Are there voices that are more difficult for a speaker to render faithfully?

[weaver]

Two of my favourite classical voices are those of Bryn Terfel and Jessye Norman. Over the years I have heard a number of their recordings but have generally been disappointed - as compared to what they sound like live.

Clearly there are a number of factors involved: a live performance is not just down to the sound being made and expectation must play a large part in the potential for disappointment in any given recording. That said there are other artists whose recordings I can listen to happily and not feel that too much has been lost in the process; Emma Kirkby is one who comes to mind.

So, are there voices, or indeed voice types that in general are better served, that loose less of their original character, that are perhaps simpler in their make up and are thereby less prone to the compromises made by speaker designers? As both of these singers are highly successful recording artists I can only assume that there is nothing inherently wrong with their discs, that it is purely a subjective response to what I hear.

Secondly, given that voice is so central to the development of a Harbeth speaker (and to the appraisal of any speaker) is their a distinction to be made between reproduced voices that sound natural, ie. free from the artefacts that you referred to recently

The real-live human voice doesn't 'do' spitty, wiry, gritty, grainy, pinched, peaky, barking, biting etc.

and voices that not only sound natural but also faithful to the original. Strictly speaking unless we have the original to compare to, judgement of this faithfulness is dependent on memory (flawed as it is) but we do retain a strong subjective impression of what people 'sound like' and this is obviously what I am comparing to in this instance.

Of course it may well be that the speakers that I have been using were not suited to what I wanted from them!

[yeecn]

There is the factor of recording venue, mixing and recording techniques/instruments to consider as well. I have a few CD of Pavarotti - recorded in various occasions - and they do sound quite different!

[A.S.]

Over the years I have heard a number of their recordings but have generally been disappointed - as compared to what they sound like live...

When you say live, what exactly do you mean?

[weaver]

Live live.

As in standing six feet away from Ms Norman in the front row of the Proms Arena, or sitting in the very back row of the Royal Opera House to hear Bryn sing in Die Walküre. Very different presentations and indeed sensations, but nonetheless live.

Is that what you were asking?

[A.S.]

Yes. So you're reasonably sure that Ms Norman's live voice is clean then. Many singing voices have odd qualities: Ella Fitzgerald's voice has a slightly reedy quality emphasised when she was rather close to the mic and/or the mic itself didn't suit her voice.

But listening at home is at the far end of a long recording chain. Each stage of the chain, from the microphone to the loudspeakers, adds (or removes) something. Considering the extreme loudness a professional singer is capable of generating, just one of innumerable variables is how much distortion the microphone generates and the subjective and objective characteristics of the microphone.

Just searching for microphone distortion .... interesting

[STHLS5]

That’s interesting because all recorded vocals are not [/[S]natural[/S] real to me. Most of the time, I listen to music facing the opposite direction while doing work at my computer but SHL5 one a few occasions did make me turn my head to see if “I heard someone’s voice in the room”*. On one occasion, it was rather spooky when I thought I heard my daughter behind me. I have to go out of the room looking for my daughter to make sure it wasn’t her. After that, I replayed the track but somehow I did not get the same feeling. That’s the closest to [s]natural[/s] feel like real person voice without a mic I can get. My previous speakers were nowhere close to that.

* 90% of the type of music I listen to are soundtracks, i.e. songs conceptualized, recorded and mastered to be played in cinemas**. The irony is, all singers must undergo “Mic Test” to see if their voice is suitable for the studio’s recording mic before they are selected for the song recording!

**It is common for these recordings to contain other sounds, vocals besides the singers’, musical instruments and sound effects to convey the essence of the song and the story line of the movie. So don’t expect your Miles Davis, Patricia Barber or Diana Krall recordings to have someone’s presence in the room ;)

ST

[STHLS5]

I think I used the wrong choice of word. I am reposting hoping it makes some sense.


That’s interesting because all recorded vocals are not real (as in live) to me. Most of the time, I listen to music facing the opposite direction while doing work at my computer but SHL5 one a few occasions did make me turn my head to see if “I heard someone’s voice in the room”*. On one occasion, it was rather spooky when I thought I heard my daughter behind me. I have to go out of the room looking for my daughter to make sure it wasn’t her. After that, I replayed the track but somehow I did not get the same feeling. That’s the closest to feel like real person voice without a mic I can get. My previous speakers were nowhere close to that.

* 90% of the type of music I listen to are soundtracks, i.e. songs conceptualized, recorded and mastered to be played in cinemas**. The irony is, all singers must undergo “Mic Test” to see if their voice is suitable for the studio’s recording mic before they are selected for the song recording!

**It is common for these recordings to contain other sounds, vocals besides the singers’, musical instruments and sound effects to convey the essence of the song and the story line of the movie. So don’t expect your Miles Davis, Patricia Barber or Diana Krall recordings to have someone’s presence in the room ;)

[Thanos]

A very difficult yet absolutely stunning recording, indeed, although I don't know the exact conditions under which it was done, is STEPHAN MICUS - ATHOS - A JOURNEY TO THE HOLY MOUNTAIN. The voices of the chorus are incredible... Who can find it, please give it a listen. The SHL5s surpassed my expectations with it...

[weaver]

So the answer to the question:

Are there voices that are more difficult to record and reproduce faithfully? would appear to be yes.

As an end user though, the only bit of the long chain of innumerable variables that I have a degree of control over is indeed the bit at the very end. Which is why I asked:

Are there voices that are more difficult for a speaker to render faithfully? But maybe that's not relevant.

How good should I reasonably expect the sound to be though? If it doesn't live up to my hopes and expectations then what should I do?

Should I blame the recording; or should I start on the hamster wheel of chasing that elusive quality that I feel should be there?

Previously I have used live acoustic sound as a benchmark against which I (mentally, unreliably) compared reproduced sound, perhaps that was a mistake.

Ella Fitzgerald is interesting; she had a really quite remarkable instrument but how many people would ever have heard it direct - without mic and amplification?

I'm actually very happy with the way my Ella records sound.


N.B. the above queries are posed in the general spirit of 'the de-mystification of hi-fi'.

[A.S.]

... How good should I reasonably expect the sound to be though? If it doesn't live up to my hopes and expectations then what should I do?

The first thing to do is disconnect one speaker completely and place the other one well away from the walls, asymmetrically in the room, mono the source (you do have a mono button don't you?) and listen again.

I'm surprised you don't find Ella's voice a little reedy.

[weaver]

I guess Ella I have always just accepted the way she is - of course that may change now!

No, no mono button (or tone controls, or balance, sorry).

Does using a mono LP and taking a single channel (just in case of channel imbalance) from the phono stage achieve the same?

[A.S.]

Does using a mono LP and taking a single channel (just in case of channel imbalance) from the phono stage achieve the same?

Yes it does. Another way would be a phono two-into-one Y adaptor, so you can combine both output channels from the CD player into one channel on the amplifier. DO NOT combine the two speaker feed wires together - you'll destroy the power amplifier.

If you've never listened to 'single speaker mono' on voice before then you should. It's a completely different experience to a phantom voice generated (in your mind) in the centre of a stereo sound stage. It's far more like listening to the original, real voice, but of course, you lose all the spaciousness we associate with stereo reproduction. The replay loudness has to be just right.

[STHLS5]

... It's far more like listening to the original, real voice, .......

Thank you , thank you and thank you. Everyone was thinking I am crazy when I said somehow I feel vocal through one speaker sounded more correct.

ST

[A.S.]

...I feel vocal through one speaker sounded more correct...

As we've said here many times before, always revert to nature and to evolution to better understand how we interact with our environment.

Our hearing system has been evolving for millions of years. Let's guess at 25 million years. Now, the circumference of the earth is about 25,000 miles. If one mile = one million years, our hearing has been evolving for the equivalent of going around the world one thousand times. For almost all of that thousand-times-around journey, the sound that we humans have been programmed by evolution to hear and interpret is that of another human, directly, live, near us, in the same cave as us.

Then electric sound reproduction first appeared about one hundred years ago. But the voice we heard was still a single 'point source' from the single (horn) speaker. Then, only about fifty years ago stereophonic sound with two speakers was invented whereby the listener has to invent in his mind the illusion that there is a real voice hovering in space between the loudspeakers. That is clearly a psychoacoustic trick in the brain. Not everyone will be able to perform this trick. It depends how evolution has influenced their individual DNA.

Now, let's look at the timeline. Our ears have been developing for the equivalent of one thousand times around the earth. And the fifty years that stereo sound has been around - that's a journey equivalent to just fifty miles or 80km. In evolutionary terms that is a blink of an eye. It's so short a timeframe that it's utterly irreleveant relative to a thousand loops of the earth. It's not long enough (just two generations) for the process of natural selection to positively discriminate in favour of those with 'superior' hearing against the rest of us. So of course, our hearing is not (yet) optimised for interpretation of that 'phantom' voice, beamed at us from two different places in the room, the two loudspeakers.

To put these distances into perspective, the earth from space here. And our tiny island from space here. And now, the M25 motorway which circles London: it's about equivalent in length to the entire timeline of reproduced sound, since the cylinder reproducer - here. Half way around the M25 equates to the entire time that humans have been exposed to stereo sound.

Incidentally, one of the reasons that I listen to portable radios at home (and of just adequate quality, not hi-fidelity) rather than hi-fi equipment is because I am always busy with something - like typing here on the HUG - and I do not want to be pulled-in to sitting and listening to quality audio which would demand all of my attention. Even turning the hi-fi down low doesn't fool my ears; there are enough cues in the sound to make me curious, and try to get me involved!

[weaver]

Haven't got very far with the mono suggestion as yet; the issue is actually getting time at home to myself.

I did get twenty minutes the other afternoon and perhaps chose an unsuitable recording to try, voice with full orchestra. Oddly perhaps, given what has been said above, I found listening to a single source highly disconcerting in this instance. Even with eyes closed the effect of all the sound coming from a single place would I think take rather longer to acclimatise to than I had on this occasion, but as I said I don't think the orchestra was helping here.

I will revisit this with more time, but it did throw up a supplementary question which is to do with the sound dispersion characteristics of a given instrument (or individual).

As I understand it, a microphone needs to be placed sufficiently far from an instrument to pick up the 'full' sound, too close and the sound will be unbalanced due to the dispersion characteristics of that instrument.

Is it known whether we perceive anything distinct about the way sound is dispersed and use that information along with timbre to identify what is producing the sound? From an evolutionary point of view, knowing what kind of thing is making the sound: footfall of a large animal, rocks tumbling, trees falling, is very useful and I'd guess much of this identification is down to the quality of the sound, to timbre. But each of these things would also create a distinctive sound dispersion signature wouldn't it?

Coming back to voices, although we are all more physically similar to each other in the way we produce sound than we are to say a piano - are we sufficiently different to each other in the way our vocal instruments disperse sound for that to be a major element in the way we distinguish between different voices?

In designing a speaker, the way it will disperse sound in a typical domestic room is obviously of great importance. In making a recording is the only reference to the dispersion characteristics of the original instruments the degree to which the ambient acoustic (if there is one) is preserved? What I mean is: a speaker cannot disperse sound in the way that a piano does, but it can replay a recording of how that dispersed sound sounded in a particular space. If we then control the amount that that replayed sound is allowed to interact with our own room we may get an impression of what the original 'live' instrument sounded like - entirely dependent on how it was recorded of course.