I agree that this is an oversimplified representation (Le Cleac'h on "distorsion de phase, Melaudia 2005), but it opposes well the mandatory spatial distorsion between the real event and it's restitution on a couple of conventional speakers.
Some people can be satisfied with it, after all it's already better than the first Edison's experiments, but some others can have higher expectations.
The goal of having the same restitution that the producer had in his cabin is the simplest one on earth : just buy the same speakers and build the same room. No mystery.
The goal of reproducing something looking more real, on the basis of a 2 channels record (mixed as described on the lines before) is much more challenging and then much more interesting.
PS : as my post follow your's, please Dave don't take it as an aggression, I always enjoy your thoughts so much.
Some people can be satisfied with it, after all it's already better than the first Edison's experiments, but some others can have higher expectations.
The goal of having the same restitution that the producer had in his cabin is the simplest one on earth : just buy the same speakers and build the same room. No mystery.
The goal of reproducing something looking more real, on the basis of a 2 channels record (mixed as described on the lines before) is much more challenging and then much more interesting.
PS : as my post follow your's, please Dave don't take it as an aggression, I always enjoy your thoughts so much.
If the reproduction comes out like the picture on the right--that can definitely be a recording thing. I've got some very spacious recordings that came with my software that will blow your mind. Don't ask me how to do it, but it's not as simple as adding reverb. 2 channel stereo can be quite compelling I assure you. 3D even. Without any formal training, I can record better than that. I'm not even good for an amateur and my speakers are nothing special!
Dan
Dan
Indeed, one of the limitations of the picture I posted is written in french "enregistrement en stéréophonie d'amplitude"...pan potted stereo.
I guess that in your records you do some phase manipulations, or you got it with the mike pattern. Just guess, because the other day it was impossible to DL your song and run a proper autopsy.
I guess that in your records you do some phase manipulations, or you got it with the mike pattern. Just guess, because the other day it was impossible to DL your song and run a proper autopsy.
No need to risk a headache.It is most likely that I have used phrases like "When I listen to my speakers, I find ..." more than once. It didn't occur to me how literally such terms could be taken by a member of this forum.
Rudolf
Let's make it clear now ! Do you advocate speakers to be localisable, or not ?
If you want make the speakers localisable, it is totally different matter how you would design a speaker than if you want to make the speakers to disappear
- Elias
check out also this article by EJ Jordan for Wireless World, February 1971:
http://www.ejjordan.co.uk/PDFs/Jordan_WW_Feb_71.pdf
He discusses cross-in-front setup, center channel setup and even stereolit-like reflecting setups of the kind of Elias' and Radugazon's
it is when it allows for identification and localisation of the speaker as real sound source with spatial cues conflicting with spatial cues for virtual images - in consequence spatial cues for loudspeakers are superimposed on spatial cues for virtual images and width, depth and height of the soundstage as recorded cannot be reproduced because they are restricted by the width of the stereo base spread, the distance of the speakers from the listener and the height at which their acoustic centers are situated
yes, BTW stereolit-like setup reveals this perfectly - soundstage size is perfectly proportionate to reverberation, with just left-right cues with no reverb the soundstage collapses - no wonder as there is no stereo base spread in stereolit-like setup, OTOH recordings with real distance cues can sound unbelievably spectacular
yeah, bad recording (and every music loving producer would agree that music recording in an anechoic chamber is bad, quite absurd recording approach) should sound bad on good equipment - the better equipment the worse
BTW from this perspective stereolit-like setup must be the best because best recordings sound best on it and the worst - worst
oh yeah, and Moulton is not a Grammy winning producer and Bob Ohlsson is not a legend but both are just naive unexperienced guys with bad hearing because they can hear imaging precision where there cannot be one as we are told authoritatively by Dave S.
it seems that there are things You have never even dreamed of
check this article by EJ Jordan for Wireless World, February 1971:
http://www.ejjordan.co.uk/PDFs/Jordan_WW_Feb_71.pdf
He discusses cross-in-front setup, center channel setup and even stereolit-like reflecting setups of the kind of Elias' and Radugazon's
You have heard of Jordan, have You? Precise imaging is very high on His priority list for over 50 years now
Last edited:
nah
You have already taken the red pill - out of the matrix of "cannot have both" - You are in the Wonderland, aren't You? The answer to the original question seems to be that the best directivity pattern is a speaker that has directivity and is not omni. Both the people who do not like early reflections and those who do not like to hear the original speaker (flooder and sterolith) need a directional speaker the only difference is in the positioning.
Recordings usually have these envelopment ques embedded, but under normal 2 speaker stereo get reproduced from the wrong direction(from front 30deg).
You can argue for 38 pages how to correctly illuminate sidewalls to get some envelopment from the recording, or upmix the 2 channel material to ie. 7 channels. Using centre speaker gets mostly rid of the rigid listening spot requirement (and some crosstalk problems) of 2 channel stereo. Correctly processed side speakers (60-80deg. placement) reproduce envelopment ques from optimal location - so you can have it both ways, envelopment AND imaging. Also a plus is a generation of listening area, instead of listening 'spot'.
Technology is here, called Logic 7 and Dolby-s PL2(music).
I agree. I've heard some really mind blowing "stereo" recordings that played back on on two speakers at standard angles is totally enveloping with a wide sense of depth and space in every direction, and yet not at all artificial or gimmicky sounding - and this is with fairly directional speakers that don't excite the side-walls too much, so it's not a function of the room.
Clearly the magic is in the processing in the recording, and if I had to guess I would say that they used every trick in the book to encode the right cues in the recording to place the sounds in 3D space - including crosstalk cancellation, cues such as ITD, HRTF, reverberation, and so on. Basically an electronically generated version of binaural audio but with crosstalk cancellation encoded in the mix so that it works on speakers rather than headphones.
If we all agree that envelopment comes from sound sources that are at a fairly wide angle (The Audessy page suggests 120 degrees eg +/- 60 degrees) then its relatively easy to take individual components of the "sound-scape" and apply a 30 to 60 degrees horizontal HRTF transformation to make these components appear to be coming from +/- 60 degrees, and thus get a sense of envelopment from them, even though the physical speakers are only at 30 degrees off centre.
(Even the "width" processor on my DEQ2496 seems to do something similar, and quite successfully on most recordings)
What's surprising in some of these recordings is you can still turn your head as much as 20-30 degrees away from the centre line without the the illusion falling apart, and even moving sideways significantly only seems to affect localization of sounds from behind you - they shift somewhat, but the overall sense of envelopment and space remains.
Pan-potted stereo really needs to die, it just doesn't sound any good and has a very narrow sweet spot with the range of localizable images remaining between the speakers, clearly there are far superior encoding systems being used on some recordings, but we as listeners don't know what techniques were used on what recordings, thus greatly confusing the whole discussion about imaging, envelopment, and what speaker/room configurations are need to get it. (With what type of recording ? etc...)
Let's make it clear now ! Do you advocate speakers to be localisable, or not ?
Elias
Elias,
in your quotation I was talking about a single speaker. That would have been clear immediately, if you hadn't omitted my following sentence from your quotation.
My position is clear as daylight: If I listen to a single speaker, that speaker (not the drivers!) should be precisely locatable.
If I listen to a stereo pair of speakers, those speakers should not be locatable - except for the case, that a sound is panned exactly and solely to one of the speakers.
I'm somewhat irritated that I even have to mention such basics in this thread. It should have gone without saying IMHO.
Rudolf
My position is clear as daylight: If I listen to a single speaker, that speaker (not the drivers!) should be precisely locatable.
If I listen to a stereo pair of speakers, those speakers should not be locatable - except for the case, that a sound is panned exactly and solely to one of the speakers.
I'm somewhat irritated that I even have to mention such basics in this thread. It should have gone without saying IMHO.
Hey Rudolf! basics is to know that if a speaker is designed in such a way so that it is perfectly localizable in mono then it will remain perfectly localizable in stereo as well - after all it will be the same speaker!
the difference is that in stereo localizability of the speakers will not lead to hearing speakers instead of virtual sound sources but that it will lead to superimposition of spatial cues for the speakers on spatial cues for virtual sound sources and in consequence it will lead to known limitations of the conventional setup as to reproducing the 3D of the soundstage - width, depth and height of it
Last edited:
Would aspects look (sound) different, if we were to place the speakers
behind a curtain consequently ?
A phantom panned solely to one of the speakers is likely to have the
least fuzzyness in lateralisation when using pure intensity stereophonic
recording and processing.
We maybe won't be irritated to same extent by sounds coming
"from the speaker" or even "from one speaker" then, because we cannot
see the speaker's anymore.
But if there is some fuzziness in phantom lateralisation for phantoms
inbetween the extreme left and right position, phantoms at the extremes
may nevertheless stand out because of being more definetely localized
than other phantoms.
In case a system is able to pinpoint imageing at every angle, a phantom
at the extremes may not stand out.
My first goal would be to have all phantoms presented with about the
same degree of definition, may that be a somewhat higher or lower
degree of definition.
Having that definition of phantoms consistent all over the soundstage is
a plus of its own in "realism" IMO.
Of course this is not a goal to be achieved in the reproduction domain solely,
as it is also due to the recording technique.
In case of pure intensity stereophony phantoms will be more definitely lateralized
at the extremes and in case of time delay stereophony phantoms will be more fuzzy
at the extremes.
From a consistency point of view, in a (recording and reproduction) system not
able to produce phantoms lateralized sharply inbetween, a speaker
"fuzzyficating" its own location to the right amount would appear like
having an advantage ... given that definition of the inbetween phantoms
does not suffer from that moreover ...
Last edited:
This seems to be the essential debated "fact' of this thread. It is perfectly sensible that a dry recording sent to a single speaker should sound like a source present in the room at the point of the speaker. This was my example of spoken voice recorded in an anechoic chamber. Not that this was desirable recording technique, but that it represents one extreme of possible stereo effect: dry and present.
Yet some think that it should be transformed into a spaciously broad stereo effect on its own. Not only should we get spaciousness and diffusion when the recording contains it, but also when it doesn't. If the recording covers a range of sources, some near and others far, but the speaker forces a large amount of diffusion, then you are clearly losing the ability to bring sources present into the room. You may pick up on the spacious end of the scale but must lose on the presence end.
I've stated two tests to differentiate between systems with varying stereo clarity. How different is sound when you switch from stereo to mono? And how great is the difference between in phase and out of phase? Clearly, the more we accentuate the role of the room and the less we hear the direct sound of the speaker, the more we diminish the difference.
What is your definition of accuracy?
David S.
Unfortunately it works the other way around too: "You may pick up on the presence end of the scale but must lose on the spacious end."
Oliver,
I have reached a certain level of phantom image "sharpness". At this level it becomes very obvious how much recordings can differ in how they present the stereo stage. That reaches from excellent recordings that take full advantage of interchannel phase and intensity difference to rather bad recordings that are purely (and poorly) intensity-panned to left, center and right – with nothing in between. By increasing the rear radiation of my dipoles I could increase the fuzziness of those “bad” phantom images up to a point where the stereo stage sounds more life-like.
But this would compromise the sharp definition of phantom images, which is present in better recordings. I understand very well your point of view, but I’d rather go for listening to less recordings in best quality than more recordings in compromised quality.
You are mentioning a changing fuzziness between phantoms at the extremes and phantoms between the extreme left and right position (but NOT the center in my case). I can’t deny that it exists in my system too, but the system induced differencies are much less than the differences due to recording issues. Yet I have not found a way to “fuzzyficate” phantoms, which are "glued” to the speaker position, without blurring the phantoms in between.
Rudolf
actually smart intellectual provocation is what I value most as it inspires the mind
OTOH stupid bashing provocations discouraging people from going off the beaten tracks to try new things and discover new ways is what I hate in discussions most
__________________
the above is to be understood as purely and sincerely constructive
your signature is quite ironic.
Dave,
I subscribe to most what you have written above. And I would like to add my definition of what distinguishes spaciousness from diffusion:
Spaciousness is the ability to "free" the space between instruments (/voices/effects/whatsoever) from erratic sound. Diffuseness is the ability to "fill" the space between instruments with erratic sound.
We all know how the quality of music is defined by the space between the notes more than by the notes itself.
... to regain in my room as much as possible from the recording. This would encompass spatial aspects as well as tonal or even emotional ones.
I am very well aware that I cannot recreate the mixing/monitoring situation of the recording in my room. But even without THAT accuracy, there is usually more "information" in a good recording than most playback chains can deliver. I remember listening to Nils Lofgrens "Keith don't go" through a pair of (budget) speakers, which were switched between "normal" and "phase linear" at the listening position. What an incredible difference in spatial resolution! What a gain of small details. I would not want to loose any of it in some diffusive "fog".
Rudolf
I subscribe to most what you have written above. And I would like to add my definition of what distinguishes spaciousness from diffusion:
Spaciousness is the ability to "free" the space between instruments (/voices/effects/whatsoever) from erratic sound. Diffuseness is the ability to "fill" the space between instruments with erratic sound.
We all know how the quality of music is defined by the space between the notes more than by the notes itself.
... to regain in my room as much as possible from the recording. This would encompass spatial aspects as well as tonal or even emotional ones.
I am very well aware that I cannot recreate the mixing/monitoring situation of the recording in my room. But even without THAT accuracy, there is usually more "information" in a good recording than most playback chains can deliver. I remember listening to Nils Lofgrens "Keith don't go" through a pair of (budget) speakers, which were switched between "normal" and "phase linear" at the listening position. What an incredible difference in spatial resolution! What a gain of small details. I would not want to loose any of it in some diffusive "fog".
Rudolf
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.