I loved the 2 cu ft box variations, the BC1 was chosen using my voiceand several others, but I was not keen on it. The forerunner the LS 3/4 was a lovely speaker, and that Celestion unit absolutely blew me and many others, away at that time - way ahead. I had a 3/6 here a couple of years ago, and a friend has just offered me a loan of the Rogers Export Monitor.
I nearly bought a pair of Revel Salon2s a few weeks ago, the specs are amazing, and the mid on that F208 is very good looking.
It is hard trying as an amateur to do design at home, lack of resources, inconvenience to domestics, and in my case considerable insecurity about computer measurement, come to think of it about computers. But after hearing my Rogers/ESS Heil attempt, A.K of The Funk Firm said that he could sell it for £4k. It cost me about £1.5k, and much, and I mean a lot, of work. I would post a pic, but i have concerns about putting stuff on public hosting sites. (not here).
I nearly bought a pair of Revel Salon2s a few weeks ago, the specs are amazing, and the mid on that F208 is very good looking.
It is hard trying as an amateur to do design at home, lack of resources, inconvenience to domestics, and in my case considerable insecurity about computer measurement, come to think of it about computers. But after hearing my Rogers/ESS Heil attempt, A.K of The Funk Firm said that he could sell it for £4k. It cost me about £1.5k, and much, and I mean a lot, of work. I would post a pic, but i have concerns about putting stuff on public hosting sites. (not here).
I would politely suggest again that the 'correct' was answer given by Harwood (post #68), not by us guessing about what he and his other eminent engineering colleagues at the BBC might have been thinking instead.
No. The stereo difference channel (S) contains the first order lateral information. When replayed from the front, the reason for requiring a dip will still be present.
The first (frequency-wise) collapse of the stereo image is not the cause of the differences in hearing acuity with direction that the dip compensates. The phantom image collapse happens in repeating frequency intervals above the dip commensurate with the comb-filtering in two loudspeaker reproduction of the stereo sum channel (M).
Last edited:
"The first (frequency-wise) collapse of the stereo image is not the cause of the differences in hearing acuity with direction that the dip compensates. "
Surely a collapse of image is a result rather than a cause; it is an end result which we are discussing as a failure of stereo. Isn't any difference in hearing acuity causally contributory, and what any compensation maybe be an attempt to ameliorate?
Surely a collapse of image is a result rather than a cause; it is an end result which we are discussing as a failure of stereo. Isn't any difference in hearing acuity causally contributory, and what any compensation maybe be an attempt to ameliorate?
I never implied (or intended) otherwise. My apologies if it was not clear.
The difference in hearing acuity re direction is the cause of the use of the dip to ameliorate the inability of stereo to reproduce higher frequency directional information encoded in the difference channel because the loudspeakers are in front of the listener.
The cause of the phantom image collapse is comb-filtering in the stereo sum signal because there are two loudspeakers.
They are then different issues that are nevertheless mathematically related.
Notably in three speaker stereo reproduction, the collapse of the phantom image can be prevented altogether, but the need for the dip (if it is needed) will remain.
That is about as much help as I can offer on the semantics...
Tangent RS4s deserve a special mention.They were very "BBC" sounding and were deliberately designed to have a dip at the crossover point [around 2000-3500 hz].
They were my first speakers and I loved them.They did have a recessed sort of sound but you get used to that sort of sound and other speakers then sound too forward.People commented on how they imaged way back behind the plane of the speakers .So you could say that they had collapsed stereo image but they still had reasonable image depth.
That is also how the Revel Performa F208 sound which is why I mentioned them.
It seems to me a lot of people prefer that sort of presentation.It works especially well with human voice and classical music.
It seems to me a lot of modern speakers sound too forward and maybe would benefit from a slightly attenuated presence region.That sort of presentation can get pretty tiresome.
They were my first speakers and I loved them.They did have a recessed sort of sound but you get used to that sort of sound and other speakers then sound too forward.People commented on how they imaged way back behind the plane of the speakers .So you could say that they had collapsed stereo image but they still had reasonable image depth.
That is also how the Revel Performa F208 sound which is why I mentioned them.
It seems to me a lot of people prefer that sort of presentation.It works especially well with human voice and classical music.
It seems to me a lot of modern speakers sound too forward and maybe would benefit from a slightly attenuated presence region.That sort of presentation can get pretty tiresome.
Last edited:
I do not think that anything I have stated has any semantic ambiguities, but that some of your posts are inferential soundbloke.
You refer to post 68, and so do I as follows;
"
The next point to be discussed is the question of optimum axial frequency response. This question is not concerned with how wide a frequency range should be covered, but what shape the response curve should be. First of all the underlying assumption must be clearly stated. This is that both the microphone and all associated amplifiers have a uniform, frequency response. The usual conclusion is that the loudspeaker should also have a uniform axial frequency response but this is precisely what is being challenged. Not even in stereo reproduction are the sound wave- fronts produced in a listening room similar to those heard in the studio or concert hall and it therefore seems clear that if by "bending" the axial response curve of the loudspeaker a more realistic psychological impression is obtained, then this is entirely justified. Thus, for example, if a uniform output is maintained at all frequencies an orchestra sounds extremely close.
This condition is quite unnatural and a much better sense of perspective is obtained if a slight dip in the 1 to 3kHz region is applied. About 2dB is sufficient to provide the more distant perspective without destroying the sound quality. It may well be that as techniques progress other such tricks will follow. All that is intended at this stage is to get away from the rigid idea that a uniform axial response is necessarily the best."
I respect Harwood greatly, but I think this is an entirely subjective assertion by him, based I presume on how things sound to him.
You state;
"The difference in hearing acuity re direction is the cause of the use of the dip "
Yes that is the assertion, though cause seems to imply some direct link when in fact it is a human response to what is perceived to be an erroneous situation.
I feel sure that comb filtering is a cause of loss of imaging to an extent.
I still do not see a coherent argument for implementation other than 'it sounds better'.
You refer to post 68, and so do I as follows;
"
The next point to be discussed is the question of optimum axial frequency response. This question is not concerned with how wide a frequency range should be covered, but what shape the response curve should be. First of all the underlying assumption must be clearly stated. This is that both the microphone and all associated amplifiers have a uniform, frequency response. The usual conclusion is that the loudspeaker should also have a uniform axial frequency response but this is precisely what is being challenged. Not even in stereo reproduction are the sound wave- fronts produced in a listening room similar to those heard in the studio or concert hall and it therefore seems clear that if by "bending" the axial response curve of the loudspeaker a more realistic psychological impression is obtained, then this is entirely justified. Thus, for example, if a uniform output is maintained at all frequencies an orchestra sounds extremely close.
This condition is quite unnatural and a much better sense of perspective is obtained if a slight dip in the 1 to 3kHz region is applied. About 2dB is sufficient to provide the more distant perspective without destroying the sound quality. It may well be that as techniques progress other such tricks will follow. All that is intended at this stage is to get away from the rigid idea that a uniform axial response is necessarily the best."
I respect Harwood greatly, but I think this is an entirely subjective assertion by him, based I presume on how things sound to him.
You state;
"The difference in hearing acuity re direction is the cause of the use of the dip "
Yes that is the assertion, though cause seems to imply some direct link when in fact it is a human response to what is perceived to be an erroneous situation.
I feel sure that comb filtering is a cause of loss of imaging to an extent.
I still do not see a coherent argument for implementation other than 'it sounds better'.
If you're looking for a mathematical quantification of equalisation curves, the better known types are generally derived from controlled listening tests and (usually) a statistically significant preference from the participants, preferably a large number taken over a reasonable amount of time to account for differences in health & whatever other external factors cannot be controlled under the available conditions. The test methodology is usually provided in relevant papers when published or indicated in associated articles.
That is typically the realistic limit of mathematics in this specific regard, whether for the mild midband EQ referred to in this thread, Blauert & other EQ curves. While you can show, for example, a graph with a target curve or set of equations, human hearing itself cannot be expressed purely in those terms, since we are inherently variable biological animals. So the various papers naturally provide variations on the theme of '[statistically significant] recorded responses', because that is what they are and how they have to be conducted. The same applies to thousands of medical studies on human hearing which attempt to provide a mathematical analysis on various auditory perceptions: the human being is still the variable, so these studies too usually employ a variation on a group of test subjects and statistically significant responses (however taken) to whatever matter happens to be in question.
That is typically the realistic limit of mathematics in this specific regard, whether for the mild midband EQ referred to in this thread, Blauert & other EQ curves. While you can show, for example, a graph with a target curve or set of equations, human hearing itself cannot be expressed purely in those terms, since we are inherently variable biological animals. So the various papers naturally provide variations on the theme of '[statistically significant] recorded responses', because that is what they are and how they have to be conducted. The same applies to thousands of medical studies on human hearing which attempt to provide a mathematical analysis on various auditory perceptions: the human being is still the variable, so these studies too usually employ a variation on a group of test subjects and statistically significant responses (however taken) to whatever matter happens to be in question.
Last edited:
Well, this is more or less where they were at during the mid '70s:
...This raises the whole question of what we are trying to measure and why. In the BBC the spherical response of a number of loudspeakers has been measured and efforts made to correlate it with sound quality in a live room, but with very little result.
When for example we listen in a room of normal reverberation time to a rather directional loudspeaker on its axis, it is common experience that the sound quality does not change drastically when in the near or reverberant sound field. On the other hand if we were really listening simply to the sound pressure at these two points then the direct response and the spherical response would indeed be the determining factors. Furthermore a similar factor must be involved in the fact that with such a loudspeaker in a live room the directional properties are clearly audible even when listening well into the reverberant field.
These experiences indicate clearly that the spherical response is not the predominating factor in determining sound quality under live listening conditions and to check this a formal experiment was carried out at BBC Research Department. A monitoring loudspeaker was taken having three units and representing as omnidirectional a device as was possessed at the time, and for comparison an 8in wide range unit representing as directional a device as was likely to be met. Listening on axis in a free field room and using speech and a team of experienced observers, the two were equalized by ear to sound as closely similar as possible. They were then transferred to a listening room well away from the walls; the room had a reverberation time of about 0.4s, and the loudspeakers were again compared, listening on the axis. The results in the two conditions were almost identical within the experimental error, although a small change towards the known spherical performance could be discerned but not guaranteed. The conclusion therefore was that it is essentially the direct sound which determines the sound quality and not the spherical response. The measurement of frequency response at various angles in a free –field room is therefore a much better indication of performance than the spherical response even when listening in the reverberant field, and this has been confirmed by careful listening tests many times since.
Harwood, H. D., BBC Research Department, 'Some Factors in Loudspeaker Quality' Wireless World May 1976 pp. 47-48
He also referred to some of the factors involved in (for example) horizontal directivity & the use of the slot-load over the LF unit of the LS5/5 to enhance this, but as noted, I can't quote the whole article here or the post would take up a fair bit of space.
Interesting discussion really. You usually go for flat frequency response in a studio monitor, and a downward slope in a domestic speaker.
Robin Marshall's solution to integrating drive units with very different dispersion was shallow slopes, as in the Epos range.
But maybe I am onto something here:
Bigger tweeter, simple as.
And don't forget that 90 degree phase Butterworth fills the power hole anyway in either polarity, the dotted line being negative polarity:
Wouldn't surprise me if John DeVore did this in the Orangutan o/96:
Robin Marshall's solution to integrating drive units with very different dispersion was shallow slopes, as in the Epos range.
But maybe I am onto something here:
Bigger tweeter, simple as.
And don't forget that 90 degree phase Butterworth fills the power hole anyway in either polarity, the dotted line being negative polarity:
Wouldn't surprise me if John DeVore did this in the Orangutan o/96:
That last excerpt seems to clarify one thing at least, so now the question is clearly why, having eliminated the concerns about indirect sound, the direct sound should need to be adulterated. But I can see clearly that recordings done with near field mics. may be too bright, because the mics. pic up more presence and top. (Used by Previn and praised.)
So now, if a mic. records at a distance of say, 10', and we replay that recording at a listening distance of 10', where and how are errors introduced?
Simple is also one design aim. There has been a school moving towards low order filters, (Wilson Benesch), but of course the problem is getting wide band enough drivers. I often find very abrupt Xovers quite detectable.
Lots of people use just a cap to couple a tweeter, and it seems to me that this is a great risk both from physical damage and distortion.
So now, if a mic. records at a distance of say, 10', and we replay that recording at a listening distance of 10', where and how are errors introduced?
Simple is also one design aim. There has been a school moving towards low order filters, (Wilson Benesch), but of course the problem is getting wide band enough drivers. I often find very abrupt Xovers quite detectable.
Lots of people use just a cap to couple a tweeter, and it seems to me that this is a great risk both from physical damage and distortion.
Last edited:
I will take that as a compliment. All of my posts are inferential as they are based on evidence and (hopefully) well-thought out reasoning.
I disagree it is subjective because two microphones mounted either side of a head shaped ball will result in remarkably similar effects. Differences between us all introduce variations for sure, but the basic premise for the dip will remain.
It is symptomatic of the entire cause of the loss of imaging in stereo, namely an insufficient number of information channels.
It compensates for a three-dimensional acoustic event being replayed via two or more loudspeakers in front of the listener - and specifically for how laterally encoded information is presented.
My argument is that a single objective measure to describe how deep the dip should be would be dependent on the source positions and directivities, the recording space, the microphone arrangement, the loudspeaker arrangement and to some extent the listening environment also. It might well be that an optimal dip would vary throughout a recording - which would render it non-optimal.
But nevertheless, I maintain the dip has an objective basis, just one that has not been fully characterised yet.
"I will take that as a compliment. All of my posts are inferential as they are based on evidence and (hopefully) well-thought out reasoning."
Evidence itself negates the need for inference, but we make make one from it.
"I disagree it is subjective because two microphones mounted either side of a head shaped ball will result in remarkably similar effects. Differences between us all introduce variations for sure, but the basic premise for the dip will remain."
The relationship of the ears to the head also applies in normal hearing perception.
"It is symptomatic of the entire cause of the loss of imaging in stereo, namely an insufficient number of information channels."
Symptomatic, yes, but maybe not a complete analysis, but I had hoped for a clear piece by piece mechanistic analysis of how, if it is so, it happens.
"My argument is that a single objective measure to describe how deep the dip should be would be dependent on the source positions and directivities, the recording space, the microphone arrangement, the loudspeaker arrangement and to some extent the listening environment also. It might well be that an optimal dip would vary throughout a recording - which would render it non-optimal.
But nevertheless, I maintain the dip has an objective basis, just one that has not been fully characterised yet."
Pretty well agree with that.
Evidence itself negates the need for inference, but we make make one from it.
"I disagree it is subjective because two microphones mounted either side of a head shaped ball will result in remarkably similar effects. Differences between us all introduce variations for sure, but the basic premise for the dip will remain."
The relationship of the ears to the head also applies in normal hearing perception.
"It is symptomatic of the entire cause of the loss of imaging in stereo, namely an insufficient number of information channels."
Symptomatic, yes, but maybe not a complete analysis, but I had hoped for a clear piece by piece mechanistic analysis of how, if it is so, it happens.
"My argument is that a single objective measure to describe how deep the dip should be would be dependent on the source positions and directivities, the recording space, the microphone arrangement, the loudspeaker arrangement and to some extent the listening environment also. It might well be that an optimal dip would vary throughout a recording - which would render it non-optimal.
But nevertheless, I maintain the dip has an objective basis, just one that has not been fully characterised yet."
Pretty well agree with that.
I used the definition of inferential from a dictionary. Evidence does not negate the need for inference, however, as we infer knowledge from what we perceive to be evidence. But that is semantic.
Yes. And the dip is used because the reception of stereo information at a listener's ears is different to that which would have been presented to the listener had they been sat at the microphone array position.
The most simple treatise I have come across is that of Edeko et al in their JAES paper. I still regard this as somewhat off-topic, if related.
Essentially our hearing is less sensitive to laterally originating sound sources in the frequency band in question. When such energy is encoded in a stereo recording and then replayed from (predominantly) the front, the reproduction will appear overly-bright. Hence a dip compensates for the apparent excess energy in this band (to a degree dependent on all the factors I have described previously).
in chapter 9.1.3 in tooles book 'sound reproduction' he describes a simple experiment with pink noise, in short you can find the listening sweet spot by simply listen to when the pink noise sound is dullest, moving even so slightly left or right of the sweet spot causes the sound to get audible brighter. this is simply stereo as it is flawed. the cause of the dullness in the phantom center image is destructive acoustical interference
is it not this dullness we might want to compensate for?
is it not this dullness we might want to compensate for?
Are the loudspeakers in phase or out of phase?
In-phase (M) will demonstrate the collapse of the centre image at multiple frequencies.
Out-of-phase (S) will appear brighter than would the same signal developed from a genuinely laterally positioned sound source.
As said above, the two effects are mathematically related, but they are not the same effect. Using three speaker stereo can remove the in-phase problems, but will leave the out-of-phase issue - and therefore the dip might still be applicable.
"Yes. And the dip is used because the reception of stereo information at a listener's ears is different to that which would have been presented to the listener had they been sat at the microphone array position."
That I am very much in agreement with, and it is entirely consistent with my earlier post;
"That last excerpt seems to clarify one thing at least, so now the question is clearly why, having eliminated the concerns about indirect sound, the direct sound should need to be adulterated. But I can see clearly that recordings done with near field mics. may be too bright, because the mics. pic up more presence and top. (Used by Previn and praised.)"
That I am very much in agreement with, and it is entirely consistent with my earlier post;
"That last excerpt seems to clarify one thing at least, so now the question is clearly why, having eliminated the concerns about indirect sound, the direct sound should need to be adulterated. But I can see clearly that recordings done with near field mics. may be too bright, because the mics. pic up more presence and top. (Used by Previn and praised.)"
Now my instincts are again befuddled soundbloke.
Because;
The BBC used a mono signal for testing stereo quality, which surely must have been in phase, Yet you are saying that a mono signal of pink noise will exhibit a collapsed stereo image.
You are also saying that an out of phase signal will appear brighter then that developed form a genuinely laterally positioned sound source.
These both seem to me to be counter intuitive given the nature of constructive and destructive interference.
Celef, was the pink noise coherent?
Because;
The BBC used a mono signal for testing stereo quality, which surely must have been in phase, Yet you are saying that a mono signal of pink noise will exhibit a collapsed stereo image.
You are also saying that an out of phase signal will appear brighter then that developed form a genuinely laterally positioned sound source.
These both seem to me to be counter intuitive given the nature of constructive and destructive interference.
Celef, was the pink noise coherent?
Because lateral information in the recording is replayed from a predominantly frontal location. It therefore appears brighter than it was if you were sat at the microphone array. Hence the frequency balance is altered to try and compensate. Obviously it is incorrect for a centre-front sound, so the dip will be a compromise dependent largely on the recording sources and venue.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.