Re: First I've gotta say -
Ken,
I can't see why he has to choose in this way. It seems to me that different people approach music and HT in different ways. What one seeks in music, another may look for in HT. I would say rather, that he should choose his goals and their level of priority. This may in fact be more helpful than putting HT or music as priority.
For HT one person may be happy with 100 db SPL with clear vocals and authoritative bass down to 30 Hz. Another may want bass down to 20 Hz with clean output anywhere in the room to 110 db SPL.
Thanks for the suggestion. Perth is right across the other side of Australia, but if there was someone in Melbourne who has one I'd certainly be interested in hearing it.
Most HE drivers I have looked at don't have a very flat response and are very expensive. They also tend to look unappealing. It seems to me that if you take a HE woofer, filter the response to be as flat as a hifi woofer, what you have is a more expensive ugly looking speaker that is only more efficient due to its size. For the same cost and with less difficulty, I suspect you would be able to use a number of hifi woofers to get the same output and efficiency. This would depend greatly on the drivers in question of course.
As far as horns go I can't speak from experience in a home environment. My inclination at this point is to look for other ways to get high output, given what I have heard. I would like to hear some good DIY horns that might change my mind ...
Thanks for the reference. I had a look at two of the JBL PDF files with some interest. One thing that surprised me was the minimum subwoofer recommendations! A single JBL 18" subwoofer for a space of 2500 m2!!! As far as LFE is concerned, many HT enthusiasts have considerably more output in their living room. A high excursion 18" PA subwoofer driver has about 3L of VD which is about double that of the Shiva, most have in fact considerably less. A typical HT room may be say 25 m2.
Perhaps comercial cinemas are less concerned with the output at 20 Hz than many DIY HT enthusiasts ....
This discussion so far has led me to consider new ways to balance fidelity and high output. The following approach appeals to me most at the moment:
1. Dome tweeter fitted with waveguide for constant directivity and increased efficiency and output, possibly multiple tweeters for higher output in an array
2. Open baffle dipole mids using 8" hifi drivers similar to the Linkwitz Phoenix, possibly in a small array to cover the midrange for vocal clarity and near constant directivity. Open baffle speakers are actually very efficienct if the rear wave is not damped above the fequal cut-off where the first order rolloff begins due to cancellation.
3. Lab Horns for the bass, or if possible, larger 20 or 25 Hz basement horns.
Such a design would have very high output, it would be very dynamic and have exceptional clarity.
Ken L said:
Ken,
I can't see why he has to choose in this way. It seems to me that different people approach music and HT in different ways. What one seeks in music, another may look for in HT. I would say rather, that he should choose his goals and their level of priority. This may in fact be more helpful than putting HT or music as priority.
For HT one person may be happy with 100 db SPL with clear vocals and authoritative bass down to 30 Hz. Another may want bass down to 20 Hz with clean output anywhere in the room to 110 db SPL.
Ken L said:
Thanks for the suggestion. Perth is right across the other side of Australia, but if there was someone in Melbourne who has one I'd certainly be interested in hearing it.
GM said:
Most HE drivers I have looked at don't have a very flat response and are very expensive. They also tend to look unappealing. It seems to me that if you take a HE woofer, filter the response to be as flat as a hifi woofer, what you have is a more expensive ugly looking speaker that is only more efficient due to its size. For the same cost and with less difficulty, I suspect you would be able to use a number of hifi woofers to get the same output and efficiency. This would depend greatly on the drivers in question of course.
As far as horns go I can't speak from experience in a home environment. My inclination at this point is to look for other ways to get high output, given what I have heard. I would like to hear some good DIY horns that might change my mind ...
GM said:
Thanks for the reference. I had a look at two of the JBL PDF files with some interest. One thing that surprised me was the minimum subwoofer recommendations! A single JBL 18" subwoofer for a space of 2500 m2!!! As far as LFE is concerned, many HT enthusiasts have considerably more output in their living room. A high excursion 18" PA subwoofer driver has about 3L of VD which is about double that of the Shiva, most have in fact considerably less. A typical HT room may be say 25 m2.
Perhaps comercial cinemas are less concerned with the output at 20 Hz than many DIY HT enthusiasts ....
This discussion so far has led me to consider new ways to balance fidelity and high output. The following approach appeals to me most at the moment:
1. Dome tweeter fitted with waveguide for constant directivity and increased efficiency and output, possibly multiple tweeters for higher output in an array
2. Open baffle dipole mids using 8" hifi drivers similar to the Linkwitz Phoenix, possibly in a small array to cover the midrange for vocal clarity and near constant directivity. Open baffle speakers are actually very efficienct if the rear wave is not damped above the fequal cut-off where the first order rolloff begins due to cancellation.
3. Lab Horns for the bass, or if possible, larger 20 or 25 Hz basement horns.
Such a design would have very high output, it would be very dynamic and have exceptional clarity.
Accurate reproduction is accurate reproduction. The only difference is that HT has set requirements as to what consitutes it. Music is much more subjective, and having listened to some abominations that other people thought were great music speakers, I'll go with the HT speakers.
I can understand the anti-HT bias, a lot of the speakers commonly used are designed with output and output alone in mind.
I can understand the anti-HT bias, a lot of the speakers commonly used are designed with output and output alone in mind.
>GM,
My mistake, I read it a couple of weeks ago and my memory is none too good.
====
Neither is mine, and why I try to check my facts before responding.
====
>But back to my point. Peak dialogue intelligiblity is at around 80dB IIRC.
====
Never heard of this, got any references? As I noted, we are ampitude oriented, so S/N ratio and limited BW is what determines intelligibility in the research I've read. Indeed, maximum speech intelligibility requires that the response above 2kHz rolls off 3dB for each octave, i.e. -3dB at 4kHz; -6dB at 8kHz; -9dB at 16kHz. The response from 125Hz to 2kHz is flat with a window of ±2dB. Frequencies below 125Hz are rolled off at 6dB/octave to minimize rumble/boominess.
On several occasions on various forums I've hypothezied that a CC with this FR might be the best for a 'fill' to 'anchor' the screen sound when in phantom or 2.1 mode, but AFAIK no one has tried it yet.
====
>Imagine what dialogue would be like closer to the front of the theatre if it was reproduced at 85dB 2/3 of the way back. Unpleasant would be putting it mildy.
====
I don't have to, I've experienced it often enough at the cinema down the street.
Can't vouch for cinemas in Merry ol' England, but around here the horns are medium throw and aimed at the 2/3 point so you're progressively more off axis as you move forward. IOW, it's acceptable.
====
>Also consider that 75dB requires 1/10th the power of 85dB, and this is nearly continuous power.
====
True, but how is it relevant to the discussion?
====
>On the subject of dynamic range, in digital sources that is theoretically determined by the sample bit depth if I'm not mistaken, and DD is normally sampled at 20Bits per sample, giving 12dB more range than 16bit PCM (if I've done the maths right). Of course, depending on the reconstruction filter you might lose some of that range.
====
But how does this make film soundtracks have wider dynamic range than vinyl/CDs?
====
>The dynamic range of the subwoofer channel in the home is 115dB if used solely for the LFE channel and 121dB if all bass is redirected to it.
====
~122dB actually, but what's a dB among friends?
====
>And I understand exactly why people have problems with that, and I think you do too if your sig is anything to go by. Loud is truly beautiful if it's clean Many people miss out on that.
====
Yeah, they do. Folks are always surprised at how much louder they listen to my much more efficient system than their own or other cone/dome ones.
====
>I'm happy to be proved wrong of course if anyone can dig up Dolby's mixing guidelines or anything else that contradicts me.
====
So go to their site and download it and the peripheral docs. I get so frustrated with these types of threads where folks don't bother to read up on the process, etc., before engaging keyboard.
GM
My mistake, I read it a couple of weeks ago and my memory is none too good.
====
Neither is mine, and why I try to check my facts before responding.
====
>But back to my point. Peak dialogue intelligiblity is at around 80dB IIRC.
====
Never heard of this, got any references? As I noted, we are ampitude oriented, so S/N ratio and limited BW is what determines intelligibility in the research I've read. Indeed, maximum speech intelligibility requires that the response above 2kHz rolls off 3dB for each octave, i.e. -3dB at 4kHz; -6dB at 8kHz; -9dB at 16kHz. The response from 125Hz to 2kHz is flat with a window of ±2dB. Frequencies below 125Hz are rolled off at 6dB/octave to minimize rumble/boominess.
On several occasions on various forums I've hypothezied that a CC with this FR might be the best for a 'fill' to 'anchor' the screen sound when in phantom or 2.1 mode, but AFAIK no one has tried it yet.
====
>Imagine what dialogue would be like closer to the front of the theatre if it was reproduced at 85dB 2/3 of the way back. Unpleasant would be putting it mildy.
====
I don't have to, I've experienced it often enough at the cinema down the street.
Can't vouch for cinemas in Merry ol' England, but around here the horns are medium throw and aimed at the 2/3 point so you're progressively more off axis as you move forward. IOW, it's acceptable.====
>Also consider that 75dB requires 1/10th the power of 85dB, and this is nearly continuous power.
====
True, but how is it relevant to the discussion?
====
>On the subject of dynamic range, in digital sources that is theoretically determined by the sample bit depth if I'm not mistaken, and DD is normally sampled at 20Bits per sample, giving 12dB more range than 16bit PCM (if I've done the maths right). Of course, depending on the reconstruction filter you might lose some of that range.
====
But how does this make film soundtracks have wider dynamic range than vinyl/CDs?
====
>The dynamic range of the subwoofer channel in the home is 115dB if used solely for the LFE channel and 121dB if all bass is redirected to it.
====
~122dB actually, but what's a dB among friends?
====
>And I understand exactly why people have problems with that, and I think you do too if your sig is anything to go by. Loud is truly beautiful if it's clean Many people miss out on that.
====
Yeah, they do. Folks are always surprised at how much louder they listen to my much more efficient system than their own or other cone/dome ones.
====
>I'm happy to be proved wrong of course if anyone can dig up Dolby's mixing guidelines or anything else that contradicts me.
====
So go to their site and download it and the peripheral docs. I get so frustrated with these types of threads where folks don't bother to read up on the process, etc., before engaging keyboard.
GM
It's actually 74dB with correctly encoded material and a calibrated system.
Home systems are calibrated with -30dBFS pink noise to yield 75dB SPL at the seats and theatrical systems are calibrated with -20dBFS pin noise to produce 85dB - so 0dBFS produces 105dB SPL (Dolby Reference)
AC-3 encodes the level of dialog as a five bit quantity relative to 0dBFS : -31 to -1dBFS (0 is invalid).
Decoders then attenuate by 31dB - this level.
With 31dB of headroom you end up with 105dB - 0dB - 31dB = 74dB SPL. With 1dB of headroom you'd end up with 105 -1dB - 30dB = 74dB SPL.
Dolby's encoders default to 27dB of headroom beyond the dialog level, meaning you usually get peaks no higher than 101/111dB.
Sorry, I have read up on the whole reproduction of theatrical soundtrack in the home issue, just not for awhile. If you want I will go back and dig up all the relevant notes if you want to read them. Very long list though. And yes, they don't all agree before you ask.GM said:
It's kind of a moot point though, as most reproduction systems don't come close to the amplitude experienced in a theatre without much higher distortion. What is needed is the same amplitude with lower distortion. On that I think we can all agree.
Mudge said:
Recently I purchased Dr Earl's book (referenced at the start of the thread) on Home Theatre design and it addresses some issues that you won't find in most sources as most available material is on large cinemas, not a home environment.
Some of the differences include:
- larger surface to volume ratio in a small room
- small rooms are more acoustically dead
- small rooms have room mode problems below the shroeder frequency which makes accurate bass reproduction much more difficult
Mudge said:
I find that I like the sound of my speakers better than the sound of a cinema. This is even without my subwoofers. However, one thing I'd like to achieve is more output and lower distortion, this is certain.
I think open baffle dipoles actually have a lot of potential for high output. They are 6db more efficient than a monopole above fequal. They are known for reduced output when the bass is also dipole. If mated with horn loaded bass with an array of tweeters and midrange drivers, they can be designed for extreme high output. This appeals to me as usually efficiency is increased by:
- increasing radiating area
- waveguides and horns
Dipoles achieve this without increasing the number of drivers or doing anything which adds coloration, unlike horns. Considering this it is a surprise that it is not more commonly used, or that when open baffle is used, that efficiency is not usually stated as one of the advantages.
Something else to consider in doing an array is that you'll pick up another 6db of efficiency at a 4M listening distance in addition to the overall increased efficiency due to additional drivers and the almost 3db increase by going dipole. This is due to the line array effect and output decreasing by only 3db per doubling of distance instead of 6db with a single speaker. This benefit is almost never discussed in a home application.
Another benefit of arrays is that there are less reflections to deal with. Ceiling and floor reflections are insignificant because the sound is focused in the vertical plain of the array. This will be a problem if you have a center channel, because with horizontal mounting a center channel array will need to be as wide as the seating area.
Another benefit of arrays is that there are less reflections to deal with. Ceiling and floor reflections are insignificant because the sound is focused in the vertical plain of the array. This will be a problem if you have a center channel, because with horizontal mounting a center channel array will need to be as wide as the seating area.
That depends upon distance and frequency. All finite arrays begin to act like a point source at some distance. In a typical room a bigger concern for you will be more of how tall do you want your listening area to be. eg If your array is only 4ft tall, which would be fine for sitting, but when you stand up with your ears above the plain of the array, it's like someone pushed the mute button. It's not silence, but the difference is dramatic, so if you want music for parties or have a tiered seating area in your HT, you want your arrays to be tall. Also, smaller drivers are better and need to be as close together as possible, because the line array effect is dependent upon wavelength and spacing of the sound sources.
This is the chart from Jim Griffin's white paper, with some grid lines added for clarity.
An externally hosted image should be here but it was not working when we last tested it.
Near Field Line Array White Paper
Near Field Line Array White Paper
Paul and Others,
The questions about near field extent (3 dB decrease per doubling of distance from the array), height of the array, and much more are addresed in my white paper.
Jim
Near Field Line Array White Paper
An externally hosted image should be here but it was not working when we last tested it.
Paul and Others,
The questions about near field extent (3 dB decrease per doubling of distance from the array), height of the array, and much more are addresed in my white paper.
Jim
Note what the chart above implies for relatively short arrays. At average listening distances, you will be in the near field for high frequencies and the far field for low frequencies. That means you will experience a shift in the tonal balance as you move front to back. The highs fall off at 3dB as you move away but the lows fall off at 6dB. I don't see this 'problem' discussed much but it's very real.
While I appreciate the theory, but I don't believe that chart to be even close to accurate. At the nearfield/farfield change the dispersion changes too. I built a pair of 5ft line arrays (12 4" drivers on 5" centers) for a local bar and I can tell you for a fact that at 30ft there is still a clear boundary for being inside the plain of the array or outside of it and it's not anywhere near a 3 or 4Khz cutoff. Also, I noticed no gross change in tonal balance at that distance. Maybe that's supposed to be meters in distance, not feet.
Now I'm going to have to go over there and measure for myself using a tone generator and SPL meter, because I like arrays and have plans to build more.
While I respect that Dr. Griffin is vastly more knowledgeable than I in this area, there must be something to reconcile my real world experience with the information in his white paper.
Now I'm going to have to go over there and measure for myself using a tone generator and SPL meter, because I like arrays and have plans to build more.
While I respect that Dr. Griffin is vastly more knowledgeable than I in this area, there must be something to reconcile my real world experience with the information in his white paper.
Right. The array will start to "beam" vertically where the wavelength is shorter than the array -- somewhere in the 200-300 Hz range for a 5 ft. array.
A check with pink noise and an RTA will be more revealing than an SPL meter. Start at 5 ft., walk away from the speaker, and see what happens. I could hear a change in tone with a Linus array I had a chance to listen to but YMMV.
Comments on Line Arrays
On the near to far field transition distance raised in the above posts one should realize that we are not talking of a black and white issue. Around the near/far field transition distance you will have progressively different fall off rates as you go from the 3 dB per doubling of distance decline to the 6 dB per doubling of distance decrease rate. This is also the region wherein the greatest variation in amplitude versus distance occurs (see the undulating nature of the response versus distance plots in some of the JBL plots which are referenced in my white papers). Bottom line for a line array in your home you should design to have it operate in the near field only, i.e., extend the transition distance outward from the source to cover the listening area.
Additionally, any low frequency limitation of a line array is often migiated by coupling to the room boundaries--especially the floor and ceiling. You will realize a mirror image extention of the line's height in a room that allows extend the line's active height if the array is within a couple of feet of the ceiling or floor. Effectively, the equivalent line height becomes triple its length or 3X. Look at the distance plot with this 3 times multiplier in mind to get a sense for what the low frequency extension would be. For example, a 6 feet high array would respond like an 18 feet high array in the room if the ceiling and floor are within 1-2 feet of the ends of the line. Bottom line is that the low frequency extension of a line array is a room dependent variable.
On the near to far field transition distance raised in the above posts one should realize that we are not talking of a black and white issue. Around the near/far field transition distance you will have progressively different fall off rates as you go from the 3 dB per doubling of distance decline to the 6 dB per doubling of distance decrease rate. This is also the region wherein the greatest variation in amplitude versus distance occurs (see the undulating nature of the response versus distance plots in some of the JBL plots which are referenced in my white papers). Bottom line for a line array in your home you should design to have it operate in the near field only, i.e., extend the transition distance outward from the source to cover the listening area.
Additionally, any low frequency limitation of a line array is often migiated by coupling to the room boundaries--especially the floor and ceiling. You will realize a mirror image extention of the line's height in a room that allows extend the line's active height if the array is within a couple of feet of the ceiling or floor. Effectively, the equivalent line height becomes triple its length or 3X. Look at the distance plot with this 3 times multiplier in mind to get a sense for what the low frequency extension would be. For example, a 6 feet high array would respond like an 18 feet high array in the room if the ceiling and floor are within 1-2 feet of the ends of the line. Bottom line is that the low frequency extension of a line array is a room dependent variable.
I forgot to mention that these were OB arrays. Maybe that had some effect. Regarding beaming, I don't think that was the issue because I didn't notice a big change in off axis response with a change in distance. The sloping roof starting high where the speakers were located may have affected the response as well. I guess the bottom line is that I need to go do some more critical listening and in the future build taller arrays for big areas.
This was a specific application for ambiance music and speakers for the projector screen used mostly for watching soccer games in a large half covered area 40ft deep by 70ft long. An OB array was my logical choice for clear dialogue out into the open area without being so much louder up close. Their effective response is 150-10Khz.
The results exceeded my expectations with his existing speakers adding nice bass fill and and the arrays producing very clear vocals and dialogue even out in the open area. The bar owner was very happy with the results as well. Of course some cheap 2 way PA speakers that he already had were just no competition for an OB line array even using $1.50 ea drivers.
This was a specific application for ambiance music and speakers for the projector screen used mostly for watching soccer games in a large half covered area 40ft deep by 70ft long. An OB array was my logical choice for clear dialogue out into the open area without being so much louder up close. Their effective response is 150-10Khz.
The results exceeded my expectations with his existing speakers adding nice bass fill and and the arrays producing very clear vocals and dialogue even out in the open area. The bar owner was very happy with the results as well. Of course some cheap 2 way PA speakers that he already had were just no competition for an OB line array even using $1.50 ea drivers.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.