunately it is not a problem that I have, or have noticed, in my systems.
It should however be possible to observe if a problem like this was related
to the recordings, the hardware or room acoustics, by doing some
experiments.

KE

BEAR
September 14th 03, 12:29 AM
Richard D Pierce wrote:

> In article <Btp8b.426353$Ho3.68021@sccrnsc03>,
> All Ears > wrote:
> >"Vocalists magically appear right in front of you. Strings are never harsh
> >or bright unless the recording is. Even then it won't drive you from the
> >room as other, more analytical transducers may. I do not hear any metallic
> >overlay of subliminal ringing that seems to plague many metal domes like my

<snip>

> So, let's see if we understand this: Mr. Ehnes, when he plays,
> keeps his violing in precisely one spot, never moving it here or
> there, not up nor down, narry left and certainly not to the
> right. It is kept rigidly in place, locked in position, and
> never slides back and forth.
>
> And this reviewer noted this fact.
>
> A couple of questions, if I may:
>
> 1. How does the reviewer KNOW that this is the way Mr. Ehnes
> plays?

a. Maybe he read the liner notes?
b. Maybe Mr. Ehnes uses an outboard clamp to hold the instrument rigid?
c. Maybe the recording was made with ONE microphone on the Violin??
d. other

>
>
> 2. Why does Mr. Ehnes NOT play like every other violinist I have
> ever seen, where the violin, at the hands of the supposed
> master in chargem moves about?

a. he does
b. he doesn't
c. it is recorded with one mic.
d. other

>
>
> 3. How, possibly, could the reviewer EVER follow the music
> played by a violinist who, through inconsideration it would
> seem, moves the violin around, making it difficult to follow
> the music?

a. ask the reviewer not the poster
b. it doesn't matter what the reviewer said
c. the reviewer was reviewing the speaker, not the music
d. other

>
>
> 4. How does the reviewer account for the fact that, possibly
> with the exception of the violin played by Mr. Ehnes, an
> apparently extraordinary instrument, the radiation pattern
> of violins changes quite dramatically with frequency as
> different vibrational modes of the violin are excited.
> This, combined with the unpredictability of the recording
> or performance venue, leads to a chaotic "image" at best
> of where the violin "is."

a. it was recorded with one mic.
b. it was recorded with one mic.
c. it was recorded with one mic.
d. other.

>
>
> 5. Does this reviewer simply have too much spare time and not
> enough productive work to occupy it?

a. ask the reviewer
b. don't ask the poster
c. write the publisher or editor
d. perhaps the person who posted this last post fails on the same account?
e. other

>

<snip>

_-_-bear

> --
> | Dick Pierce |
> | Professional Audio Development |
> | 1-781/826-4953 Voice and FAX |
> | |