Thats another big can of worms when you are talking about small rooms. There is no concensus or even that much information. So we can bounce back and forth on what we each "believe" to be true, but the directivity topic shows that doesn't get us very far.
Basically, I don't do any room measurements except at LFs. I design the loudspeakers for the direct field and a nuetral power response via a constant and narrow directivity and let it go at that. I have measured my room, and the results are actually posted on my web site, but I did not find anything of use in those measurements, basically they didn't tell me anything new. Well let me corect that - the impulse response did tell me a lot about problem VER from the floor and ceiling and led me to the screen problem. But that was all time domain stuff. The frequency domain is not very useful.
At LFs, one has to do room measurements or you really have no idea what is going on.
I think if you try a little harder you will understand. Actually I'm a bit disappointed with your response. I was hoping for a clear statement why you feel constant directivity above 1kHz is preferable to a smooth increasing directivity.
I think it's got to do with early reflections mostly. Dr. Geddes may differ. I know little of these matters, but it makes sense in my wee brain.
Dan
I think that goes for both of us.
My explaination is given above to Markus. It does involve the early reflections as well as the toe-in. Minimizing the VER requires narrow directivity and large toe-in requires that the directivity be constant. Seems pretty clear to me.
What your describing should have been captured in the recording. Minimizing variability with listener position is obviously beneficial in ensuring not only that many people can listen and enjoy, but also, that one doesn't shift the tonal balance when they move slightly while listening. Laserbeam imaging not only limits the 'sweet spot', it means that when you move 1cm, shift in your seat, lean back, whatever, the FR changes.
And I tend to like 8" full range drivers, in the right cabs, despite this issue. Always using supertweeters and subs.
We already discussed that aspect. In psychoacoustics it's called "trading", varies from person to person and is signal dependent.
I agree that it's better then nothing but it's also nothing you can rely on (you can rely on a center speaker though). When I toe-in my Nathans I do get a more stable sound stage for the price of less defined localization. But maybe that's just because I'm sitting close to the back wall.
Best, Markus
Last edited:
My speakers and my room produce a precise and stable image throughout the entire listening area. This is not a small thing to simply be written off, but a major design challenge that only CD can achieve. "Trading" does not vary as much person to person as it varies with the speaker design. In other words, done right, toe-in will always work better to some degree for everyone than using non CD speakers and requiring on a sweet spot. So in that sense it can be "relied on".
I'm not sure that your loudspeakers provide the correct level values for trading to work. We would need level differences up to 18dB (see Eberhard Sengpiel).
Furthermore localization will be negatively affected when a recording makes use of interchannel time differences. If you're leaving the sweet spot then level and time information of a recording don't "fit" anymore. This results in a blurred phantom image.
Best, Markus
Furthermore localization will be negatively affected when a recording makes use of interchannel time differences. If you're leaving the sweet spot then level and time information of a recording don't "fit" anymore. This results in a blurred phantom image.
Best, Markus
Also, I've found at least that adding a hard center along with toe-in gives one both soundstage stability and (in recordings that have very defined images) well-defined localization. Obviously, one doesn't always want the speakers to present such images, because many recordings don't have them and in many situations (large symphonies, for instance) the real-life experience doesn't have them.
Keep in mind I'm not using Dr. Geddes' speakers but 12" Tannoy dual concentrics. The pattern control isn't quite as good as with his separate waveguides, but still way better than most stuff out there. The main benefit compared to a separate waveguide is the vertical polar response, which allows one to have all three front speakers off axis by placing them above the listening position.
I've long believed that a center channel was a must, but HOW its done is an issue to me. I did a little research on the subject and I was not happy with what I found. Those that tell you what they are doing are not doing it right and if someone is doing it right then they don't want to tell you how they are doing it, so they don't tell you anything. Left me with the impression that I better just avoid the processors for the time being. Hence two channel is still two channel on my system. Of course the 7.1 stuff coming off of BD is incredible - got that working finally.
I don't like centers personally. Except when broad coverage of a large audience is required - which most cases is already slightly compromised. The whole ITU arrangement is at odds with the fact that most people have a screen in the exact location that the center channel needs to be. Speakers like yours already have a strong phantom center which displays very little of the comb filtering effects of older speakers so I do not think it is needed personally. If there is to be a center I think why not add a new dimension and make it a height center channel.
I haven't heard a whole lot of horizontal centers that actually beat the phantom center I already have. So I skipped the center and went for 6 channel equal spaced surround sound. This makes way more sense in a keep it simple stupid sort of way to me. And I sort of wonder why it isn't a standard since it's easy to set up and has no hollow spots on the horizontal plane.
So yeah I am all about the toe in. 60 degress speaker to speaker exactly most of the time. But once in a while I will do the "less toe in" trick. It's just I don't see it as being much different than the sit a little closer trick.
I haven't heard a whole lot of horizontal centers that actually beat the phantom center I already have. So I skipped the center and went for 6 channel equal spaced surround sound. This makes way more sense in a keep it simple stupid sort of way to me. And I sort of wonder why it isn't a standard since it's easy to set up and has no hollow spots on the horizontal plane.
So yeah I am all about the toe in. 60 degress speaker to speaker exactly most of the time. But once in a while I will do the "less toe in" trick. It's just I don't see it as being much different than the sit a little closer trick.
Last edited:
It can scare you that's for sure. Honestly the imaxes I've been too haven't hit me like that - probably just the wrong ones I am going to or poor movies for the experience that I've seen. The one that I can remember being as good or better than my home system was when I saw Independence Day in some suburb outside of Milwaukee. I always forget which format that was in the alphabet soup of surround sound formats but it was minidisc/atrac based decoder that I think could do 7 or 10 channels. But outside of the specific format I just think they had some nice speakers and subs.
Once you get stuff setup to sound like that at home and you have a calibrated monitor where the colors look near perfect the home sort of has the edge over the theater since you can sit exactly in the sweet spot drink eat whatever etc.. You get a little spoiled with the small details.
Once you get stuff setup to sound like that at home and you have a calibrated monitor where the colors look near perfect the home sort of has the edge over the theater since you can sit exactly in the sweet spot drink eat whatever etc.. You get a little spoiled with the small details.
Well my point was directed at the general shape of the response. With pink noise this would seem to me to be indicative of the uniformity of reverberant field. Above about 500 Hz the measurement is basically independent of position, unless close to the speaker. Yes, the range is 120 dB, but it is also pretty obvious that the the level is within a couple of dB from 500 to 5k Hz, and rolls off smoothly after that to 20 K Hz.
How else would you look at a reverberant field other that using a signal with constant energy per octave and performing some type of time average? The display in 1/3 octave bands seems reasonable as well. I guess 1/6 ot 1/12 ocyabe bands could also be used if morre frequency resoltion was desired.
Have you done comparisons between toe-in (listening to the Summa's "outer" axis) and toe-out (listening to the "inner" axis) to see if there's a change in phantom image width?
Best, Markus
Hmm. Would you mind elaborating on what you don't like about current implementations? I assume Dolby Pro Logic II is the industry standard. It's built into most multichannel audio hardware, and it's what I use for basically all of my home listening. (Even my nearfield system is 3 channel up front.) That is to say, everything but concerts on DVD or the few DVD-As/SACDs I happen to have.
Also, if you happen to remember do you have any recordings you can recommend that would illuminate possible flaws?
That's why vertical polar response is important to me. My center is above the screen, with the center of the Tulip waveguide being maybe 50"-55" off the ground. The other two front speakers are, of course, at the same height.
While that's true, the other benefit I've found to using a center is that it lets me spread my mains much wider apart, which allows for a wider soundstage. One "knock" on the better speakers I've heard (which is to say, the ones designed with constant directivity in mind) is that none of them seem to throw images outside of the edges of the left and right speakers.
I don't imagine you would, either. The big problem I see with multichannel audio today, even more so than the masses of crappy-sounding speakers generally, is the rampant hackery that goes on with center channels. Toppled-MTMs, while being the most glaring example of audio stupidity in action, are just the top 2/3 of the iceberg.
People who care about sound never use different speakers for right and left, or put them at different heights or different orientations, so why don't they do the same with the center?
Yep. As Dr. Geddes has mentioned a few times, it's kind of like the difference between sitting right up front and sitting mid-hall. With an "overtoed" setup one can get both, simply by moving forward a little bit when the urge to sit up front strikes.
Yeah now that you press my memory
The first time I went to imax it was flat out way too bright. Similar to the bad mastering now but I didn't know it then. I was dead center in what you would think would be the sweetest spot in the theater but I wasn't getting that subsonic thrilling theatre experience it was just glaring me in the face with 5kHz. The rumor I always hear is that in most places there will be one with the A+ sound and another 2 with the C sound. Kind of a gamble which one you get. You would think though with imax it's a little bit better quality control. Ooh ooh can I answer this one? DPLII I don't think is any kind of industry standard. It's probably the most popular stereo to surround sound decoders that is on the shelves now. Most movies are mixed discretely into separate channels now. But the problem I have had with this is the mixers are way too conservative for my taste. I always thought that discrete mixers with unlimited channels would beat a decoder but in my experience I find that stereo converted to surround sound somehow more consistent.
The problem with DPLII is too much unneeded processing and too many user controls - it's like they can't decide what it actually is so they just give you all these controls instead of calibrating the thing. No offense Mr Fosgate! You'll always be a sound guru to me even if I don't like the implementations of your decoder
The reverberant field is always smooth, after all it's an average response. The narrower the measurement the more spatial averaging that you need to use to eliminate the inate frequency variance in any single point measurement. But as I said, I don't look at data like this very often. Time domain yes, frequency domain no.
On the 120 dB scale it looked more like a 10 dB variation to me, but on that scale its hard to tell.
No
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.