Obviously, I have missed something... I have thought a 'db' shows a ratio. How to devide pressure by energy? 🙂Kuei Yang Wang said:
OK, if not be a pedant, I understand it this way: if we increase omni sound source power DI times, we will get the same SPL as at case
- without icreasing a source power,
- with given acoustic source design,
- at this design axis.
Is it valid interpretation?
If valid: DI tells about integral feature (like, say, THD does). To have a whole picture, we must have SPL=SPL(frequency, vertical angle, horizontal angle) function (for given radius). So, for example, for given frequency, DI = 4 * pi * SPL_at_axis / (polar integral of cited function). Is it so? (I have omited converting to db)
Konnichiwa,
Not really.
A true omni has a DI of 0db, meaning if you measure the SPL on the design axis and you measure in any other direction you measure the same.
A dipole for example has a DI of 4.8db meaning it radiates in total, integrated across the whole 360 degree spherical radition angle 4.8db LESS sound than on axis or than a true Omni Directional radiator.
A typhical "HiFi" speaker has a DI of 0db at low frequencies and if have the usual 1" tweeter dome and a 6" Woofer/Midrange the DI varies wildly across the midrange and eventually hits well over 10db at high frequencies (10KHz +).
I am not sure if that helps, but the key to good and realistic sound is to not only have a reasonably high DI (I prefer even more than the 4.8db given by a dipole) but also to avoid drastic changes in DI and to ideally keep it constant to at least around 5KHz and above at least 250Hz with some increase in DI towards high frequencies being allowed.
Sayonara
anli said:
Not really.
A true omni has a DI of 0db, meaning if you measure the SPL on the design axis and you measure in any other direction you measure the same.
A dipole for example has a DI of 4.8db meaning it radiates in total, integrated across the whole 360 degree spherical radition angle 4.8db LESS sound than on axis or than a true Omni Directional radiator.
A typhical "HiFi" speaker has a DI of 0db at low frequencies and if have the usual 1" tweeter dome and a 6" Woofer/Midrange the DI varies wildly across the midrange and eventually hits well over 10db at high frequencies (10KHz +).
I am not sure if that helps, but the key to good and realistic sound is to not only have a reasonably high DI (I prefer even more than the 4.8db given by a dipole) but also to avoid drastic changes in DI and to ideally keep it constant to at least around 5KHz and above at least 250Hz with some increase in DI towards high frequencies being allowed.
Sayonara
Kuei Yang Wang,
I still don't realize all the "from DI-related suggestions to concrete design" chain in practical way. Two days ago I have got three Beyma pairs (for 3-way speakers), and now is under thinking about design (at any case LF sections will be separated from mid and HF). Can you give some practical DI-related tips "by example"? If you will (I hope, it is so, as you love pro speakers 🙂), I'll list Beyma models.
I still don't realize all the "from DI-related suggestions to concrete design" chain in practical way. Two days ago I have got three Beyma pairs (for 3-way speakers), and now is under thinking about design (at any case LF sections will be separated from mid and HF). Can you give some practical DI-related tips "by example"? If you will (I hope, it is so, as you love pro speakers 🙂), I'll list Beyma models.
Konnichiwa,
Well, first of all you MUST discard what you think you know about speaker design. You then MUST consider the off-axis response as relevant and essential as the on axis response.
To give you a design example:
The crossover points are 550Hz and 2.8KHz. The 1" tweeter contains a "phaseplate" to widen it's dispersion at high frequencies and somewhat inset into a small short waveguide, narrowing dispersion at lower frequencies.
The 5" Midrange is mounted as dipole with some extra absorbtion behind the driver, together with the rear radiation reflecting delayed and attenuated (absorbtive material) back makes the whole radiation fairly directional.
The 15" Driver itself "beams" quite nicely above 300Hz and thus above 300Hz we have a deliberatly controlled dispersion aimed at making the off axis frequency response as flat as possible also.
This bit of the design is actually quite antique by now, it was started by MEG in the Mid 1980's. The result here with the on axis and 45Degree horizontal frequency response (note the vertical scale BTW and the fact that not much smoothing is applied to the curves:
Here for comparison the DI vs. frequency curve:
With that you can also understand somewhat the relation between DI and FR....
At 45 degrees we find that above around 250Hz the 45 degrees off axis response is evenly attenuated by around 3db and that overall in this range the speaker radiates around 9db less sound power into the room overall compared to an equal on axis frequency response omnidirectional speaker or indeed to say that ratio between room reverbrant sound and direct sound is 9db more weighted towards the recording and not the room, compared to a true omni.
Helps?
Sayonara
anli said:
Well, first of all you MUST discard what you think you know about speaker design. You then MUST consider the off-axis response as relevant and essential as the on axis response.
To give you a design example:
An externally hosted image should be here but it was not working when we last tested it.
The crossover points are 550Hz and 2.8KHz. The 1" tweeter contains a "phaseplate" to widen it's dispersion at high frequencies and somewhat inset into a small short waveguide, narrowing dispersion at lower frequencies.
The 5" Midrange is mounted as dipole with some extra absorbtion behind the driver, together with the rear radiation reflecting delayed and attenuated (absorbtive material) back makes the whole radiation fairly directional.
The 15" Driver itself "beams" quite nicely above 300Hz and thus above 300Hz we have a deliberatly controlled dispersion aimed at making the off axis frequency response as flat as possible also.
This bit of the design is actually quite antique by now, it was started by MEG in the Mid 1980's. The result here with the on axis and 45Degree horizontal frequency response (note the vertical scale BTW and the fact that not much smoothing is applied to the curves:
An externally hosted image should be here but it was not working when we last tested it.
Here for comparison the DI vs. frequency curve:
An externally hosted image should be here but it was not working when we last tested it.
With that you can also understand somewhat the relation between DI and FR....
At 45 degrees we find that above around 250Hz the 45 degrees off axis response is evenly attenuated by around 3db and that overall in this range the speaker radiates around 9db less sound power into the room overall compared to an equal on axis frequency response omnidirectional speaker or indeed to say that ratio between room reverbrant sound and direct sound is 9db more weighted towards the recording and not the room, compared to a true omni.
Helps?
Sayonara
Altec....
Moin,
another hint for such great engineering science in the past are the specs for the Altec Lansing VOT A7 and the electrical twin brother Valencia.
You can find the specifications here:
http://www.audioheritage.org/
Carsten
Moin,
another hint for such great engineering science in the past are the specs for the Altec Lansing VOT A7 and the electrical twin brother Valencia.
You can find the specifications here:
http://www.audioheritage.org/
Carsten
Kuei Yang Wang,Kuei Yang Wang said:
Thanks, helps! 🙂 I see my speakers choice (18" woofer, 6.5 mid and CP21F) is very suitable to follow your suggestions. The only thing to experiment with more closely is rather wide tweeter horisontal dispersion. Probably some little dumping 'pillow' left and right from tweeter mouth is a way to try to avoid too small DI at HF. Or, as another experiment, it is possible to rotate a tweeter pi/2.
Re: Re: Re: Re: What exactly is beaming and why does it matter?
But any way you do it you have to make comprimises. Do you see any advantage to wide dispersion up to about 1 Khz or so?
You left out that part in your post (crossfiring). That's why I said "seems like", without having your speakers (at least the tweeters) setup crossfiring, my first reply is quite true.Kuei Yang Wang said:
But any way you do it you have to make comprimises. Do you see any advantage to wide dispersion up to about 1 Khz or so?
Re: Re: Re: Re: Re: What exactly is beaming and why does it matter?
Konnichiwa,
It is an example which allowes this to be done using speakers with symetrical radiation patterns, the old JBL Everest used less toe-in and horizontally an asymetric radiation pattern. Ever which way, a wide area of stable imaging makes controlled directivity speakers essential, together with suitable setup.
Depends on the application. For the reproduction of musical recordings either minimally miked or artificially created in the laboratory, in domestic settings, in the far field? Absolutely no.
But that's just me. Some people like the "they are here" feeling omni's and related devices give. I don't.
Sayonara
Konnichiwa,
critofur said:
It is an example which allowes this to be done using speakers with symetrical radiation patterns, the old JBL Everest used less toe-in and horizontally an asymetric radiation pattern. Ever which way, a wide area of stable imaging makes controlled directivity speakers essential, together with suitable setup.
critofur said:
Depends on the application. For the reproduction of musical recordings either minimally miked or artificially created in the laboratory, in domestic settings, in the far field? Absolutely no.
But that's just me. Some people like the "they are here" feeling omni's and related devices give. I don't.
Sayonara
Well it would seem to me that if you wanted a recording of a cello to sound as if it was playing right there in your house that an omni speaker would be the only way to accomplish that. It does seem that most often people rather prefer to have the illusion that they are listening to music being played in the venue where it was recorded.
Using good speakers designed and setup the way you are suggesting can be nice, but, unless you use some fancy DSP, does it really fool you into thinking you're also hearing the reflected sounds that you would experience in that venue, as if those reflections were coming from the proper places?
Are you saying that's the area where you prefer to make comprimises for the sake of other aspects of the sound?
Only once in my life have I heard a speaker that could make it sound as though a trumpet was actually playing right there in front of us. We were listining to a particular uncomon type of omni speaker in a very large room not close to any walls, just one speaker (so obvoiusly mono).
Using good speakers designed and setup the way you are suggesting can be nice, but, unless you use some fancy DSP, does it really fool you into thinking you're also hearing the reflected sounds that you would experience in that venue, as if those reflections were coming from the proper places?
Are you saying that's the area where you prefer to make comprimises for the sake of other aspects of the sound?
Only once in my life have I heard a speaker that could make it sound as though a trumpet was actually playing right there in front of us. We were listining to a particular uncomon type of omni speaker in a very large room not close to any walls, just one speaker (so obvoiusly mono).
Konnichiwa,
Not quiet. A cello is not quiet omnidirectional...
So it seems.
This works largely due to the large arrival time differences in real concert halls and the way we humans hear spatially (virtual placement vs. level/time difference).
I am not sure how you conclude this. By replacing the recording venues acoustics with reverbration in the room you are not improving things, you introduce an effect. If you like it, fine, but it remains an effect.
But how a speaker/system that sounded as if you where in the concert hall in front of the orchestra? That is what interrests me both on the recording and playback side.
Sayonara
critofur said:
Not quiet. A cello is not quiet omnidirectional...
critofur said:
So it seems.
critofur said:
This works largely due to the large arrival time differences in real concert halls and the way we humans hear spatially (virtual placement vs. level/time difference).
critofur said:
I am not sure how you conclude this. By replacing the recording venues acoustics with reverbration in the room you are not improving things, you introduce an effect. If you like it, fine, but it remains an effect.
critofur said:
But how a speaker/system that sounded as if you where in the concert hall in front of the orchestra? That is what interrests me both on the recording and playback side.
Sayonara
Not entirely omnidirectional, but certainly less like a speaker which is narrow dispersion on both horiz. and vert. axis. One of the important aspects of cello to me is feeling (not just hearing) freq. which are somewhat higher than what you normally consider to be in that range where you feel it. There must be a high dynamic range to the cello. I like to feel the sound in my skin, my hair, in the floor and the arms of the chair I'm sitting in (as you do if you sit close to a someone playing cello). I do not want to have to use an excessive average SPL to achieve this.Kuei Yang Wang said:
Hmm, well certainly some of the audio information recorded was affected by the venue in which it was recorded... But, that aspect was so overshadowed by the previously unexperienced vibrancy and natural sounding timbre that it wasn't really noticed at the time. The CD we listened to was Miles Davis - "Kind Of Blue" and we felt suddenly, here is Miles playing in front of us...
Do you like recordings made with mics on either side of a wooden sphere, or where the "ears" on a manequin head are?
Most loudspeakers certainly do not have narrow dispersion, that's kind of the point of this thread.
Dynamic range is the difference between soft and loud sounds. It has nothing to do with the play(back) mechanism.
Your question is - could a loudspeaker ever reproduce the feeling of sitting next to a cello being played? I hope not. Was she hot? I like the feeling that Mingus is playing in my living room - but never, in one million years, would that be the same as Mingus actually playing in my living room.
Kind of Blue sounds good on almost any system - it's acoustically quite simple. It's just a few instruments in a fairly dry studio (if memory serves). Good recordings are like that - they are more involving on many different types of systems (good and bad).
Omnidirectional speakers are giving you the maximum amount of acoustic signature of the room they're playing in - not the room they were recorded in or the instruments themselves. That can lean towards good or bad depending on the room, material and listener.
KYW can be unpleasant =) but he has done his homework. Listen to what he's saying and try to understand it. Or buy some PVC and build a Pluto and hang out with Miles all you like. No one will question you for enjoying it.
Peace.
Dynamic range is the difference between soft and loud sounds. It has nothing to do with the play(back) mechanism.
Your question is - could a loudspeaker ever reproduce the feeling of sitting next to a cello being played? I hope not. Was she hot? I like the feeling that Mingus is playing in my living room - but never, in one million years, would that be the same as Mingus actually playing in my living room.
Kind of Blue sounds good on almost any system - it's acoustically quite simple. It's just a few instruments in a fairly dry studio (if memory serves). Good recordings are like that - they are more involving on many different types of systems (good and bad).
Omnidirectional speakers are giving you the maximum amount of acoustic signature of the room they're playing in - not the room they were recorded in or the instruments themselves. That can lean towards good or bad depending on the room, material and listener.
KYW can be unpleasant =) but he has done his homework. Listen to what he's saying and try to understand it. Or buy some PVC and build a Pluto and hang out with Miles all you like. No one will question you for enjoying it.
Peace.
It seems like a main idea isn't to have narrow dispersion, rather not too wide, and more or less constant for critical (above 300Hz) range with smooth narrowing at HF. The aim is to have off axis FR without peaks and dips.
Indeed, suggested Directivity Index supposes _slightly more_ narrow dispersion rather common hi-fi speakers have at _some_ ranges, but I have not found any suggestions about "narrow dispersion" itself. At any case DI=5db can not be interpreted as "narrow dispersion"!
Indeed, suggested Directivity Index supposes _slightly more_ narrow dispersion rather common hi-fi speakers have at _some_ ranges, but I have not found any suggestions about "narrow dispersion" itself. At any case DI=5db can not be interpreted as "narrow dispersion"!
Konnichiwa,
What you write about strikes me as actually being a comment on dynamic behaviour, not the dispersion. This has much more to do withy the con struction of the drivers magnet system and voice coil.
Well, Miles Davis KOB (I have the original 6 Eye Vinyl and several reissues inculding the recent Sony CD) was a studio recording with more than two tracks. It was fairly dry (not much reverb in the studio).
Not particulary. It depends entierly on what is being recorded. For small ensembles the ORTF arrangement or indeed the Jeklin Disk can work well, but other arrangements can give better results. How you record depends on the size of the ensemble/orchestra, the room and finally the goals set by the producer/director/artist(s).
Sayonara
critofur said:
What you write about strikes me as actually being a comment on dynamic behaviour, not the dispersion. This has much more to do withy the con struction of the drivers magnet system and voice coil.
critofur said:
Well, Miles Davis KOB (I have the original 6 Eye Vinyl and several reissues inculding the recent Sony CD) was a studio recording with more than two tracks. It was fairly dry (not much reverb in the studio).
critofur said:
Not particulary. It depends entierly on what is being recorded. For small ensembles the ORTF arrangement or indeed the Jeklin Disk can work well, but other arrangements can give better results. How you record depends on the size of the ensemble/orchestra, the room and finally the goals set by the producer/director/artist(s).
Sayonara
Konnichiwa,
Yes. Ideally we would be able to achieve a dispersion of maybe 60 degrees horizontally and 20 degrees vertically (which is a DI of around 10db IIRC) flat across the frequency range. That failing we would like to approach this as closely as we can.
I propose a DI of much more than 5db, however readily achievable, with simple DIY methodes is a DI of 5db at low frequencies, increasing smoothly to 10db in the midrange and above, though at very high frequencies the DI is likely higher than that.
Sayonara
anli said:
Yes. Ideally we would be able to achieve a dispersion of maybe 60 degrees horizontally and 20 degrees vertically (which is a DI of around 10db IIRC) flat across the frequency range. That failing we would like to approach this as closely as we can.
anli said:
I propose a DI of much more than 5db, however readily achievable, with simple DIY methodes is a DI of 5db at low frequencies, increasing smoothly to 10db in the midrange and above, though at very high frequencies the DI is likely higher than that.
Sayonara
Kuei Yang Wang,
OK, from 5db to 10db - at any case it isn't narrow, I think 🙂
Kuei, are there some published results about using low (till zero) order xovers to smooth off axis responce between multiway speaker bands?
OK, from 5db to 10db - at any case it isn't narrow, I think 🙂
Kuei, are there some published results about using low (till zero) order xovers to smooth off axis responce between multiway speaker bands?
Konnichiwa,
Well, yes, 10db DI is not that narrow.
But a common "HiFi" speaker has a DI that varies from 0db (around 100Hz and down) to 16db (15KHz and up, if a 1" dome is used) and often that variation is not smooth either....
Nothing I am readily aware off. I have seen some listening tests that seemed to favour 3rd order Butterworth....
Sayonara
anli said:
Well, yes, 10db DI is not that narrow.
But a common "HiFi" speaker has a DI that varies from 0db (around 100Hz and down) to 16db (15KHz and up, if a 1" dome is used) and often that variation is not smooth either....
anli said:
Nothing I am readily aware off. I have seen some listening tests that seemed to favour 3rd order Butterworth....
Sayonara
Kuei,Kuei Yang Wang said:
OK. And you have a big experience and "healthy" own opinion 🙂 Do you think it is possible (and, may be, attractive) way to smooth off-axis responces via widing of a band with two speakers simultanious radiation?
Konnichiwa,
This seems akin to B&O's filler driver. I once reviewed a speaker (4-way, two ways with filler) that did this, it was rather good too.
Sayanara
anli said:
This seems akin to B&O's filler driver. I once reviewed a speaker (4-way, two ways with filler) that did this, it was rather good too.
Sayanara
- Status
- Not open for further replies.