We all wish that this would be true but it isn't. It depends on the overall direct-reflecting ratio and the number, angle, delay, spectrum and level of reflections. The relationship of these parameters in acoustically small rooms is not sufficiently investigated. Linkwitz does not deliver hard facts for his 6ms limit.
That only shows that loudspeakers creating a similar diffuse field spectrum sound similar (the room itself plays a major role for this to be true). But do they sound accurate? Probably not when the mixing/mastering engineer used loudspeakers with a different off-axis response.
Because he only investigated "preference" and not "reference". Talking about preference is a bottomless pit.
Best, Markus
Last edited:
Below their dipole peak, acoustically small dipoles produce weaker lateral reflections than conventional speakers due to the cosine alpha polar response.
Assuming one had a 13x19' room with Orions 4' off the front wall 8' between tweeters and sat 11' off the front-wall toed in to face the listener, before the magnet/basket structure became significant the sidewall reflection would be 67 degrees off-axis and therefore -8.2dB down from the direct sound.
Obviously, the trade-off is a stronger front-wall reflection at some frequencies although in this example that's happening at 8ms with 6dB of attenuation from distance alone (plus 3.7dB from dipole polar response) versus 3.4ms for the side wall and 4ms for the ceiling. One mostly new reflection with a lot of delay seems like a good trade off for cutting two out of the three near reflections which arrive soon at relatively high amplitude.
I don't know how important directivity is at lower versus higher frequencies, although subjectively a little more toe-in is enough to avoid image shift in extremely asymmetric rooms (2' to left wall, 15' to right).
They're no more spacious sounding than conventional box speakers. One might expect that given a minimum directivity of 4.8dB instead of 0dB.
They're a lot more accurate compared to unamplified live jazz or symphonic music compared to any box speaker I've heard and various other open baffle attempts.
I have no first hand experience with how the rear tweeter or placement less than 4' off the front wall change the perception of spaciousness or accuracy for better or worse.
Last edited:
A dipole will generate a very strong lateral reflection depending on its location and toe-in:
An externally hosted image should be here but it was not working when we last tested it.
Best, Markus
Well I have a hunch some of the big players have already done some controlled tests. This is from my speakers manual - they site no sources just say it like it's a known fact to them.
"For these reasons, we do not generally recommend the equalization of monitoring systems.
So called “Room EQ” is considerably less necessary with a good Near Field Monitoring (NFM)
system. Since the whole principle of NFM is to remove the influence of room interactions
through the placement of speakers close to the monitoring position, and thus the timing and
level relationship of originating signal to any room reflections allows our built-in psychoacoustic
processing to “ignore” much of the room effect. Our standard, built-in Human
Psycho-acoustic processor is extremely powerful, and is able to naturally ignore late arriving
signals in favor of the originating source within certain time envelopes. These are generally
15ms and 10-15dB. In NFM applications the close proximity of the speakers means that any
room returns are late, thus ignored! We designed the Diamond Pro to be a naturally balanced
product, with smooth, low Q, changes in radiation pattern over its operating range. All the
transducers were purpose built and integrated into the design as integral parts. The natural
balance of the speaker is such that EQ is detrimental."
The whole idea of room treatment and in room measurements has never correlated with what I am actually hearing/perceiving. There is a whole process of error correction going on that most engineers and theoreticians are probably over looking. Linkwitz does go into greater depth on room treatment on his site and I think he punches sufficient holes in the idea that you can somehow control or absorb or predict room reflections to a useful degree in terms of technology. And it's really not clear that you have to if these claims about our brain removing reflections from the sound is true - which in my subjective experience and testing it is.
That's the problem. Everybody speaks authoritatively about what is "right" without showing any objective data that would back up their claims.
How did you test for that? You would need to do double-blind tests to generate useful data.
They just don't care enough. Otherwise there would be much more listening tests like the ones by Naqvi (Naqvi et al. (2005), “The active listening room- a novel approach to early reflection manipulation in critical listening rooms”, Journal of the Audio Engineering Society)
Of course you can but nobody knows exactly what the sound field has to look like to be "useful", i.e. optimal.
Toole calls it "adaption" - the ability to listen "through" a room's acoustics. Another issue that needs further examination.
Best, Markus
Yeah I used to just call it acclimation so people wouldn't think I am crazy. But if you start to break down why you perceive certain things a certain way you can easily come to the processor/error correction solution. It makes sense out of a lot of things that I have personally experienced with sound.
I've come up with a bunch of anecdotal ways to test it with surround sound and just trial and error. It's actually kind of easy to test for perception of reflections with surround sound even though no one would ever accept my conclusions.
I've come up with a bunch of anecdotal ways to test it with surround sound and just trial and error. It's actually kind of easy to test for perception of reflections with surround sound even though no one would ever accept my conclusions.
Last edited:
Yes, our brain is probably pretty good at making sense of the sounds that reach our ears. Just like our eyesight under different ligthing conditions. But to achieve the most accurate reproduction there needs to be a standard like "D65". We don't have that (yet) in audio reproduction.
Best, Markus
Best, Markus
Anyone know what type of gating is appropriate to measure loudspeakers with? Say you are 3ft away from your elevated and in the center of the room speaker. You gate is set at 6ms. Should you use hamming, Tukey, flat-top, etc... ? Anyone know how these different types of gating work? Some of them allow more low frequency content, some less, some smooth, some rough. That's the best I have right now.
Thanks,
Dan
Thanks,
Dan
Exactly. I was actually thinking closer to ICC profiles though. And to tell the truth I just now read about that standard but you know what conclusion I came to independently with my own monitors, eyes, and colorimeter? That 6500k is the most natural white point to calibrate to if you are doing color correction. See my senses aren't that off in terms of judging spectral balance and I have anecdotally proved this to myself over the years. Just another little instance where my intuition seems to somehow prove to be empirically correct.
Gating in the time domain on an impulse resonse just needs to be truncated, but programs like Holm taper it off gradually.
As to the perception of sound in small rooms, its a tough subject for which almost everyone with any expertise in it disagree on the details while we all tend to agree on the major points (which are dramtically differnt that the prevailing dogma). I think that Linkwitz is probably the most extreme in his position and also probably the least experienced - audio is his hobby, it was never his occupation like it has been for Toole, Olive and myself.
As to the perception of sound in small rooms, its a tough subject for which almost everyone with any expertise in it disagree on the details while we all tend to agree on the major points (which are dramtically differnt that the prevailing dogma). I think that Linkwitz is probably the most extreme in his position and also probably the least experienced - audio is his hobby, it was never his occupation like it has been for Toole, Olive and myself.
Well this makes him more credible in my book. I find that artists who's goal is to impress themselves first tend to be the ones which go the furthest because they don't have to consider what they do as marketable or fashionable. And I mean the guy had his hand in pretty much solving one of the hardest problems in audio already and he wasn't in the game back then.
Anyway it's just a kooky idea I had popping around in my head. And since I am sort of searching for a bare minimum I tend to think that I might have one of the most extreme opinions on what can be judged negligible - waiting for someone to ask me "is there anything you think is important?!" And well I have zero credibility in a professional capacity.
I tend to think the audio world is too caught up in chasing perfection to get around to consider making it easily accessible to the little guy. I don't trust any of the big manufacturers that have been testing this stuff for close to a decade will be offering up too much insight because it's too profitable to keep a level of confusion in the market. I could be wrong about some of this stuff but I want to know what I am wrong about. And what needs to be done just so someone can make and playback an album with more ease and precision. Right now it's a pointing game if your recording doesn't sound right - the speaker maker blames the mastering engineer, the mastering engineer blames your room or the mixing engineer who then blames the musicians. At some point we need to sort it out
An honorable statement, but not one that I have found to be true in practice. People seldom appreciate your pointing out where they are wrong. Read Toole, but understand where he is coming from - preference NOT accuracy and he works for a manufacturer. he also listens exclusivley to classical music and holds the recreation of space as of preminate importance. Imaging is not his thing.
I went to hear Messiah the other day. When I closed my eyes there was no "image". There was no localization of instruments. The room and the "spaciousness" swamped out all imaging aspects. This would be quite different in a small Jazz club for example.
I think this is possible to capture and playback possibly independent of certain deviations in the polar pattern. It's just figuring out what mic placement, speaker placement, and possibly what decoder. It's just recorded music is bit trickier than making a reflection filled concert hall sound coherent.
Haha. I have actually thrown quite a bit of my audio belief system out the window so a couple more things being changed isn't that big of a deal to me. Thinking back on aesthetics that I liked and got somewhat snobbish about as a teenager is kind of an embarrassment. I had a million ** reasons why I thought the crappy music I heard was crap including a lack of creativity, drum machines, syths, samples etc.. etc.. I have since programed my own beats realized that a great deal of the music I loved before I got snobbish used all of these things. So maybe with me it's more likely that I will become more squirrely and latch on to my next fixation/obsession 🙂
well these aren't done with Holm--maybe next week--but they are truncated and look terrible. 1 meter, 6ms gate for all.
on axis:
11.25 off axis:
22.5 off axis:
33.75 off axis:
45 off axis:
56.25 off axis:
67.5 off axis:
and then finally a 90 degree off axis shot:
I feel like I shouldn't even post these, but I did them and they should be more useful than before. If not I won't ask you to waste your time. The bass reduces nicely as you go off to the side. but the 2-4k hump goes nowhere. I'm sure that's still not great data, but it's better than before. A step in the right direction. Overall the shape doesn't look that different from the last graphs except the bass disappears.
So if the rest of the speaker had the same behavior as the bass, I'd be in luck?
Do they make absorption that can get down to 2kHz? Maybe a lot of that out to the sides would be useful? or a waste of time/money?
Back to the drawing board,
Dan
on axis:
11.25 off axis:
22.5 off axis:
33.75 off axis:
45 off axis:
56.25 off axis:
67.5 off axis:
and then finally a 90 degree off axis shot:
I feel like I shouldn't even post these, but I did them and they should be more useful than before. If not I won't ask you to waste your time. The bass reduces nicely as you go off to the side. but the 2-4k hump goes nowhere. I'm sure that's still not great data, but it's better than before. A step in the right direction. Overall the shape doesn't look that different from the last graphs except the bass disappears.
So if the rest of the speaker had the same behavior as the bass, I'd be in luck?
Do they make absorption that can get down to 2kHz? Maybe a lot of that out to the sides would be useful? or a waste of time/money?
Back to the drawing board,
Dan
Dan
You are learning fast. Rooms tend to average out things yielding much better looking data and sometimes worse, but never very meaningful. As you can see things are not so pretty when you see them 'in the light" - kind of like an aging hooker in a low class brothel.
Now the problem is still that 6 ms is likely to contain some reflections. You just can't guess at the window time you have to set it precisely and Holm does that for you. Let me know when you get some Holm data and I'll comment, but your already learning that things are no so clear as they once appeared to be.
You are learning fast. Rooms tend to average out things yielding much better looking data and sometimes worse, but never very meaningful. As you can see things are not so pretty when you see them 'in the light" - kind of like an aging hooker in a low class brothel.
Now the problem is still that 6 ms is likely to contain some reflections. You just can't guess at the window time you have to set it precisely and Holm does that for you. Let me know when you get some Holm data and I'll comment, but your already learning that things are no so clear as they once appeared to be.
Sure, what don't we agree on? If you mean Linkwitz, I agree with most of his principles - like CD, etc. - but certainly not his implimentation. Once when I saw him speak on his concepts, he started with an assumption of 85 dB listening levels. I asked what happens to all his concepts if 85 dB isn't enough. He responded that "I never listen louder than that". Well I guess that kind of puts things in perspective. You can't expect a derivation of anything to hold when the assumptions don't hold. And listening to his speakers, they failed miserably when the levels got higher. Thats pretty well universally accepted that the Orions are strictly a low level system. I like the option.
The indirect sound field created by your (CD), Linkwitz's (dipole) and Toole's (smooth directivity) loudspeakers doesn't have much in common.
Best, Markus
yea, this time window thing is shedding a lot of light on this subject and a couple other subjects as well--as do the polars of course. The only thing I actually like about what I've measured is the crossover now that I think about it. Most people think that's the hard part. It ain't--if you go active and have a lot of overlap to play with. I think you can fix a bad crossover with DSP--perhaps not as elegant, but......... This has been elucidating though I'm not out of the mud yet.
Thank you again Dr. Geddes. You may not think of it this way, but you are a true friend. You've been harder on me than I have been on myself and I've learned more about audio in the last week that I would have imagined. People who talk on these forums should really do their own research. I few hours of doing has to be worth a decade of conjectures.
Dan
Thank you again Dr. Geddes. You may not think of it this way, but you are a true friend. You've been harder on me than I have been on myself and I've learned more about audio in the last week that I would have imagined. People who talk on these forums should really do their own research. I few hours of doing has to be worth a decade of conjectures.
Dan
Last edited:
- Status
- Not open for further replies.