Sigview here is one package I have, and yes it can do what I thought. Last time I was able to do this was June (last time my computer really worked right), so excuse me if I'm a bit gray on the details. I also forgot to mention that I often use filters to remove things I'm not concerned with looking at, and to make the statistical analysis easier. For instance, if I want to know if any content exists above 5khz, just set a highpass filter at 5khz, then all you get is content over that. Then you can calculate proportion scores of content over 5khz to total content. You can also calculate parameters related to frequency and amplitude, say an interaction for instance of high frequency content and high amplitude.
I've also done some comparisons of surround channels to mains, using this same software, and calculating statistics on the differences. The most interesting for me was looking at the High def surround content of the Transformers Movie durring some flyby scenes. I could compare the pan from rear right to front center (which involved the right front as well). If all channels were being used full bandwidth and meant to represent what a giant flying robot jet would sound like flying overhead at very low altitude, I would expect that the surround would roughly match the front in dynamic range of this content as well as frequency. However, this was not the case, as the front had content down into the single digits and up to over 15khz, covering roughly 30db's of dynamic range as well (Difference between softest and loudest noise above the noise floor). The surrounds had a little energy in the 80hz range, but really started around 100hz (engineer put in a filter I think) and then stopped around 8khz, with some but much less content to 12khz. This was the particular soundtrack I mentioned which was using the new format. For that particular setup I took a toshiba HD DVD player and output the analogue signal to my soundcard (Mark of the Unicorn firewire outboard 8ch) with the built in decoder set to all large full range speakers. Now, if the Toshiba contains some built in processing that filters the signal, or was an issue of the analogue signal, then this could all be just measurement artifacts, but I don't believe thats the case. I have looked inside and the analogue outs are all identical circuits, I know I can measure above and below these points, so I think its the soundtrack. This by the way, is by far the widest bandwidth I have ever measured in a surround soundtrack.
I've also done some comparisons of surround channels to mains, using this same software, and calculating statistics on the differences. The most interesting for me was looking at the High def surround content of the Transformers Movie durring some flyby scenes. I could compare the pan from rear right to front center (which involved the right front as well). If all channels were being used full bandwidth and meant to represent what a giant flying robot jet would sound like flying overhead at very low altitude, I would expect that the surround would roughly match the front in dynamic range of this content as well as frequency. However, this was not the case, as the front had content down into the single digits and up to over 15khz, covering roughly 30db's of dynamic range as well (Difference between softest and loudest noise above the noise floor). The surrounds had a little energy in the 80hz range, but really started around 100hz (engineer put in a filter I think) and then stopped around 8khz, with some but much less content to 12khz. This was the particular soundtrack I mentioned which was using the new format. For that particular setup I took a toshiba HD DVD player and output the analogue signal to my soundcard (Mark of the Unicorn firewire outboard 8ch) with the built in decoder set to all large full range speakers. Now, if the Toshiba contains some built in processing that filters the signal, or was an issue of the analogue signal, then this could all be just measurement artifacts, but I don't believe thats the case. I have looked inside and the analogue outs are all identical circuits, I know I can measure above and below these points, so I think its the soundtrack. This by the way, is by far the widest bandwidth I have ever measured in a surround soundtrack.
markus76 said:
This is a little harsh don't you think?
Mr. Holman is the one who recommends dipoles. He, like myself, believe that surrounds are only for adding effects that should be non-obtrusive and non-localized. Dipoles yield no direct sound only reverberant energy, which in a surround is, by my estimation, ideal (otherwise why would we delay them by so much?). As I said before there are no hard and fast rules about surrounds, so do what you want. I use dipoles - but only for film. Otherwise I don't use them at all.
"The point to consider is that in my setup the sides are dipoles and the listener is on the dipole axis. This is the situation recommended by Thomason Holman, but quite the opposite of Floyd Toole."
The use of dipoles was first recommended by TH because the surrounds were severely limited by being matrixed and mono, and when hard discrete effects like we get now were but a dream.
As Markus alluded to, the ambience is supposed to be in the program material and/or processing.
"With a speaker like the Nathan as a side surround it would basically be like wearing headphones. In my room they would only be a few feet away pointed directly into you ear - ouch! The direct field would dominate any reverberant field and there would be no spacial effect from the surrounds at all."
Why not do the analogue to toeing in the L/R's - aim them up more or less at the wall/ceiling junction of the opposite side of the room.
Impractical/ungainly with something the Nathan's size, but doable with a coax.
I also think the even power response would be beneficial.
"the software will give you proportions of total time that x-y range of frequencies exist,"
That seems like a poor criterion. Highs are crucial but very brief components of percussive sounds.
The use of dipoles was first recommended by TH because the surrounds were severely limited by being matrixed and mono, and when hard discrete effects like we get now were but a dream.
As Markus alluded to, the ambience is supposed to be in the program material and/or processing.
"With a speaker like the Nathan as a side surround it would basically be like wearing headphones. In my room they would only be a few feet away pointed directly into you ear - ouch! The direct field would dominate any reverberant field and there would be no spacial effect from the surrounds at all."
Why not do the analogue to toeing in the L/R's - aim them up more or less at the wall/ceiling junction of the opposite side of the room.
Impractical/ungainly with something the Nathan's size, but doable with a coax.
I also think the even power response would be beneficial.
"the software will give you proportions of total time that x-y range of frequencies exist,"
That seems like a poor criterion. Highs are crucial but very brief components of percussive sounds.
noah katz said:
But Mr. Holman has not changed his recommendation to my knowledge. He is the expert in film audio.
When I lived in Thailand I almost couldn't go to see a movie because they always set the surround levels too high. They seemed to believe that you had to hear them commensurate with the mains. It basically ruined the whole audio for me. Maybe I'm unique (except eveyone seems to like my setup) but I can't stand to know that there are surrounds. The action is a movie and the focal point in music is all front and center. All surrounds do is add a feeling of a larger space or the impression of "something" happening at the sides or back, but the main "image" is always frontal. There is no "image" in the surrounds. Its not like one needs to worry about "early reflections" or anything like what has to occur in the front channels. How can it ever be argued that surrounds need the same things as the mains? Have you ever seen a theater with surrounds comparable to the mains? NO!
And as far as multi-channel audio is concerned its basically not significant enough in the marketplace to have any impact. My system does two channel audio and 7.1 for film. As far as I am concerned thats all the media that there is. Maybe this will change in the future, although I doubt it, but for the time being there is no real multi-channel audio, only film.
That should solely be a property of the recording and not a property of the speakers. Historically I understand the use of dipoles to decorrelate a monophonic surround channel but time moved on.
And why is that? When looking at history then things become clearer. There was no way to do anything more sophisticated because it was technically and financially not enforceable.
"With a speaker like the Nathan as a side surround it would basically be like wearing headphones. In my room they would only be a few feet away pointed directly into you ear - ouch! The direct field would dominate any reverberant field and there would be no spacial effect from the surrounds at all."
Why not do the analogue to toeing in the L/R's - aim them up more or less at the wall/ceiling junction of the opposite side of the room.
Impractical/ungainly with something the Nathan's size, but doable with a coax.
I also think the even power response would be beneficial.
"the software will give you proportions of total time that x-y range of frequencies exist,"
That seems like a poor criterion. Highs are crucial but very brief components of percussive sounds. [/B][/QUOTE]
Ok as per your comment about toeing the speakers up to the corner, given the response you would then have on your listening axis, I don't see how this would give much benefit over Dr. Geddes or my idea. The response would now be much rolled off, much diffused, etc.
As per your comment on my software measurements, I was really reaching for ways to make meaningful comparisons, so I wont say everything I did was great. my field is child development, I use statistics to measure things dealing with that, this is a hobby for me and nothing more, so I claim no real field expertise. None the less, my reason was that I could create these proportion scores for a series of movies, and enter them as a variable into a data set which would have the same proportion scores for all the channels. You can then do things like ANOVA's or MANOVA's to compare the different channels (treated as groups) and find group differences. I know you can't just compare an entire movies soundtrack like that, of course the front has more info, so I chose front/rear and rear/front pans as my measure.
While I can't attest to knowing all that many movie studio sound engineers, I do know a few, and have toured a few of the major mix rooms. What I find is that in the main mix rooms the speakers are usually pretty junky, and often the engineers will readily admit its just for getting things roughly right. They often have screening rooms though which is used for final listening. While the larger ones don't ever have dipoles, I have seen more than one home sized screening room which did use dipole speakers. Actually, now that I think about it, I'm pretty sure that one of the smaller screening rooms at Lucas uses a THX M&K setup with tripole surrounds. In fact, I thought a lot of their mixing rooms did as well, the self powered pro models as well. A few I have been in used Klipsch THX stuff as well. Again though, in the tiny mixing rooms, typically they just use the small thx studio monitors, which aren't usually great for much.
I was in Cherry Beach Sound once for a quick tour, they have a really nice Genelec surround setup, but I think they mostly do surround music and soundtracks music stuff. Those Genelecs are pretty nice, probably be ok in a home too. None the less, as I recall, the guy sitting at the console (couldn't tell you who he was, so I won't say engineer for sure) felt that surround mixing was compromised in their setup since it didn't match what he felt should be in homes. There has been a lot of 5.1,7.1,9.5 discussions earlier, so I don't know for sure what he was referencing. My point would only be, I don't know that what studios use is necessarily was engineers are actually using when coming up with final mixes (I think screening rooms are used for that).
Why not do the analogue to toeing in the L/R's - aim them up more or less at the wall/ceiling junction of the opposite side of the room.
Impractical/ungainly with something the Nathan's size, but doable with a coax.
I also think the even power response would be beneficial.
"the software will give you proportions of total time that x-y range of frequencies exist,"
That seems like a poor criterion. Highs are crucial but very brief components of percussive sounds. [/B][/QUOTE]
Ok as per your comment about toeing the speakers up to the corner, given the response you would then have on your listening axis, I don't see how this would give much benefit over Dr. Geddes or my idea. The response would now be much rolled off, much diffused, etc.
As per your comment on my software measurements, I was really reaching for ways to make meaningful comparisons, so I wont say everything I did was great. my field is child development, I use statistics to measure things dealing with that, this is a hobby for me and nothing more, so I claim no real field expertise. None the less, my reason was that I could create these proportion scores for a series of movies, and enter them as a variable into a data set which would have the same proportion scores for all the channels. You can then do things like ANOVA's or MANOVA's to compare the different channels (treated as groups) and find group differences. I know you can't just compare an entire movies soundtrack like that, of course the front has more info, so I chose front/rear and rear/front pans as my measure.
While I can't attest to knowing all that many movie studio sound engineers, I do know a few, and have toured a few of the major mix rooms. What I find is that in the main mix rooms the speakers are usually pretty junky, and often the engineers will readily admit its just for getting things roughly right. They often have screening rooms though which is used for final listening. While the larger ones don't ever have dipoles, I have seen more than one home sized screening room which did use dipole speakers. Actually, now that I think about it, I'm pretty sure that one of the smaller screening rooms at Lucas uses a THX M&K setup with tripole surrounds. In fact, I thought a lot of their mixing rooms did as well, the self powered pro models as well. A few I have been in used Klipsch THX stuff as well. Again though, in the tiny mixing rooms, typically they just use the small thx studio monitors, which aren't usually great for much.
I was in Cherry Beach Sound once for a quick tour, they have a really nice Genelec surround setup, but I think they mostly do surround music and soundtracks music stuff. Those Genelecs are pretty nice, probably be ok in a home too. None the less, as I recall, the guy sitting at the console (couldn't tell you who he was, so I won't say engineer for sure) felt that surround mixing was compromised in their setup since it didn't match what he felt should be in homes. There has been a lot of 5.1,7.1,9.5 discussions earlier, so I don't know for sure what he was referencing. My point would only be, I don't know that what studios use is necessarily was engineers are actually using when coming up with final mixes (I think screening rooms are used for that).
Assuming that you are looking to re-create a public cinema experience in your home, it makes sense to look at the commercial gear that you are emulating.
Here is the DTS player that has been in common use at your neighborhood theater for the last 6 years: http://www.dtsdcinema.com/Support/Products/XD10/DTS_XD10.pdf
Please note that all 8 channels run flat from 20Hz to 20kHz. DTS has always been wide band in their surround channels. Older Dolby gear was band limited at 8k in their surround channels, but most of that gear has been replaced with newer full range solutions. All newer Dolby players are wide band in their surrounds.
The point of the surround channels is to create a even and difuse sound field for the purpose of psychacoutic cues. The surround channels rarely exceed 20 percent of amplitude coming out of the front channels. Cinemas typical create this diffuse field using 8 or more loudspeakers. In your home the best way to create a diffuse soundfield is to use multiple small dipole loudpeakers. You use multiple surround loudspeakers for the same reason you use multiple subwoofers. This information is common knowledge amongst 'high end' theatre designers and is described in the THX guidelines as well.
As HD Audio finds its way in the mass market surround channel mixes have gotten more complex and contain more information important to the experience. A truly diffuse soundfield coming from your surround loudspeakers becomes more important.
Here is the DTS player that has been in common use at your neighborhood theater for the last 6 years: http://www.dtsdcinema.com/Support/Products/XD10/DTS_XD10.pdf
Please note that all 8 channels run flat from 20Hz to 20kHz. DTS has always been wide band in their surround channels. Older Dolby gear was band limited at 8k in their surround channels, but most of that gear has been replaced with newer full range solutions. All newer Dolby players are wide band in their surrounds.
The point of the surround channels is to create a even and difuse sound field for the purpose of psychacoutic cues. The surround channels rarely exceed 20 percent of amplitude coming out of the front channels. Cinemas typical create this diffuse field using 8 or more loudspeakers. In your home the best way to create a diffuse soundfield is to use multiple small dipole loudpeakers. You use multiple surround loudspeakers for the same reason you use multiple subwoofers. This information is common knowledge amongst 'high end' theatre designers and is described in the THX guidelines as well.
As HD Audio finds its way in the mass market surround channel mixes have gotten more complex and contain more information important to the experience. A truly diffuse soundfield coming from your surround loudspeakers becomes more important.
Carl, a truely diffuse (homogeneous and isotropic) soundfield is something that can't be created in a acoustically small room. What advantages would a dipole have over a direct radiating speaker?
"What advantages would a dipole have over a direct radiating speaker?"
Dipoles are preferred because at any point in time more of the loudspeakers output is spraying around the room than hitting your ear. What hits your ear using dipoles has more reflected content and that is useful for psychacoustic purposes.
Picture a film scene where you are looking to re-create the soundfield of a civil war battle. Subtle accoustic cues that describe the general activity to your side and behind you should not have a point source for your brain to best construct that scene in your head.
Dipoles are preferred because at any point in time more of the loudspeakers output is spraying around the room than hitting your ear. What hits your ear using dipoles has more reflected content and that is useful for psychacoustic purposes.
Picture a film scene where you are looking to re-create the soundfield of a civil war battle. Subtle accoustic cues that describe the general activity to your side and behind you should not have a point source for your brain to best construct that scene in your head.
They say that a picture is worth a thousand words. Here is a picture describing what I was attempting to describe in words.
An externally hosted image should be here but it was not working when we last tested it.
What "psychoacoustic purposes" would that be? Dipoles make the acoustics of the listening room dominate. This is not what we want. Multichannel audio wants to create a believeable simulation of acoustics that is different from the room you're in. Only direct radiating speakers deliver signals as intended by the mixing engineer.
A dipole fires energy to the front and back. These reflections are least effective in creating a sense of spaciousness and envelopment. This is what psychoacoustic studies have shown.
There may be a compatibility problem when a movie soundtrack was mixed in a big control room with a setup very similar to a cinema and with a room curve applied. But that's not something your speakers should take care of. This is what the mixing or mastering engineer has to take care of.
A dipole fires energy to the front and back. These reflections are least effective in creating a sense of spaciousness and envelopment. This is what psychoacoustic studies have shown.
There may be a compatibility problem when a movie soundtrack was mixed in a big control room with a setup very similar to a cinema and with a room curve applied. But that's not something your speakers should take care of. This is what the mixing or mastering engineer has to take care of.
Markus
I disagree with you here.
"Dipoles make the acoustics of the listening room dominate. "
"These reflections are least effective in creating a sense of spaciousness and envelopment. "
The listening room acoustics will dominate only to the extent of the level of the surrounds. Dipoles in the cancellation plane have no direct field only a reverberant one so they add primarily reverberation to the room which increases spaciousness and envelopment by adding in late signal energy which is primarily difuse. This is why the surrounds need to be low in level so as to not overly excite the room with late arriving reverberant energy (this would not sound natural).
Reflections, after the first few will tend to be random but augmented by how the room absorption is placed. Hence if the side walls are reflective, and the ends absorptive, the reflections from the dipoles will become lateral even though they don't face those directions.
I see the role of the surrounds to be completely different from the role of the mains and see no reason at all to apply psychoacoustics applicable to the mains to the surrounds.
I disagree with you here.
"Dipoles make the acoustics of the listening room dominate. "
"These reflections are least effective in creating a sense of spaciousness and envelopment. "
The listening room acoustics will dominate only to the extent of the level of the surrounds. Dipoles in the cancellation plane have no direct field only a reverberant one so they add primarily reverberation to the room which increases spaciousness and envelopment by adding in late signal energy which is primarily difuse. This is why the surrounds need to be low in level so as to not overly excite the room with late arriving reverberant energy (this would not sound natural).
Reflections, after the first few will tend to be random but augmented by how the room absorption is placed. Hence if the side walls are reflective, and the ends absorptive, the reflections from the dipoles will become lateral even though they don't face those directions.
I see the role of the surrounds to be completely different from the role of the mains and see no reason at all to apply psychoacoustics applicable to the mains to the surrounds.
Markus, I understand you are very adamant about this, but "Says Who?" As far as I am concerned, there is no right or wrong answer, its a preference thing. If you want to argue industry standards, they do exist for some surround formats. THX Ultra 2, the newest of the THX standards, designed around the modern discrete surround channels, still dictates dipole surrounds. Thom Holman, even in his most up to date book, the industry standard book used by most academic institutions in teaching cinema sound, still dictates the use of diffuse surrounds (dipole/bipole). In fact, the newest iteration he has worked on, 10.2 requires four direct and four diffuse. Dolby Digital still considers the side speakers in even their newest digital formats as diffuse field surrounds, they should be bipole or dipole, with the rear most pear being direct. That is the industry standards that exist for home theaters and small room cinema's. Then you have DTS, they are the one exception, they are much less clear, but seem to dictate direct radiators. I would argue that actually comes from history, and is done because of DTS in theaters (DD and D PL both were develop with separate standards for home and cinema, DTS did not).
If you don't like diffuse surrounds, then thats fine, but I don't think its correct to say that direct surrounds are the current correct way to do surround. I don't agree that the rear surrounds in a small room, such as most home theaters, has the ability to create pin point imaging accuracy like the fronts, given their compromised locations. This problem is, in my opinion, made much worse when people sit off center, or worse yet, if the surrounds can not be put at or slightly behind the listener, and 2 feet above their head. Since most theaters don't stick the surrounds in the right place, and most people can't sit in the one perfect seat, dipoles can help in masking that, by created the "wrap around" effect of ambiance during the movie. Placing speakers pointed toward your head (thats what the industry standard setup guide would have you do even with direct radiators) and 2 feet above the highest listener would create, as Dr. Geddes has mentioned, a Headphone effect, but even then, only for the one person sitting directly in the right spot.
If you talk about pointing the speakers at walls, at corners, at the ceiling, then at that point you have to recognize that you are now arguing for diffuse surrounds, not against them. All of those techniques would create more reflected sound than direct (almost no direct in that case really).
If you don't like diffuse surrounds, then thats fine, but I don't think its correct to say that direct surrounds are the current correct way to do surround. I don't agree that the rear surrounds in a small room, such as most home theaters, has the ability to create pin point imaging accuracy like the fronts, given their compromised locations. This problem is, in my opinion, made much worse when people sit off center, or worse yet, if the surrounds can not be put at or slightly behind the listener, and 2 feet above their head. Since most theaters don't stick the surrounds in the right place, and most people can't sit in the one perfect seat, dipoles can help in masking that, by created the "wrap around" effect of ambiance during the movie. Placing speakers pointed toward your head (thats what the industry standard setup guide would have you do even with direct radiators) and 2 feet above the highest listener would create, as Dr. Geddes has mentioned, a Headphone effect, but even then, only for the one person sitting directly in the right spot.
If you talk about pointing the speakers at walls, at corners, at the ceiling, then at that point you have to recognize that you are now arguing for diffuse surrounds, not against them. All of those techniques would create more reflected sound than direct (almost no direct in that case really).
When talking about living rooms - which is the one room most surround material is listened to - we have no control on how absorption is placed. In that environment dipoles behave unpredictable, direct radiating speakers don't.
I would say that most of the time surround speakers are solely used to add spaciousness and envelopment to the front signals. But that is something that is already in the signal the surrounds receive and should not lead to a speaker design that tries to create those kind of signals on top of the level of spaciousness defined by the mixing engineer. The results will be unpredictable. With discrete multichannel becoming reality, there will be full range signals and then dipols will fail.
Markusmarkus76 said:
This isn't correct. It can be done ABOVE the Schroeder frequency. ALL rooms act the same above this frequency, there are no differences, its below this frequency that rooms act diferently. A "perfectly" difuse sound field is probably not possible anywhere, but a reasonably difuse one is just as reasonable in a small room as a large one - with some caveats. Sound cannot be difuse near absorption - this takes up a larger percentage of a small room than a large one and a significant reason why one should not sit near sound absorption. But the center of a small room above Fs *can be* just as disfuse as a large one.
I don't think that "homogeneous" is a requirement for "difuse". A room can be "difuse" at one point but not at another. Difuse is a point property, not a bulk property.
markus76 said:
What you are "hoping for" may become true, but its not now. And as Matt has said, like it or not dipoles for the sides are the standard.
You are arguing from a point of "should be" and we are coming from a point of "this is how it is". If dipoles are the standard AND sound is mixed using this standard then they won't "fail" until the standard changes. I don't ever see that happening for film because film uses the surrounds differently than music. Personally I see a "war" coming over this issue between the music people and the film people (to wit this and every other thread on this topic). My money is on film as the far far bigger market.
My money is on film too. But the technical requirements won't be driven by requirements rooms in cinemas have but by the requirements the typical living room has. This is because of the fact that downloads will become the main distribution channel in the near future.
gedlee said:
So what is the point you want to make? Is it diffuse or is it not? Gover has shown that there is no diffuse sound field in an acoustically small room. You can look it up in JASA 116, 2004: Gover et al. "Measurments of Directional Properties of Reverberant Sound Fields in Rooms Using a Spherical Microphone Array"
- Status
- Not open for further replies.