Setting up dipole surrounds. [Archive] - Audio & Video Forums

PDA

View Full Version : Setting up dipole surrounds.



Coldwater
02-25-2005, 05:08 PM
I just bought a pair of Paradigm cinema ADPs would like to know how to configure them in terms of distance and level settings. Do I set them up as if they were direct radiating?

cam
02-25-2005, 05:23 PM
For a di-pole to work the way it is intended you must put them to your sides up on the wall 3 to 4 feet above your ears at your listening position. Use a sound pressure meter to match the volume with your center and mains or do the best you can with out one. For your distance setting, measure from the speaker box to your listening position. Some people have put di-poles in back corners or on back walls because of limitations for mounting them in the correct position, but the best position is directly to your sides that way you will be in the null. Remember, the forward facing speakers are in phase and the rearward facing speakers are out of phase. If not put in the correct spot you could get some undesireable results.

RobotCzar
02-25-2005, 05:32 PM
I know you are not going to like this, but dipole speakers are not appropriate for multichannel setups. All (true multichannel) fomats I have heard about require direct radiators (and narrow ones at that). The old Dolby Pro Logic format had a diffuse single rear channel (often using two speakers) for which a dipole might make sense.

True multichannel systems rely on direct radiation from the speaker to form the ambient sound field, reflections will only muck up the plan. Does this stop people from using di and bipoles in multichannel setups? No. And don't expect dipole makers to admit their speakers are not ideal.

I suggest you consult the manual that came with your speakers. If a manual did not come with your speakers, then returning them might be the best plan.

Woochifer
02-25-2005, 06:29 PM
I know you are not going to like this, but dipole speakers are not appropriate for multichannel setups. All (true multichannel) fomats I have heard about require direct radiators (and narrow ones at that). The old Dolby Pro Logic format had a diffuse single rear channel (often using two speakers) for which a dipole might make sense.

True multichannel systems rely on direct radiation from the speaker to form the ambient sound field, reflections will only muck up the plan. Does this stop people from using di and bipoles in multichannel setups? No. And don't expect dipole makers to admit their speakers are not ideal.

I suggest you consult the manual that came with your speakers. If a manual did not come with your speakers, then returning them might be the best plan.

Yes and no. The issue is not nearly as cut and dry as you make it out to be. I agree with you about direct radiating speakers as the best option for discrete 5.1 setups, but it doesn't mean that dipoles are wholly without merit.

The viability of using dipoles depends on the source material, even if it's 5.1. In my listenings, dipoles can be the way to go if the sounds mixed into the original soundtrack are largely ambient in nature, and you don't have a lot of sounds that are mixed into both the front and surround channels at comparable levels. Even with directional cues, so long as you don't have sounds mixed between the main and surround channels to try and create a side imaging effect, they can work with dipoles.

The formats themselves do not require direct radiators. The ITU multichannel reference placement that specifies the positioning for direct radiating speakers in monitoring setups are applicable to how 5.1 music mixes are done, but not necessarily for how movies are mixed. Movies soundtracks are mixed on dubbing stages with large arrays of surround speakers. The type of positioning with direct radiators that gives you ideal imaging with multichannel music will leave noticeable holes in the soundfield with movie soundtracks that are mixed with a lot of ambient cues in the back.

This is why Dolby's placement guidelines recommend that direct radiators be raised at least two feet above ear level and pointed directly at one another. The whole purpose of this is to diffuse the sound with mostly ambient movie soundtracks, while retaining the strong imaging cues that you get with multichannel music and an increasing number of movie soundtracks that get repurposed for the DVD releases. The ambient movie soundtracks are actually better conveyed with dipoles, and direct radiators compromise this even when those compromises are minimized by using the Dolby recommended placement. However, those soundtracks with stronger imaging cues are more compromised by dipoles, and are ideally rendered using direct radiators.

Paradigm is one of the manufacturers that recommends dipolar surrounds. They even post a paper on their website written by Tomlinson Holman of THX that touts the reasons why dipoles are desirable for multichannel music. I don't agree with it (I own a set of Paradigm speakers, but I opted to use bookshelf speakers as surrounds), but that's coming straight from someone who's had a major role in creating the current multichannel standards. Basically, neither approach is ideal for all sources, so it all comes down to which set of compromises you're more willing to make. Given that more and more movie soundtracks are getting mixed with well defined side imaging cues created through the surround speakers, the viability of direct radiators has risen, but there's still a lot of material out there that does sound better with dipoles.

RobotCzar
02-25-2005, 06:53 PM
Yes and no. The issue is not nearly as cut and dry as you make it out to be. I agree with you about direct radiating speakers as the best option for discrete 5.1 setups, but it doesn't mean that dipoles are wholly without merit. .

I agree that encoding (= recording setup) can make a difference. If one intentionally tried to use some omnidirectional mikes to capture room ambience and put that on the rear channels, then one might get a reasonable result with dipole rear speakers. But that is a waste of true multichannel. The idea is to have the rear speaker directly recreate the sound field reflections that were recorded by the multichannel array of mikes. Trying the utilize the trick affect of dipoles to recreate sonic ambience is redundant and very likely to muck things up.

In a way, the whole idea of diploles (or bipoles) is for the non-direct sound to simulate recording venue ambience (reflected sound). They, of course, are not doing this based on any real plan or encoding, they just give a suggestion of 3D sonic space. (Note: I love how my dipoles sound.) Unfortunately, this effect is exactly what rear (and side) channels are trying to do, but they expect to do it (establish a 3D sound field) with direct sound (i.e., the direct sound carries the reflected sound of the recording venue). The reflections of your listening room are just going to mess up the intended (directly projected) reflections carried on the rear channels. Ideally, a multichannel system should be set up in an anechoic chamber. With multichannel, you want to MINIMIZE the reflections of your room. Needless to say, dipoles depend on reflections from your room for their effect. Hence, dipoles and multichannel don't mix.

Sir Terrence the Terrible
02-25-2005, 09:08 PM
In a way, the whole idea of diploles (or bipoles) is for the non-direct sound to simulate recording venue ambience (reflected sound).


Dipoles are itilized to mimick the soundfield that is created in the theater using mulitple speakers in a array. They were not ment to simulate any type of ambience, but to create a larger picture with fewer speakers.



They, of course, are not doing this based on any real plan or encoding, they just give a suggestion of 3D sonic space. (Note: I love how my dipoles sound.) Unfortunately, this effect is exactly what rear (and side) channels are trying to do, but they expect to do it (establish a 3D sound field) with direct sound (i.e., the direct sound carries the reflected sound of the recording venue).

The concept is actually to create two zones in your listening room. One of controlled directivity in the front of the room(great for imaging and clarity) and another is a reverberant soundfield(one of space by reflections). This is what is found on the dubbing stage, but implemented quite differently.


The reflections of your listening room are just going to mess up the intended (directly projected) reflections carried on the rear channels.

Actually, they spread them out, as if mimicking a surround array. That doesn't necessarily destroy the effect, but it sure changes its proportions.


Ideally, a multichannel system should be set up in an anechoic chamber. With multichannel, you want to MINIMIZE the reflections of your room. Needless to say, dipoles depend on reflections from your room for their effect. Hence, dipoles and multichannel don't mix.

You don't every want to created a overdamped room, and it really isn't neccessary to minimize reflections as much as it is important to control them. Dipoles do a very effective job of creating a reverberant soundfield just like you would experience in a concert hall. The reality is instruments and voices were never ment to be projected from the rear, unless called for by the score(Berlioz wanted a brass and voice choir for his reqieum spl?). Were dipoles become unsuitable is when you mess with the phase of the signal and that is why I prefer bipolar speakers for the surrounds rather than dipoles. Bipolar speakers do not have phase mismatches with the fronts, but can project a very wide reverberant field. Side and Center rear phantom imaging is easily heard and stable(something not possible with dipoles). Even when instruments are mixed in them, its position can be audiblly perceived.

RobotCzar
02-26-2005, 09:27 AM
Dipoles are itilized to mimick the soundfield that is created in the theater using mulitple speakers in a array. They were not ment to simulate any type of ambience, but to create a larger picture with fewer speakers.
I don't really want to debate, but your statement is not my understanding of dipoles nor my experience in listening to them. I don't know what a "larger picture" is other than acoustical abience unless one means that the sound of an instrument is coming from a particular postition which is probably the goal of HT multichannel and not really approriate for music.



The concept is actually to create two zones in your listening room. One of controlled directivity in the front of the room(great for imaging and clarity) and another is a reverberant soundfield(one of space by reflections). This is what is found on the dubbing stage, but implemented quite differently.
You are describing the older dolby pro logic approach, which, again, is appropriate for special effects in movies, not music. There are not two zones in the original location. I don't doubt that some recording "engineers" are attempting to do what you say. The attempt is simply a misuse of multichannel and one reason that I think the commercial audio biz will blow it with multicannel (in regard to music listening).


Dipoles do a very effective job of creating a reverberant soundfield just like you would experience in a concert hall. .
I also disagree with this statement. Dipoles use your room reflections to make what you hear sound more reverbant and more lifelike. The effect, however, is artificial in that it is NOT based on what you would hear at the recording venue. There is no way for bipoles to accurately decode what happened in the recording venue. So, ultimately, dipole effects are recreations or, if you prefer, illusions. I really like my dipoles because they give a more satisfying sound with the appearance of "realistic" sonic abience. But, I am not fooling myself that what I hear is accurate to the recording venue. True multichannel has the potential to recreate the ambience of the recording venue, but not if you create artificial zones or go for dramatic spatial effects.

Sir Terrence the Terrible
02-26-2005, 12:37 PM
I don't really want to debate, but your statement is not my understanding of dipoles nor my experience in listening to them. I don't know what a "larger picture" is other than acoustical abience unless one means that the sound of an instrument is coming from a particular postition which is probably the goal of HT multichannel and not really approriate for music.

If you want to understand dipoles usage in both music and HT, go back to the individual who brought them to the forefront. Tomlinson Holman. His goal in creating the Home THX standard was to use them to mimick the 12(or more) surround speakers that make up the surroundfield of a typical professional theater and dubbing stage. When signals are panned to the surrounds, it expands the sound just like a surround array would(going from one speaker to a spaced 4 will expand the sound), and that reduces localization. The dipole does the same thing, but uses phase variances between the front half of the speaker, and the rear half to produce the spread(spreading cannot be helped when you have a speaker with a front panel of drivers and a rear panel of drivers) and the the delocalization. This will happen whether its music, or film soundtracks.




You are describing the older dolby pro logic approach, which, again, is appropriate for special effects in movies, not music. There are not two zones in the original location. I don't doubt that some recording "engineers" are attempting to do what you say. The attempt is simply a misuse of multichannel and one reason that I think the commercial audio biz will blow it with multicannel (in regard to music listening).

No that is not the old prologic approach at all, that describes what is currently found in professional theaters and dubbing stages. Also many acousticicans believe that for better imaging the front, and the front 1/3 of the side walls should be treated with absorption(minimizes early reflections and enhances L/R and front to back imaging), and create a diffused rear soundfield using diffusion and abfusion which enhances envelopement.

Read this PDF by Peter D'Antonio of RPG

http://www.rpginc.com/news/reflections/drv4i2.pdf

Notice his comments in the multichannel area.


I also disagree with this statement. Dipoles use your room reflections to make what you hear sound more reverbant and more lifelike.

I believe I stated this very point in my post.


The effect, however, is artificial in that it is NOT based on what you would hear at the recording venue.

While it is not EXACTLY to what you will hear in a recording venue(every recording venue is different, and will have different acoustics), the concept is based on the acoustics of the dubbing stage, and concert hall. In the concert hall, our sense of spatial spaciousness comes from reflections that originate from the sides of the room. That is why it is recommended that the rear speakers in a hometheater(and in professional theaters) setup be position there. In the theater, the use of multiple spaced surround speakers create the spatial spaciousness via a comb filtering effect that happens as a result of the different arrival times of the output of these speakers to the ears. The dipole is used to mimick these kinds spatial characteristics even if the reflections they generate are user room based and are not found in the concert hall at all.



here is no way for bipoles to accurately decode what happened in the recording venue.

No speaker(not even a monopole) can do that.


So, ultimately, dipole effects are recreations or, if you prefer, illusions. I really like my dipoles because they give a more satisfying sound with the appearance of "realistic" sonic abience.

The signal it reproduces is not a recreation, the spatial effects created by the reflections are. All a dipoles do is take the signal that is fed to it, and spread it by way of room reflections to enhance envelopement. That is not necessarily restricted to just ambience, but anything fed to it.


True multichannel has the potential to recreate the ambience of the recording venue, but not if you create artificial zones or go for dramatic spatial effects.

Dipoles as used in the THX standard(in the rear) were never designed for music playback, it was designed to playback soundtracks. Their very standard(and every other acousticians) was to create a non reflective, and highly reflective zones. That is why limited despersion speakers are used in the front(just like in movie theaters) and highly reflective speakers are used in the rear(to simulate the surround array in theaters)

http://www.thx.com/mod/techlib/speakers.html

http://www.thx.com/mod/techLib/how.html

The best multichannel can do is approximate the ambience of a concert hall. Two spaced speakers cannot recreate the complex reflection pattern, or the different times these reflections hit our ears. That is why 10.2, 20.4 and every other multiple speaker configuration has been proposed for multichannel music.

Creating acoustical zones is the only way to get good acoustics in small rooms. Creating dramatic spatial effects is as much a part of soundtrack creation, as balance and clarity.
Dramatic spatial effects are often utilized by early composers in their music. Berlioz used it in his requiem as did Beethoven in his.

Coldwater
02-26-2005, 01:53 PM
I never knew dipoles were so complex. I find they work perfect for my setup, as my surrounds are 3-4 feet away from me against a wall.

RobotCzar
02-26-2005, 09:46 PM
"No speaker(not even a monopole) can do that."

There are several approaches to mutltichannel. While I have no interest in mimicing theater audio systems, I don't doubt that is the goal of HT multichannel systems. I am interested in mimicing a concert hall or club--the venue of a live performance that is recorded. I want a realistic reproduction of a live performance, not a reproduction of what you hear in a movie theater (though I don't believe a dipole would do do even that better than direct radiators).

Mulitichannels can mimic a sound field quite well. I have suggested tha Ambisonics is a much better approach for music reproduction than THX systems. Ambisoic decoded mutltichannel can recreate the original acoustic field, including recording venue reverberation. Even unencoded multichannel can also actively record 3D aspects of the recording venue (if recorded properly). Then a direct raditating speaker will then reproduce the original acoustics. A dipole merely sends some of its sound off axis (usually 180 degrees) so that the listening room will reflect some of the sound and result in a more reverberant field. There is no distinction between what is reflected and what is directly projected. Obviously, a spearker driven by a multiple channel recording CAN reproduce sound appropriate for the soundfield of the original recording via differences in the channels. A dipole can too, but it confuses the sound by sending some off in a different directions (NOT based on the original recording). The effect you get from a dipole is determined by the room, not the recording; but the differences in the channels ARE the result of the original recording and (have the potential to) reproduce acoustic aspects of the recoding venue.

I find it very difficult to believe that a dipole can ever result in more realistic sound in a multichannel setup. If a theater system is using dipoles for some effect, then use of diploles would result in a reasonable recreation of that arrangement. But, I can't see why a theater system would do that. A theater can completely surround you with speakers, they have no need to use reflected sound. And, I don't recall ever seeing a dipole in a theater. There is just no logic (pro or otherwise, grin) that would indicate what a dipole would be a good idea in a multichannel setup IF accuracy is one's goal.

I have from the start felt that HT multichannel would get it wrong. Too bad, because multichannel could really move home audio forward if done right. The bottom line of my point is that recording room reflections (reverberation) is carried by the multiple channels, there is no need to bounce sound off walls of the listening room--that would only interfere with the sound comming from the multiple channels. I suggest you read more than one source regarding multichannel and that you look at the topic beyond THX.

Sir Terrence the Terrible
02-27-2005, 12:28 PM
"No speaker(not even a monopole) can do that."

There are several approaches to mutltichannel. While I have no interest in mimicing theater audio systems, I don't doubt that is the goal of HT multichannel systems.

That is the goal of SOME hometheater systems. Others want accuracy, and faithfulness to the soundtrack without any enhancements.


I am interested in mimicing a concert hall or club--the venue of a live performance that is recorded. I want a realistic reproduction of a live performance, not a reproduction of what you hear in a movie theater (though I don't believe a dipole would do do even that better than direct radiators).

Well good luck in trying to mimick a concert hall of club with two surround speakers. A concert hall has literally thousands of incendental and coincedental reflections arriving at different times to the ears, and that is something cannot be relayed to the listener with just two surrounds, and three front speakers.


Mulitichannels can mimic a sound field quite well.

How would you support such a statements, recording experience? I know for a fact it doesn't even come close. It can't, not with just 5 or so microphones(or even 20 for that matter), and 6 speakers for playback. Wishful thinking on your behalf.


I have suggested tha Ambisonics is a much better approach for music reproduction than THX systems.

You are mixing apples and watermelon here. THX is a speaker system, controller, and amps, b-chain stuff. Ambisonics is a recording format that works great if you sit equidistant from all speakers, but fails just like any another mapping process off axis. No matter how good the process is, it can only be optimized for a single seat because of phase issues.



Ambisoic decoded mutltichannel can recreate the original acoustic field, including recording venue reverberation.

It does a reasonable facsimile, but it has the same limitations as multichannel SACD or
DVD-A, the pipeline is too small to record every reflection heard in a typical concert hall.


Even unencoded multichannel can also actively record 3D aspects of the recording venue (if recorded properly).

Is this some more of your first hand recording experience talking to you?


Then a direct raditating speaker will then reproduce the original acoustics.

You are dreaming man, it can't be done. Now with 500-1000 direct radiators, you might have a possibility, but not with 5.1


A dipole merely sends some of its sound off axis (usually 180 degrees) so that the listening room will reflect some of the sound and result in a more reverberant field.

I believe that has been covered in the course of this dicussion already.


There is no distinction between what is reflected and what is directly projected.

Highly depends on the design of the dipole. Obviously if you have a dipole that is designed with the woofer facing into the room(that would be direct output with no reflections except from the ceiling and opposing wall), and the mid/tweets pointing at 45 degree angles, then there is a distinction for what is reflected, and what is direct.


Obviously, a spearker driven by a multiple channel recording CAN reproduce sound appropriate for the soundfield of the original recording via differences in the channels.

Can you fully explain to me how a single speaker can reproduce the appropriate soundfield(and 5 can't), and just how do you extract the differences between the channels without going to matrix technology? The recorded in phase and out of phase signals come from each channel with no distinction between them.


A dipole can too, but it confuses the sound by sending some off in a different directions (NOT based on the original recording). The effect you get from a dipole is determined by the room, not the recording;

You are dead wrong in your conclusions. If I I choose to flip the phase, or distant record a signal on a recorded channel, with a dipole, it will sound much more reverberant and open. If I close record a signal that has little phase variance, then I can make the dipole sound as dry as powder even with the reflections it creates in the room. With a dipole the in phase signals are reproduced by the front half of the panel, and the out of phase signals are reproduced by the rear panel. The combination of the two signals are projected near the walls to mix the two signals up in room. Recordings also have in phase and out of phase signals. Depending on the balance of these signals directly effect whether you hear spaciousness, or acoustical dryness(lack of reverb).


but the differences in the channels ARE the result of the original recording and (have the potential to) reproduce acoustic aspects of the recoding venue.

If you add the word "certain" between reproduce and acoustics, then we agree.


I find it very difficult to believe that a dipole can ever result in more realistic sound in a multichannel setup.

I am sorry you find it so difficult, however if ambience is all we speak of, then a dipole is much better at reproducing a soundfield than a direct radiator. If I record ambience in the rear channels, and it is reproduced by two direct radiators, it will not sound nearly as ambient than when projected by a dipole. That is because of the in room reflections of the dipole combined with the fact that a dipole seperates the in phase signals from the out of phase signals. A direct radiator cannot seperate the out of phase signals from the in phase ones, it all coming out of one pipeline. In the case of heavy in phase signals, direct radiators suffer from the "speaker in the head" phenom, the dipole can never suffer that because of its design.


If a theater system is using dipoles for some effect, then use of diploles would result in a reasonable recreation of that arrangement.

Unfortunately dipoles are not used in soundtrack creation, multiple spaced direct radiators are. In this case, dipoles are used to create an acoustical effect, not a direct translation of what is on the soundtrack.


But, I can't see why a theater system would do that. A theater can completely surround you with speakers, they have no need to use reflected sound.

If you sit in the audience in any movie theater, you are hearing more reflections than direct output. THX recommends the use of dipoles to mimick the effect of having multiple speakers surround you. It's an effect, but not a great translation at all.


And, I don't recall ever seeing a dipole in a theater. There is just no logic (pro or otherwise, grin) that would indicate what a dipole would be a good idea in a multichannel setup IF accuracy is one's goal.


Theaters don't need dipoles, it uses multiple spaced direct radiators to create a reverberant field(along with the theaters vast space), a effect that THX uses dipoles(along with other electronic means) to mimick. If you are trying to mimick the acoustics of a movie theater, then a dipole would be the most economical way.

You are correct, if accuracy is the issue, then a dipole shouldn't be used. However THX isn't going for accuracy, they are going for the effect, and that is why they chose dipoles as part of their standard.


I have from the start felt that HT multichannel would get it wrong.

Who says they got it wrong?


Too bad, because multichannel could really move home audio forward if done right.

What is right? There are many ways to mix a multichannel soundtracks, there is no right way since it is a manufacturing process from the beginning with film soundtracks and studio based recording.


The bottom line of my point is that recording room reflections (reverberation) is carried by the multiple channels, there is no need to bounce sound off walls of the listening room--that would only interfere with the sound comming from the multiple channels.

If you are trying to create a effect of expansiveness in a small room, then speaker induced reflections are a must. That is why omnipolar and electrostatic speakers are so popular. Unless you have more than two channels carrying ambience, then you are hearing only a fraction of the natural ambience in the hall. Two microphones, and two speakers is to small of a pipeline to accuractly recreate that.

Keep in mind, if you have three controlled dispersion speakers, and two highly reflective speakers, all properly calibrated in terms of volume, and delay, there is no interference between the two outputs because of the precedent effect. The output from the front speakers arrives before the rear speakers just like you would experience if you sat in the first few rows of a concert hall.




I suggest you read more than one source regarding multichannel and that you look at the topic beyond THX.

Obviously I have, which is why I can tell you how wrong you are in this subject matter. ; )

RobotCzar
03-01-2005, 04:44 PM
Let me say that I don't think I have been clear in my objection to dipole or omnidirectional speakers in multichannel, so let be take another shot.

I did read the Web references given by STT and, as I expected they don't make a lot of sense. I can't help it THX has mucked up multichannel, but perhaps people should think about it a bit more. To be charitable, THX seems to have a goal of recreating "theater sound" in the home. Now, since they get to define what "theater sound" is, they can pretty much do and say what they please about it. I have a few problems with it. The main problem is that I do not seek theather sound from my main audio system. I am after realistic concert hall sound. I want a accurate and realistic reproduction of the live sound of a concert hall (or other venue).

Now, that said, let me return to multichannel. The main potential of multichannel is that it can more accurately reproduce the ambient (reflected sounds) of a recording venue. The recorded signal can be encoded to reproduce the reflected sounds via the multiple channels. In effect, the channels carry information about the reflected sound and directly reproduce the reflected sounds of the recording venue on each speaker. With proper encoding, the ambient acoustics of the recording venue are recreated by the multiple channels. Note I have said nothing about reflections from one's listening room; THOSE ambient (reflected) sounds are undesireable and unecessary as the ability to surround the listener with multiple channels is already directly handling the ambience.

Now, diopoles are an attempt to use your room acoustics (reflections) to "simulate" a larger, more obvious space by bouncing some of the speaker's sound off the surfaces of your listening room. A dipole does not do this based on any information from the recording (i.e., information from the recording venue acoutics), the dipole merely sends a percentage (often 50%) of its sound off in another direction (usually 180 degress from the main axis). This indirect sound lends spaciousness to what one hears by using the listening room acoustics more than direct radiators do.

Now, if multichannel is already producing all the ambience within its channels, then any misdirection of sound intended to use your room acoustics will interfere with the ambient sound being produced by the multiple channel. Therefore, if your goal is accurate or realistic concert hall listening (the goal of choice for many audiophiles) then using a dipole or omni speaker is not ideal with multichannel (you will get less accurate sound).

To use the vernacular, reflecting speakers are no good for music but might be okay for home theater. The unfortunate aspect this that there is no recording company or hardware makers that support proper multichannel for music--current multichannel is driven by the HT market. So, I will end by adjusting my original statement by saying that using dipoles for multichannel sound is very unlikely to give satisfactory results for music IF your goal is concert accuracy. If you want the starship Enterprise to sound like it does in the movies, then omnis might work.

RobotCzar
03-04-2005, 11:51 AM
Dipoles as used in the THX standard(in the rear) were never designed for music playback, it was designed to playback soundtracks. Their very standard(and every other acousticians) was to create a non reflective, and highly reflective zones. That is why limited despersion speakers are used in the front(just like in movie theaters) and highly reflective speakers are used in the rear(to simulate the surround array in theaters)
We are in agreement about this, THX is s theather simulaiton, not designed for music playback. I did get around to reading the article by D'Antonio. I must say that reading articles by CEOs is almost never instructive, and I didn't find this article particularly useful.

In the section entitle "Multichannel formats" (which THX isn't) D'Antonion says

"DIPOLE SURROUNDS: In the THX format, the surrounds are dipoles located on the side walls to the left and right of the listener. THX feels this arrangement simulates the commercial cinema experience in which there are several monopole surrounds distributed on the rear and side wallssive), "

The operant phrase is "THX feels", this signals to me that THX is not about to say "don't use expensive dipoles" and offend the makers of such speakers. Note also that this quote says THX in theaters does not use dipoles (I thought not, as there is no reason to). I'd say that you will get better "simulation" of theater THX using speaks as THX does. Based on this statements, I'd say monopoles are better for theater reproduction.



The best multichannel can do is approximate the ambience of a concert hall. Two spaced speakers cannot recreate the complex reflection pattern, or the different times these reflections hit our ears. That is why 10.2, 20.4 and every other multiple speaker configuration has been proposed for multichannel music.
as did Beethoven in his.

In is indeed very difficult for two speakers to create a complex reflection patters (dipole will not help). True mutltichannel formats for music such as Ambisonics, do have the potential to approximate the ambience of a concert. Ambisonics works with only two speakers, but the approimation becomes better with an increase in the number of speakers. 10 or 20 speakers would be nice but 5 works out very well in recreating the sound field of the recorded venue.

Sir Terrence the Terrible
03-04-2005, 12:53 PM
Let me say that I don't think I have been clear in my objection to dipole or omnidirectional speakers in multichannel, so let be take another shot.

I did read the Web references given by STT and, as I expected they don't make a lot of sense. I can't help it THX has mucked up multichannel, but perhaps people should think about it a bit more. To be charitable, THX seems to have a goal of recreating "theater sound" in the home. Now, since they get to define what "theater sound" is, they can pretty much do and say what they please about it. I have a few problems with it. The main problem is that I do not seek theather sound from my main audio system. I am after realistic concert hall sound. I want a accurate and realistic reproduction of the live sound of a concert hall (or other venue).

As I mentioned to you in my previous posts, you will only get an approximation of live sound out of 5.1 channels. If you want the accurate and realistic reproduction, then be prepared to line your walls and ceiling with speakers, and be prepared to spend big time on amps. There is NO way you will hear what is realistic and accurate in a concert hall outside of that kind of solution.


Now, that said, let me return to multichannel. The main potential of multichannel is that it can more accurately reproduce the ambient (reflected sounds) of a recording venue.

Compared to what?


The recorded signal can be encoded to reproduce the reflected sounds via the multiple channels. In effect, the channels carry information about the reflected sound and directly reproduce the reflected sounds of the recording venue on each speaker.

I repeat, it cannot be done with 5.1. 10.1 brings you alot closer, but 5.1 is only a small approximation.



With proper encoding, the ambient acoustics of the recording venue are recreated by the multiple channels. Note I have said nothing about reflections from one's listening room; THOSE ambient (reflected) sounds are undesireable and unecessary as the ability to surround the listener with multiple channels is already directly handling the ambience.

Here is where I think you are completely falling short. You are attempting to simplify an extremely complex phenoma. In a concert hall(depending on the building material used to create it)one reflect creates several more before falling below the threshold of hearing. If you have an orchestra of 110, or a small ensemble in a large hall, the instruments played can create hundreds of complex reflections directly, which turns into thousands in the reverberant field before they drop below the threshold of hearing. There is no way right now to map the complex reflection pattern of a typical(if there is such a thing) concert hall. If you have ever done any live recording of orchestras, and had to mix them back in the control room, you would understand completely the difference between recorded sound, and the live real deal. The loss is staggering in the transfer.


Now, diopoles are an attempt to use your room acoustics (reflections) to "simulate" a larger, more obvious space by bouncing some of the speaker's sound off the surfaces of your listening room. A dipole does not do this based on any information from the recording (i.e., information from the recording venue acoutics), the dipole merely sends a percentage (often 50%) of its sound off in another direction (usually 180 degress from the main axis). This indirect sound lends spaciousness to what one hears by using the listening room acoustics more than direct radiators do.

You still are not correct in your understanding of how a dipole functions. First, not even a dipole could generate enough reflections in and of itself to add spaciousness. They add very little to reverberation(the paths to the ears are too small)time in small rooms. What makes a dipole sound more spacious lies in the speakers electronics itself. Dipoles seperate in phase signals(which add nothing to the effect of spaciousness) from out of phase signals(the signals where the sense of spaciousness is derived) and via reflections created by design, mix the two signals together in the room. A simple experiment would give you a great idea how this works. Take two monopolar surround speakers and wire them out of phase with two main speakers. What you will hear is a solid sense of spaciousness in the rear channels as you are only getting the difference signal in the mix. The more you turn the effect up, the more spacious(but fuzzy and indistinct) the sound becomes. You would get the same effect running these signals through a prologic decoder, and listening to what is coming out of the surrounds. Without the out of phase signals, a dipole would sound like a bipolar speaker which is not nearly as ambient sounding as a dipole as it is not wired for out of phase signals. In small rooms it is not the reflections of a dipole that make it sound spacious, it is the amount of out of phase signals(and their overall amplitude in the mix) that makes it sound spacious. The reflections from the dipole just spreads them out


Now, if multichannel is already producing all the ambience within its channels, then any misdirection of sound intended to use your room acoustics will interfere with the ambient sound being produced by the multiple channel.

It is clear that you have never heard of the "presidence"effect. If those signals reproduced by the dipole are delayed, then the signals that arrive first would be heard the loudest followed by any signals arriving afterwards. So it is possible for a dipole or a bipolar speaker to be effectively used with multichannel, as long as the only signals being mixed into the surrounds are only ambient signals, are sufficently delayed in respect to the direct signals from the front speakers.. These days that is not the case, and why dipoles are not advised for multichannel music.




Therefore, if your goal is accurate or realistic concert hall listening (the goal of choice for many audiophiles) then using a dipole or omni speaker is not ideal with multichannel (you will get less accurate sound).

Let me ask you a question. Do you think that reflections arrive from a single point in space, or from multiple points in space? Your comment is totally appropriate if we are talking about the frontal hemisphere, but we are discussing the rear soundfield, and in this case your comments fall flat. Since we don't hear reflections in a concert hall from one point in space, then it could be argued that dipoles are perfect for the rear channels of even multichannel music. However, a great deal of multichannel music out there puts voices, drums, instruments, small percussion, bells, and everything else under the sun in the surrounds. This is why dipoles ARE NOT recommended for multichannel music, not because they are less accurate in protraying a ambient soundfield(we know they are superior in the effect to direct radiators)


To use the vernacular, reflecting speakers are no good for music but might be okay for home theater.

If you are talking about the frontal hemisphere, reflecting speakers are not good for either hometheater or music. If we are talking about the rear hemisphere, then they are perfect for both(music the way it is currently mixed presents a problem with using dipoles)


The unfortunate aspect this that there is no recording company or hardware makers that support proper multichannel for music--current multichannel is driven by the HT market.

Since you don't mix music the way you mix film soundtracks, then your statement is not truely supported by fact. In hometheater, the center channel dominates the rest. In music, it is the combination of all the channels collectively. Mostly all record companies mixing music have no affiliation whatsoever to companies that mix film soundtracks. I know of only a few recording engineers (myself included) who do both. The HT market paved the way for multichannel music, but that is the only drive that the two have in common.


So, I will end by adjusting my original statement by saying that using dipoles for multichannel sound is very unlikely to give satisfactory results for music IF your goal is concert accuracy. If you want the starship Enterprise to sound like it does in the movies, then omnis might work.

If "Concert accuracy" is what you are looking for, then you'll have to go to the concert hall to get it. No format at this point can "accurately" bring it to you. If the ambience of a concert hall was the sole signal mixed into the discrete surrounds, then a dipole and a bipole speakers would be much better at conveying it in your room than a monopole.
There have been literally dozens of experiments to support that done over the years

Sir Terrence the Terrible
03-04-2005, 01:15 PM
We are in agreement about this, THX is s theather simulaiton, not designed for music playback. I did get around to reading the article by D'Antonio. I must say that reading articles by CEOs is almost never instructive, and I didn't find this article particularly useful.

Rather condescending statement considering the guy knows more about acoustics than you do. His material does require that you have at least a basic understanding of room acoustics, which may be why this didn't particularly help you.


In the section entitle "Multichannel formats" (which THX isn't) D'Antonion says

"DIPOLE SURROUNDS: In the THX format, the surrounds are dipoles located on the side walls to the left and right of the listener. THX feels this arrangement simulates the commercial cinema experience in which there are several monopole surrounds distributed on the rear and side wallssive), "

The operant phrase is "THX feels", this signals to me that THX is not about to say "don't use expensive dipoles" and offend the makers of such speakers. Note also that this quote says THX in theaters does not use dipoles (I thought not, as there is no reason to). I'd say that you will get better "simulation" of theater THX using speaks as THX does. Based on this statements, I'd say monopoles are better for theater reproduction.

Now I am not about to defend THX on anything, but THX really didn't do the homework that created the standard. The homework was done by Tomlinson Holman. Since he was my professor in film school, I understand all the research he did to come to his conclusions. He did quite a few listening test that compared the array of speakers in a dubbing stage to two monopoles, and two dipolars in a small room. Based on these test, and the notes from the participants, he came to the conclusion that two dipolar speakers sounded more like that array than two direct radiators. I agree with his conclusions(but I think he still discounts the big space of the dubbing stage contributions to the sense of spaciousness)
THX in movie theaters don't need dipolar speakers because 1. the array used along with the large space already creates envelopement and the sense of spaciousness so it is not needed and 2. the long reverberation time of most theaters would make dipolars sound a mess in the seats and 3. coverage problems would be huge using dipolar speakers, as there would be now way to use coverage prediction software to install them properly.

Keep this in mind, at no point is a film soundtrack mixed or mastered using just two loudspeakers in the surrounds. So what makes you think that only using two would be more accurate. Most installs that I have done used a minimum of 6, as I would rather use multiple surrounds in my installs than dipoles. That however is not practical in 90% of listening spaces.



In is indeed very difficult for two speakers to create a complex reflection patters (dipole will not help).

Dipoles would do alot better than two monopoles, that is for certain.



True mutltichannel formats for music such as Ambisonics, do have the potential to approximate the ambience of a concert. Ambisonics works with only two speakers, but the approimation becomes better with an increase in the number of speakers. 10 or 20 speakers would be nice but 5 works out very well in recreating the sound field of the recorded venue.

If you had 10 or 20 speakers to use, then you don't need ambisonics. Ambisonics is not a real solution to creating an accurate soundfield, because it is listener position depended. Ambisonics cannot be optimized to sound the same to people off axis as it does to people on axis. So you are still tied to a very small sweet spot. That makes it nothing more than a small band aid than a total solution

blasterman_
03-07-2005, 02:16 PM
Not trying to get in between you two, but being this is one of the few discussions I've seen that treat rear channel seriously......well, my $.02

Years ago I was visiting a friend in Chicago, and we stopped at a small custom audio shop that sold a hodge-podge of speakers and other HT gear. In their main 12x18 show room they had a HT set-up consisting of Definitive-Tech speakers, which at the time was about the only serious player in the bi-polar game. The salesman popped in a laserdisc of some sci-fi movie with a scene of an explosion in space, and a violently expanding 'ring of gas' passing through the starship or whatever.

To mine and my friends utter shock, we could literally 'feel' the exploding ring pass through us just as on screen and pass through the room with near ghostly omnipresence. It was astounding display of HT technology at the time that left us awed for years.

Since then I've been in countless multi-million dollar homes and been privy to sitting in and listening to more custom $25,000 HT set-ups than I can list. Honestly? Only one system I've heard in the past decade matched the rear channel of that old rig I heard in Chicago running clunky Dolby Pro Logic and lowly Definitive Tech bi-poles on the rear walls.

You'd think with the much better speaker technology today (I wouldn't use DT now if you gave them to me), and the discrete and better dynamics of DTS and Dolby Digital we'd have some killer sounding HT set-ups. Regretfully, while I think 2-channel and certainly video technology has improved, I don't feel rear channel implementation has improved much at all. In virtually every single "high-end" HT sitting I've been in, I can hear the left rear channel localized, and I can hear the right rear channel localized. Unless I'm watching Saving Private Ryan and listening for bulltets richocheting off walls behind me, it sounds frankly 'dumb'. In these cases I'm hearing either direct radiators, or poorly implemented di-poles usually placed in corners and too low a height. Again, I dont' care how it's supposed to work, but only using my ears as a reference. I simply don't like the rear sound field when it's produced by direct radiating speakers pointed at the listeners, or 'beaming at you from a wall. Give me those old BP-2's mounted above and away from the rear wall any day.

For music, I have different conclusions. Not sure if I know what I'm talking about and please correct me if I'm not, but my understanding is that most multi-channel music recordings are simply extracted from a much simplier direct recording, and merely enhanced and doctered to sound like a multi-channel in the lab. Except for some rare instances like Delos labs and their miraculous recordings, I'd rather turn the rears off for music anyways.