A question about what is measurable.

[Karl-Heinz Fink]

Thanks KH. What do you use for measurements of prototypes? Piezo strips or something else?

We use a Laser Scanning Vibrometer like this: http://www.polytec-ltd.co.uk/uk/pro...rometers/psv-500-compact-scanning-vibrometer/
We can read in those measurements into a software we wrote in_ house and calculate the radiated power response in relation to the active driver.

Best KH

 

[amirm]

Oh, that is really cool.

 

[Cosmik]

...I would assume that "soundstage depth," is a function of the relative volumes of sounds in the mix, and "soundstage" itself is a nontechnical term for a combination of that depth, and the placement of sounds in the stereo mix. I would further assume that playback equipment's ability to reproduce that stage properly is a function of well-defined and quite measurable properties like dynamic range, noise, plus distortions like IMD, jitter or harmonic distortion that can make sounds seem more or less prominent than they are on the recording, channel separation, etc. I would assume that "timbre" is similarly a function of FR, noise and distortion.

All much more clearly defined terms than "soundstage" and "timbre," all measurable. What have I missed?

As was mentioned in another post in this thread, even a mono recording can give a perception of depth due, I would say, to the cues of volume level, reverberation and frequency response - start recording your voice close to a mic and walk away while talking in a room: the perception of changing distance is very clear. However, we could intentionally destroy that 'depth' by manipulating the frequency response and playing with the dynamics e.g. an extreme effect would involve gating the speech in order to suppress the reverberation. Any process that obscured the 'cues' would destroy the depth. Ultimately we could deconvolve the recording with the room's impulse response at that position and remove all sense of distance.

An audio system that is burdened with intermodulation distortion and dynamic compression would, presumably, affect 'depth' even if everything else was OK.

 

[fas42]

Yes, some of the old mono recordings have tremendous depth in them, because of the very simple recording techniques used: a single microphone, with everyone in a large space, stacked in layers so everyone could get some "exposure" to the mic. A competent system reveals this well, and you get interesting perspectives such as "hearing" that the drummer is way at the back of the room, behind everyone else - so as to not overload the mic, presumably.

Too much distortion, etc, kills all those acoustic clues that make this clear - hence come across as a bit of a jumble of sound.

 

[Cosmik]

Are there any analogies to be drawn with photography?

2D photographs can provide a sense of depth, and not just because of objects occluding other objects and a priori knowledge of the scale of objects. I think there are 'leap factors' concerning brightness, the scale of the image, and peripheral vision. Viewing a photograph as a projected slide rather than a print always seemed to provide a sense of '3D' I seem to remember. In recent years I found that scanning many of my old photographs and viewing them on a computer screen seemed to do this, too - which, presumably, came from seeing them larger and brighter than before.

Just as with photography, does audio need to reach the 'peripheral vision' in order to create a convincing illusion? If we sense limits to volume and frequency do we reject the illusion, even if the measurements are otherwise perfect?

 

[Fitzcaraldo215]

Are there any analogies to be drawn with photography?

2D photographs can provide a sense of depth, and not just because of objects occluding other objects and a priori knowledge of the scale of objects. I think there are 'leap factors' concerning brightness, the scale of the image, and peripheral vision. Viewing a photograph as a projected slide rather than a print always seemed to provide a sense of '3D' I seem to remember. In recent years I found that scanning many of my old photographs and viewing them on a computer screen seemed to do this, too - which, presumably, came from seeing them larger and brighter than before.

Just as with photography, does audio need to reach the 'peripheral vision' in order to create a convincing illusion? If we sense limits to volume and frequency do we reject the illusion, even if the measurements are otherwise perfect?

I think one can go too far with analogies to photography. But, yes, photos, like paintings, can suggest 3D depth from a 2D medium. Painters have known the optical illusion tricks to be able do this quite deliberately and effectively for many centuries. Photos can do it, too, depending on a number of things. I do not think peripheral vision is important in that, however. Just magnification and things like depth of field can enhance the visibility of the diminishing focus of detail off into the distance in conjunction with geometric and "atmospheric" perspective.

But, yes, audio suggests 2D depth from a pair of stereo speakers, which for direct stereo sound is 1D in idealized terms - a line between 2 point sources. However, the sound field in your room as received by your ears is actually 3D as a result of reflections. A normal photo is effectively 2D as far as your eyes are concerned, though it can suggest 3D. So, this is one area where the comparison of audio and photos/paintings breaks down.

As Amir said earlier, the perception of sonic depth is in your head as your mind pieces the raw information from your ears together. Of course, there are no simple measures in audio that we can relate to depth of image. I doubt there ever will be, though we all have our own theories as to what are the important factors in that sonic depth creation. I think "good" depth reproduction requires integration of a gamut of audio performance issues, one key one being good timing/phase accuracy. That is one reason I am interested in Bob Stuart's "timing blur" concept and his claim to be able to reduce it with MQA. It is also a reason that I rely heavily on spatial image recreation, including depth, in subjectively judging audio equipment and recordings. They have to be good in a lot of different ways in order to reproduce good depth and space.

But, getting back to the peripheral vision thing, it is clear to me perceptually that Mch delivers a greater sense of depth of image than does stereo, much as stereo can do a better job than mono. Again, it is not measurable. I also can only speculate as to why because it is complicated. The center channel plays a key role, I believe, as does the enhanced sense of the surrounding performance space, as part of our "peripheral" hearing. I suspect there is much more to it than that, such as phantom imaging created in the sound field in the room by the interaction of front and surround speakers.

 

[Phelonious Ponk]

As Amir said earlier, the perception of sonic depth is in your head as your mind pieces the raw information from your ears together. Of course, there are no simple measures in audio that we can relate to depth of image.

Are you sure? Layman's POV, correct me if I'm wrong, but it seems like the lion's share of depth comes from the mix; from the relative volume and eq of the elements in that mix, and the ability of your system to reproduce it accurately. Deliver it to the ears and let the brain do it's work. Simple? No. To really get at it, I suppose you'd have to be able to measure the instruments, or at least the tracks, individually. But if you start with good depth and dynamics in the recording, would the cleanest, most accurate reproduction of that recording give the brain the best information to deal with in order to create the perception?

Tim

 

[DonH56]

Volume and delay (and/or reverb, many ways to include or even "fake" depth).

 

[fas42]

Old mono recordings had no fancy techniques, apart from spacing the musicians from the microphone - the reverb information from that straightforward method comes through quite clearly on a competent system, and makes it possible to estimate the distance they're back, one can almost point to where they're located, by throwing a stone the right distance, so to speak.

 

[Fitzcaraldo215]

Are you sure? Layman's POV, correct me if I'm wrong, but it seems like the lion's share of depth comes from the mix; from the relative volume and eq of the elements in that mix, and the ability of your system to reproduce it accurately. Deliver it to the ears and let the brain do it's work. Simple? No. To really get at it, I suppose you'd have to be able to measure the instruments, or at least the tracks, individually. But if you start with good depth and dynamics in the recording, would the cleanest, most accurate reproduction of that recording give the brain the best information to deal with in order to create the perception?

Tim

Yes, well, I am a layman and a little bit of knowledge is a dangerous thing. So, my opinion might be entirely poorly educated speculation.

Not sure how EQ figures in directly, though it could alter frequency and timing/phase. Yes, the "mix" plays a role. But, if you start with a 2 mike stereo recording, what is there to mix? There it is just physical instrument placement, same as in mono depth. If you are mixing multitrack down to stereo, yes, relative front-rear volumes, and degrees of reverb/reflections generated by the instruments in the mix are the main cues as to depth. Mix in more reflections/reverb, and it sounds further away. But, I suspect this is all an oversimplification and there is still much more to it.

How do your ears tell the difference between the direct sound of the instruments and the slightly delayed reflections and reverb? Oversimplified, I think several things are important in that. Timing accuracy is crucial, because the reverb/echo is delayed and extended in time. There is a time and frequency spectrum of the reflected sound from any given instrument that is not the same as its direct sound, though the two are in some ways correlated for sounds from the same instrument. That is true in the venue or in the mix, as it is also in your room.

Directional cues are also important because the reflections and reverb are more diffuse than direct sound and they strike your ears from multiple different directions, even from all directions but unequally, in the hall, unlike direct sound from a single instrument. And, there are also perceived tonality differences caused by the interaction of direct with reflected sound. The same violin does not sound quite the same tonally when played in different venues or even in different seats in the same venue.

Again, it is complicated, especially when it seems so seemless, unconscious and natural when listening. I am unaware of any way to measure it, and its complexity would seem to defy objective measurement on a specific recording or in a specific playback system.

 

[Cosmik]

I am unaware of any way to measure it, and its complexity would seem to defy objective measurement on a specific recording or in a specific playback system.

I see things in a very simple way: the perception of depth in a recording may be a very complex phenomenon, but the system just has to play the recording back correctly.

For me, the recording has to be reproduced correctly as a whole - I don't see that it would be possible to have dodgy distortion, frequency response and dynamics but superb 'depth' reproduction.

However, I am sure there are people who do see things that way: a 3W valve amp playing into a single-driver paper cone (with 'whizzer'!). If they choose their music very carefully and don't play it too loud, they might just be able to ignore the obvious drawbacks and fool themselves into hearing fantastic soundstage depth.

 

[Fitzcaraldo215]

For me, the recording has to be reproduced correctly as a whole - I don't see that it would be possible to have dodgy distortion, frequency response and dynamics but superb 'depth' reproduction.

I quite agree, as far as it goes. But, I think there is even more to it than just the factors listed. Hearing, as easy as it is to do, is never simple when we try to analyze the factors that affect our perception of it.

Since depth and imaging is perceptual, I think one can also be fooled. Way back, I had a preamp with a multiband equalizer. I swore it delivered better "depth" when I attenuated a few octaves above 1K Hz a bit. By hindsight, I think it was a false illusion. The point is maybe non-linear frequency response or other factors can sometimes fool your perception of greater depth. God knows, also, what phase and timing issues were introduced by that old analog EQ.

 

[fas42]

I see things in a very simple way: the perception of depth in a recording may be a very complex phenomenon, but the system just has to play the recording back correctly.

For me, the recording has to be reproduced correctly as a whole - I don't see that it would be possible to have dodgy distortion, frequency response and dynamics but superb 'depth' reproduction.

Yes. IME this is the approach to use - every recording I have yields more, "usable" depth information when the system works well - when an audio setup is fully competent everything falls into place, no excuses have to be made for the recording techniques, the age, the original media, anything - the sound, just, works ...

Yes, there is "fooling" going on, but it's exactly the same process we use when we listen to "seamless", live sound - the brain does all the processing necessary to make full sense of the auditory universe it finds itself in - and it doesn't find anything out of kilter ... which is not the case when an audio system is not doing a good enough job of "faking" it.

 

[Phelonious Ponk]

Yes, well, I am a layman and a little bit of knowledge is a dangerous thing. So, my opinion might be entirely poorly educated speculation.

Not sure how EQ figures in directly, though it could alter frequency and timing/phase. Yes, the "mix" plays a role. But, if you start with a 2 mike stereo recording, what is there to mix? There it is just physical instrument placement, same as in mono depth. If you are mixing multitrack down to stereo, yes, relative front-rear volumes, and degrees of reverb/reflections generated by the instruments in the mix are the main cues as to depth. Mix in more reflections/reverb, and it sounds further away. But, I suspect this is all an oversimplification and there is still much more to it.

How do your ears tell the difference between the direct sound of the instruments and the slightly delayed reflections and reverb? Oversimplified, I think several things are important in that. Timing accuracy is crucial, because the reverb/echo is delayed and extended in time. There is a time and frequency spectrum of the reflected sound from any given instrument that is not the same as its direct sound, though the two are in some ways correlated for sounds from the same instrument. That is true in the venue or in the mix, as it is also in your room.

Directional cues are also important because the reflections and reverb are more diffuse than direct sound and they strike your ears from multiple different directions, even from all directions but unequally, in the hall, unlike direct sound from a single instrument. And, there are also perceived tonality differences caused by the interaction of direct with reflected sound. The same violin does not sound quite the same tonally when played in different venues or even in different seats in the same venue.

Again, it is complicated, especially when it seems so seemless, unconscious and natural when listening. I am unaware of any way to measure it, and its complexity would seem to defy objective measurement on a specific recording or in a specific playback system.

Where are you finding those?

Tim

 

[Phelonious Ponk]

I see things in a very simple way: the perception of depth in a recording may be a very complex phenomenon, but the system just has to play the recording back correctly.

For me, the recording has to be reproduced correctly as a whole - I don't see that it would be possible to have dodgy distortion, frequency response and dynamics but superb 'depth' reproduction.

However, I am sure there are people who do see things that way: a 3W valve amp playing into a single-driver paper cone (with 'whizzer'!). If they choose their music very carefully and don't play it too loud, they might just be able to ignore the obvious drawbacks and fool themselves into hearing fantastic soundstage depth.

This would be my guess.

Tim

 

[tomelex]

yep Tim.

some of you all might want to google something like adding depth to your recordings.....reverb and hf can do it, etc. it IS in a recording if put there, and then of course, the rest of your system needs to pass that info along with the least degradation as possible. You can learn more about stereo by reading up on and visiting forums where they talk about mixing. The stereo "image" is in the recording as in two channels of information, however it does not materialize in a manner we can use until our two ears and brain system captures it, and is fooled (some more than others)by it into creating the center area fill.

 

[Fitzcaraldo215]

Where are you finding those?

Tim

If you mean 2-mike minimalist stereo recordings, I doubt there is much at all, especially not in pop/rock. I know of a few very tiny classical labels that do it, but again there is not much. In the 50's, some classical labels tended that way. Others used a 3 mike to stereo mixdown. Water Lilly also did some classical in 2 mike Blumlein decades ago - interesting, and spatially revealing, but not great sounding.

But, I am not a stereo guy. I buy only Mch, which is mostly all that I listen to. Channel Classics is a small, excellent European SACD label that pretty much uses a minimalist setup, but I do not know how they do their stereo mix. Waldrep's AIX on BD is pretty minimalist from videos I have seen, but he is Mch oriented, too. He does a fair bit of non-classical stuff.

OTOH, I am also generally quite happy with a lot of highly multi miked Mch on the classical music I enjoy. It all depends on the engineering. Minimalist miking does not necessarily guarantee a better recording. It might sometimes capture better space and depth, but there are often downsides, too.

But, there is a recent non-commercial "test" recording in limited circulation I have from an engineer friend who worked with Channel Classics on their Mahler 3rd recording sessions in Budapest. It was done in Mch with just 5 new, very hi tech mikes in DSD256. It is the best recording I have ever heard. Atkinson and Rubinson of Stereophile agree, as does Quint of TAS. It was used recently in demonstrating the Merging NADAC at T.H.E. Show Newport/Irvine to much acclaim, I understand. I was not there. Also, the engineer is resistant to creating a stereo mix, but one may emerge, nonetheless. And, this excellent recording might become available as a download from Channel.