Hello,
As I am not the one who introduced the "enveloppe theme" in this disussion, preferring for my own part to speak about the relationship between harmonics and fundamental,it seems that we agree for a large part.
(see my initial message : http://www.diyaudio.com/forums/showthread.php?postid=1827566#post1827566 )
It is on such subjects that psychoacoustic studies fail for a part. As, for frequencies lower than 150Hz, it is difficult to obtain a low phase distortion acoustic signal arriving at the listener ear due to the lack of loudspeaker having such beahviour, we still don't know alot about the relation between what we hear in the bass and what we can measure.
Best regards from Paris, France
Jean-Michel Le Cléac'h
As I am not the one who introduced the "enveloppe theme" in this disussion, preferring for my own part to speak about the relationship between harmonics and fundamental,it seems that we agree for a large part.
(see my initial message : http://www.diyaudio.com/forums/showthread.php?postid=1827566#post1827566 )
It is on such subjects that psychoacoustic studies fail for a part. As, for frequencies lower than 150Hz, it is difficult to obtain a low phase distortion acoustic signal arriving at the listener ear due to the lack of loudspeaker having such beahviour, we still don't know alot about the relation between what we hear in the bass and what we can measure.
Best regards from Paris, France
Jean-Michel Le Cléac'h
Elias said:
Hello Xpert,
In the quasi general case, the use of an equalizer will linearize only the frequency response but they'll make worst the group delay curve.
As an example I smile when I read that someone saying that he plan to reduce the rise of group delay of a horn near it's cut-off just using equalization...
That's easy to understand, high frequencies travel fast inside most filters (even high pass) and grossly low frequencies are delayed versus the high frequencies. If I would do some humour I'll say that the ability of equalizers to reverse time is very small. ;-)
When the order of the transfer function of the system became large (let's say over 2) it became quite impossible to linearize phase by minimum phase equalization.
Best regards from Paris, France.
Jean-Michel Le Cléac'h
In the quasi general case, the use of an equalizer will linearize only the frequency response but they'll make worst the group delay curve.
As an example I smile when I read that someone saying that he plan to reduce the rise of group delay of a horn near it's cut-off just using equalization...
That's easy to understand, high frequencies travel fast inside most filters (even high pass) and grossly low frequencies are delayed versus the high frequencies. If I would do some humour I'll say that the ability of equalizers to reverse time is very small. ;-)
When the order of the transfer function of the system became large (let's say over 2) it became quite impossible to linearize phase by minimum phase equalization.
Best regards from Paris, France.
Jean-Michel Le Cléac'h
I remember an article in HiFi News and Record Review back in the days when it was a magazine worth reading. (cira 2000 ish please bear with me, this is all from memory)
Keith Howard took a pair of B&W 802s (I think). He used a DSP algorithm to effectively phase equalise the speaker by processing the .WAV files used for audition. This was applied between reasonable limits, I think he "let the speaker go" below port resonance.
As we would expect, the two major points of departure from minimum phase behaviour were around the crossover points and around the bass reflex tuning. The published before and after graphs showed markedly improved phase behaviour.
The subjective results he reported included an improvement in bass speed and impact. It makes sense to me that the phase relationship between the LF and directly associated instrument harmonics plays are large part in the "quality" of the bass. Over-damped IB always sound speedier than mis-tuned reflexes in the bass. I believe this is largely down to more accurate phase tracking between the instrument fundamental and harmonics, as Lynn points out LF is, in theory, the easiest bit to do!
Keith Howard took a pair of B&W 802s (I think). He used a DSP algorithm to effectively phase equalise the speaker by processing the .WAV files used for audition. This was applied between reasonable limits, I think he "let the speaker go" below port resonance.
As we would expect, the two major points of departure from minimum phase behaviour were around the crossover points and around the bass reflex tuning. The published before and after graphs showed markedly improved phase behaviour.
The subjective results he reported included an improvement in bass speed and impact. It makes sense to me that the phase relationship between the LF and directly associated instrument harmonics plays are large part in the "quality" of the bass. Over-damped IB always sound speedier than mis-tuned reflexes in the bass. I believe this is largely down to more accurate phase tracking between the instrument fundamental and harmonics, as Lynn points out LF is, in theory, the easiest bit to do!
Hello,
Is there really any reason to go higher orders than 2, if we are thinking about bass? Higher orders than that will only bring problems (BR etc) as indicated earlier.
If we stay in second order system, phase linearisation is quite straight forward by using pole/zero cancellation and generating desired new poles and zeros. Can be done in analog domain too with one opamp like in Linkwitz transform. Well, it cannot make linear phase from DC, but what it does it pushes the group delay lower in freq below audibility range. It's only mathematics of linear transfer functions. Of course it can bring other problems like increased nonlinear distortion because drivers have to work harder after equalisation.
- Elias
Jmmlc said:
Is there really any reason to go higher orders than 2, if we are thinking about bass? Higher orders than that will only bring problems (BR etc) as indicated earlier.
If we stay in second order system, phase linearisation is quite straight forward by using pole/zero cancellation and generating desired new poles and zeros. Can be done in analog domain too with one opamp like in Linkwitz transform. Well, it cannot make linear phase from DC, but what it does it pushes the group delay lower in freq below audibility range. It's only mathematics of linear transfer functions. Of course it can bring other problems like increased nonlinear distortion because drivers have to work harder after equalisation.
- Elias
Hello,
Yes it helps, but it's not feasible to do in passive crossovers. Try delaying 20kHz signal for 10ms, for example. Phase linear systems only exist in digital domain, well but that is not news. Still, even you would have a phase linear speaker, you have the room reflections which will defeat your painfully earned victory over nonlinear phase.
- Elias
xpert said:
Yes it helps, but it's not feasible to do in passive crossovers. Try delaying 20kHz signal for 10ms, for example. Phase linear systems only exist in digital domain, well but that is not news. Still, even you would have a phase linear speaker, you have the room reflections which will defeat your painfully earned victory over nonlinear phase.
- Elias
Hello Elias,
I agree with most of what you wrote. I guess the (non)feasibility to design that kind of poles/zeros cancellation with passive components only and the mainstream use of bass-reflex enclosure is surely the major limit to their generalisation.
As an example at the moment only one of my numerous friends use a close enclosure to reproduce the bass register.
Best regards from Paris, France
Jean-Michel Le Cléac'h
I agree with most of what you wrote. I guess the (non)feasibility to design that kind of poles/zeros cancellation with passive components only and the mainstream use of bass-reflex enclosure is surely the major limit to their generalisation.
As an example at the moment only one of my numerous friends use a close enclosure to reproduce the bass register.
Best regards from Paris, France
Jean-Michel Le Cléac'h
Elias said:
john k... said:
John - I can sympathize with your point, to an extent, but the fact is that I WOULD consider myself an audiophile in any sense of the word as I see it. Since we seem to agree on the points regarding these esoteric aspects of audio, I think that it might be better for us to try and direct the "impressions" that people have into a more productive line of thought.
For example, the continued use of the term "speed" is a real problem until it has a definition, and I have seen none. (And contrary to popular belief it IS NOT intuity obvious! I have no idea what people mean by its use.) The use of Psuedo-psychoacoustics (when no data from valid blind testing is used) should be avoided as its not meaningful.
In this way, maybe, just maybe, we could have a discussion here that actually gets somewhere instead of the endless circles of "well I hear this" and "that PROVES thus".
Dr. Geddes, can you imagine of a mini lexicon of typical audiophile wording and its scientific (engineering or psycho acoustic) counterpart?
I.e. someone says ''strong sound stage'' and you say, ''the good quality of recorded stereo effect rendering by the playback system as a whole, room included''.
Speed? Slam? Rhythm and Timing? blah, blah...
I.e. someone says ''strong sound stage'' and you say, ''the good quality of recorded stereo effect rendering by the playback system as a whole, room included''.
Speed? Slam? Rhythm and Timing? blah, blah...
Yes, I think that there are good terms that have meaning and can be correlated with subjective.
Coloration - pretty clear I think
Dynamics - the ability to play loud and soft passages without audible compressions or any effects which vary with level
Imaging - the ability to clearly define the location of different instruments - the image should be unambiguous, an invarient with level or passage.
Stage width - this is more dependent on the recording and the room and sometimes stage width is in direct conflict with imaging, but its a valid term non-the-less
Spaciousness - the felling of space, the impression of the rooms dimensions. This can be on the recording or in the room itself, but generally the two differnt kinds are quite clearly differentiated, or should be in a good system.
Beyond this I don't know of any terms that have a concrete enough meaning to be useful in any seriuos discussion.
I MIGHT think that "speed" for instance is a combination of low colloration with good dynamics and spaciuosness. This, to me defines good bass. Some have called this "tight bass" when listening to my system. Thank God nobody has said that it was "fast" (to my face!) Most bass that I hear is colored and not spatial at all with poor imaging. Bass should be precisely located as to the player (mostly the attack which is much higher in frequency), but the notes must be uncolored with a long spatial quality of a "hanging" note that is omnipresent - not localised.
Coloration - pretty clear I think
Dynamics - the ability to play loud and soft passages without audible compressions or any effects which vary with level
Imaging - the ability to clearly define the location of different instruments - the image should be unambiguous, an invarient with level or passage.
Stage width - this is more dependent on the recording and the room and sometimes stage width is in direct conflict with imaging, but its a valid term non-the-less
Spaciousness - the felling of space, the impression of the rooms dimensions. This can be on the recording or in the room itself, but generally the two differnt kinds are quite clearly differentiated, or should be in a good system.
Beyond this I don't know of any terms that have a concrete enough meaning to be useful in any seriuos discussion.
I MIGHT think that "speed" for instance is a combination of low colloration with good dynamics and spaciuosness. This, to me defines good bass. Some have called this "tight bass" when listening to my system. Thank God nobody has said that it was "fast" (to my face!) Most bass that I hear is colored and not spatial at all with poor imaging. Bass should be precisely located as to the player (mostly the attack which is much higher in frequency), but the notes must be uncolored with a long spatial quality of a "hanging" note that is omnipresent - not localised.
Hello Earl,
How would you define the difference of sound between a bass-reflex and a closed enclosure all parameters (low frequency cut-off, bandwith, etc.) being similar?
If dynamics and spaciousness are the same, is it just coloration?
best regards from Paris, France
Jean-Michel Le Cléac'h
How would you define the difference of sound between a bass-reflex and a closed enclosure all parameters (low frequency cut-off, bandwith, etc.) being similar?
If dynamics and spaciousness are the same, is it just coloration?
best regards from Paris, France
Jean-Michel Le Cléac'h
gedlee said:
Jmmlc said:
Jean-Michel - thanks for the question
First, the effect has to be quantified as real. I personally have not noticed this distinction as seperate from room effects. What has to happen is that there is an wide agreement of a concrete reproducable effect that depends only on the subs, not the room.
In actuality I don't think that there is any way to setup two subs of a different design such that they excite the room the same. If this is the effect that you note then its a room characteristic and NOT a speaker characteristic. This would be my guess. In fact I would classify ALL LF perception as room effects and not source effects. That's because I personally pay little attention to the source types and characteristics, but I do set them all up the same way and they all come out fine. This is not only consistant with the theory of multiple subs, but it's what I find in practice. Call it coloration, or "fastness" or whatever you want.
It all comes down to a smooth transition from the mains to the subs (and this CANNOT be done with fixed crossover points, or without measurements, because the room changes everything), multiple subs spaced arround the room and made as flat as possible at several seating locations with EQ. Works every time.
Elias said:
on reflex vs closed:
gedlee said:
To be complete on that topic please consider that even bassreflex can be "pole shifted". In spite of representing one complex conjugated pole - the resonance - a reflex box shows two of that kind. You may shift each individually to Your needs.
so long
gedlee said:
I agree but I just don't believe they will be any success in changing impressions. After all, probably a large percentage of those who are reading this believe that they can hear things which can not be measured. The truth is that instrumentation today is far more sensitive than the human ear and if a difference can be heard it most certainly can be measured. This doesn't mean that there is a convenient set of test bench measurements that serve this purpose. But most certainly, if I change the amplifier in a system and measure the voltage and/or current applied to the speaker vs. time while playing musical passage, if a difference is heard there will most certainly be a difference between the voltage or current signals delivered by the amplifier.
When it comes down to impact or slam or speed of a woofer, let's talk about the fundamentals. Ultimately also sound is created by the transformation of mechanical motion in to pressure variations in the air in direct contact with the moving elements. For there to be a difference, there must be a differences in the motion. Add to that the effects of the room, with which I basically agree with Earl and you have a situation where the factors are the room and how it couples to the transfer function of the source's motion. For one woofer to have greater impact or speed there must be something different in the woofer transfer function. For example, the stability of the transfer function at large signal levels. This is in the realm of nonlinearity, but still something that would be easily measured.
John
You are better at this than you think. And a lot of people respect you, me too, I hope. If we can get together and explain things in a rational way, that makes sense, then I think that a lot of people will follow.
And I couldn't agree with you more that there is NO WAY that someone can hear something that I cannot measure. It's just not possible. That's not to say that I know how to measure it or how to put those measurements into a form that makes things clear, but I can say with absolute certainty that there is nothing in science that cannot be measured, even emotions. (The Japanese did a huge amount of work in this area years ago in regard to the emotional impact of a car. Interesting work, mostly a dead end for reasons that you can guess.)
You are better at this than you think. And a lot of people respect you, me too, I hope. If we can get together and explain things in a rational way, that makes sense, then I think that a lot of people will follow.
And I couldn't agree with you more that there is NO WAY that someone can hear something that I cannot measure. It's just not possible. That's not to say that I know how to measure it or how to put those measurements into a form that makes things clear, but I can say with absolute certainty that there is nothing in science that cannot be measured, even emotions. (The Japanese did a huge amount of work in this area years ago in regard to the emotional impact of a car. Interesting work, mostly a dead end for reasons that you can guess.)
From market share growth, I'm sure they did something right. Bear in mind that there are lots of people out there that like good audio equipment, but quality is not totally based on measurements. Look at how much time people spend on costmetics of speakers, which makes me wonder what Lynn's next speaker would look like.gedlee said:
