On reconsideration, 1000 is optimistic. It would be a lot less than this. The limitation is "hand made". You can't do anything in volume with that technique. But when you start out with 0$ investment, you don't have many options.
As to the Behringer's - there is a reason that Uli contacted me.
Using the "null" for first side wall refection only works when there is a null. The Orions did not really have a dipole pattern in the mid-HFs (where it matters most yea that argument again!!) because the baffle was too large. The best approach to avoiding the side reflections is with a narrower directivity that does not illuminate the near wall. The far wall is no where near as big an issue as the near side wall. All speakers benefit from the wall behind them being highly damped. And dipoles, as I said, have no where near the same level of directivity control as a waveguide. Better than a monopole but less than a waveguide. And I'd have to see data that shows speakers in a typical home listening room that has "nothing under 10 ms". That just stretches credibility.
I would like to see that too. Here is the impulse response of my 3-way horn loaded speakers measured 9.5 feet away at the listening position showing early reflections.
As I understand it, the goal for a reflection free zone (RFZ) is to keep the reflections lower than -20 dB relative to the main spike. I have bass traps directly behind the speakers, absorbers on the first reflection point on the ceiling, absorbers on the back wall, and thick area rug between the speakers and listening position. I added a thick underlay to further reduce the floor bounce, but not captured in this measurement.
How do early reflections affect imaging? This short video on the HAAS effect demonstrates in which one can easily hear:
The HAAS Effect! Demonstration (use headphones) - YouTube Ironically, this is used as a mixing technique to create a fake stereo image
Last edited:
Maybe my room or speakers aren't typical, but between 1 and 18mS there are just a few reflections, about 20dB down. Don't know how close to "nothing" that is. I could probably do better with some absorbers at the right spots.
I dealt with that problem (with my ORION) with absorption in the null (easy given the limited frequency range). LX521 of course resolves that issue.
I don't believe "highly damped" particularly benefits dipoles (or omnis, it probably does for most other speakers) . . . I'd suggest for that "highly diffusive" does. It delays the front wall reflections (generally longer path length for multiple reflections) and substantially enhances the "spaciousness" effect while putting that perceived spaciousness where it belongs, behind the speakers. It does this with little or no downside for "imaging".
My listening room is perhaps a bit bigger than most at about 16x25 ft, which allows sufficient spacing from front and side walls to give that 10ms+ reflection delay. In that room dipoles, dynamic driver dipoloes in particular, do it best . . . "it" being reproduction of the classical repertoire at realistic levels while producing a plausible "acoustic scene". Oh, and with good "dynamics" too . . .
For my needs they do just fine for HT and reviving memories of my "rock and roll" days . . . but I'll certainly grant that they don't play as loud as I might once have liked. I don't throw parties like I used to, either . . .
Hi Guys
I am delighted this is getting some thought and experimentation.
Ra7, actually your measurements show a great deal, I am sorry I didn’t have time to examine them when you posted them.
First, your impulse responses, note the simpler shape of the impulse of the horn but more importantly is look how many separate reflections there are and aren't for each.
The ETC view would illustrate better the “chain of arrivals” after the direct sound. (Bill W's link shows the impulse based calculation approach and mentioned the ETC (Bill, i haven't forgotten))
The MTF measurements DO show a nice difference (and I bet the two present sound pretty different too, even if they had the same frequency response).
Each MTF is done at a frequency band, say 500Hz or 1Khz etc.
The horizontal scale is the modulation frequency and the Vertical scale is the modulation depth. A “perfect” thingy would have a flat line all the way across. The more there was limited resolution, limited, the lower it would be at the higher modulations.
With optics where one has a decreasing spacing between black and white bands, here one has a decreasing time between “loud and soft” which is the increasing modulation frequency.
Now, when you compare your MTF’s, look at the shape of each curve as the modulation frequency increases. Just like optics, the limit of resolution effects the highest frequency (modulations) first.
Also, what I saw was that if I moved the mic closer, the mtf’s got better and a junky 2 way eq’d to the same response was very different than the SH-50’s in the same location and even up close.
Hi Earl
When we were all walking around with our STIpa meters at the seminar we had to be quiet, that kind of noise was part of what could reduce the modulation depth at the microphones.
It may be because the test signal has all the bands simultaneously, but one had to have “reasonably low distortion” too (whatever that is, not gross or audible? ) because that would add energy in places and times it shouldn’t be there.
Maybe you can pick that test signal apart in matlab and see what it is inside?
Hi Pano
Dang, got to run
Tom
I am delighted this is getting some thought and experimentation.
Ra7, actually your measurements show a great deal, I am sorry I didn’t have time to examine them when you posted them.
First, your impulse responses, note the simpler shape of the impulse of the horn but more importantly is look how many separate reflections there are and aren't for each.
The ETC view would illustrate better the “chain of arrivals” after the direct sound. (Bill W's link shows the impulse based calculation approach and mentioned the ETC (Bill, i haven't forgotten))
The MTF measurements DO show a nice difference (and I bet the two present sound pretty different too, even if they had the same frequency response).
Each MTF is done at a frequency band, say 500Hz or 1Khz etc.
The horizontal scale is the modulation frequency and the Vertical scale is the modulation depth. A “perfect” thingy would have a flat line all the way across. The more there was limited resolution, limited, the lower it would be at the higher modulations.
With optics where one has a decreasing spacing between black and white bands, here one has a decreasing time between “loud and soft” which is the increasing modulation frequency.
Now, when you compare your MTF’s, look at the shape of each curve as the modulation frequency increases. Just like optics, the limit of resolution effects the highest frequency (modulations) first.
Also, what I saw was that if I moved the mic closer, the mtf’s got better and a junky 2 way eq’d to the same response was very different than the SH-50’s in the same location and even up close.
Hi Earl
When we were all walking around with our STIpa meters at the seminar we had to be quiet, that kind of noise was part of what could reduce the modulation depth at the microphones.
It may be because the test signal has all the bands simultaneously, but one had to have “reasonably low distortion” too (whatever that is, not gross or audible? ) because that would add energy in places and times it shouldn’t be there.
Maybe you can pick that test signal apart in matlab and see what it is inside?
Hi Pano
Dang, got to run
Tom
Reflections and diffraction in the very early time say < 1-2 ms lie in what is called the "fusion" zone, where the signals literaly fuse into a single signal to the brain - the precidence effect does not apply in this zone since there are not two signals for the brain to act on. If the sound is mixed to a certain location with no reflections, then the reflections being fusing with the direct signal messes this "panning" up creating a different image location and stability than was intended.
Background noise will kill the MTF. But I was assuming that the home listening space would have relatively low noise, which, of course, is far from the situation in public spaces. Hence noise in our context should not be an issue. If it is then fix the noise first!
Is there a wav file available someplace? Cool Edit would be ideal for looking at that.
Thats exactly the same size as my HT, and I can't get 10 ms of reflection free signal.
Doesn't it bother you that your beloved Orions were rated as no better than a $250 set of Behringers? How do you account for that? The results didn't surprise me, but I expect that you should be having a problem with them.
One thing that I should mention - and I am very surprised that no one else has brought this up before.
In this threads subject listening comparison, I have said that the Orion was not statistically different from the Behringer. But no one has asked what the resolution of the test was. If the results were pure noise then there would never be any "statistical" difference between any two loudspeakers, no matter how good or bad they were. I never had access to the raw data to tell whether this is true or not. But one thing is true and that is that I was asked to participate, with my speakers, in this test and I declined. I declined because I did not like the test protocol. I saw it as far too likely that the results would be so spurious as to fail to discriminate anything. And then what would that prove?
I am curious though - hasn't this dawned on anyone else?
In this threads subject listening comparison, I have said that the Orion was not statistically different from the Behringer. But no one has asked what the resolution of the test was. If the results were pure noise then there would never be any "statistical" difference between any two loudspeakers, no matter how good or bad they were. I never had access to the raw data to tell whether this is true or not. But one thing is true and that is that I was asked to participate, with my speakers, in this test and I declined. I declined because I did not like the test protocol. I saw it as far too likely that the results would be so spurious as to fail to discriminate anything. And then what would that prove?
I am curious though - hasn't this dawned on anyone else?
Earl,
I actually have noticed your comments but thought that they were a personal bias or prejudice on your part and didn't want to get into an argument about it as I haven't heard either speaker. My problem with most all of these audiophile designs is that they are nothing but fancy boxes with commodity drivers be they from Seas, Scanspeak, Audax, Focal or any number of other suppliers. You can generally look at the cone and identify the device on sight. So the only real design is the box size and shape and the crossover network. I just stay out of these fights, one designer can make a speaker sound good and the next makes the exact same speaker sound terrible with a poor implementation. I didn't want to think you had a hidden agenda and kept quite and let those who own the speakers say what they say.
I actually have noticed your comments but thought that they were a personal bias or prejudice on your part and didn't want to get into an argument about it as I haven't heard either speaker. My problem with most all of these audiophile designs is that they are nothing but fancy boxes with commodity drivers be they from Seas, Scanspeak, Audax, Focal or any number of other suppliers. You can generally look at the cone and identify the device on sight. So the only real design is the box size and shape and the crossover network. I just stay out of these fights, one designer can make a speaker sound good and the next makes the exact same speaker sound terrible with a poor implementation. I didn't want to think you had a hidden agenda and kept quite and let those who own the speakers say what they say.
Well . . . I hesitate to suggest building LX521, because I'm pretty sure they wouldn't be loud enough for you. Why you're having that problem with your horns I can't even speculate.
Not at all . . . in fact I own a pair of B2030A, and they are what I am listening to as my "beloved" (did you really write that ? ? ?) ORION are in pieces being "converted" to a sort of LX521 clone. I don't know what was wrong with the ad hoc setup (or methodology) they used for that test . . .suffice to say that I don't believe that (statistically insignificant) result is replicated with the two speakers properly set up in my room. It is possible to get the "timber" and overall tonal balance close enough to suggest "no preference", but the "quality" of the bass, and yes, of the "acoustic scene", was notably different.
And, truth be told, for most of my old "rock" collection I don't believe there's enough difference to have a preference one way or the other. I could be quite content living with just the Behringers. I'm sure your horns would do just as well, and louder, too. But I can also pretty much guarantee that if I can't get my "clones" to work out I'll be re-assembling the parts back to ORION (well, my slightly modded version thereof) . . . for classical orchestral music . . . they're enough better at that to be worth it. To me . . .
It's been brought up several times (including somewhere in this thread, if I recall correctly), including your declining to participate. For whatever reason.
But I don't believe that any of us want to get into a ""x" didn't win, so they **** on the test" sort of squabble, so we let it go. The test was what it was, and it indicates that in that room with that setup and with whatever music samples they used those listeners couldn't identify a clear preference in whatever dimension they were listening for between Behringer, IMP and ORION. And that's all.
I'd love to hear a side by side comparison of the Behringers and the Magicos in the same circumstance . . . I'll put my money on there not being a $12,000 difference . . .
Earl,
Statistic don't lie but they often ask the wrong question or have statistically small sample rates that can be very inaccurate. I don't know what the test was and I didn't start the thread from the beginning and so have no idea what the test was even supposed to be looking for except for the curious title to the thread which seemed written to draw attention and cause controversy. I only started where everyone was speaking of speaker dynamics which didn't appear to be connected to the original statement.
Statistic don't lie but they often ask the wrong question or have statistically small sample rates that can be very inaccurate. I don't know what the test was and I didn't start the thread from the beginning and so have no idea what the test was even supposed to be looking for except for the curious title to the thread which seemed written to draw attention and cause controversy. I only started where everyone was speaking of speaker dynamics which didn't appear to be connected to the original statement.
I wonder if it makes any sense to compare or not. Someone that does critical listening with me never tries to compare, he just listens and tells. If a group of listeners are on par, there can be some exchange of ideas and opinions on what they would like to see improved, but that's it. We can go to different places to see what we hear in different systems that we think are good points, but generally avoid the "which is better" point. Rather, it is always, "what would I like better". Generally, we will start picking different music to see how it sounds, each person will have his own way of trying to figure out what to listen to and what to listen for to form his own opinion. Generally there is always something that someone would like to see improved, the difficulty is trying to find what is causing that what, and whether it may be improved in or whether it is just the physical limit for certain SPL level. Often the solution may not be where you think it is. The interesting thing is, after going through the process a few times, you can more quickly figure out what was done wrong just by listening, then it's the measurement and analysis to assist in determination whether or how to optimize the fix.
If I go out and buy something, I generally like to look in the more affordable range first. Recently just got a Meridian Explorer DAC mainly because it seemed like the smallest affordable DAC that meets my sample rate needs. The store said I got was the first one to buy one, and took the time to provide a demo. Did not sound good on their system, but I still got it because I heard the potential of it in my setup. No regrets either. (BTW, if anyone finds info how to open one of these up to tweak it, it is much appreciated.)
If I go out and buy something, I generally like to look in the more affordable range first. Recently just got a Meridian Explorer DAC mainly because it seemed like the smallest affordable DAC that meets my sample rate needs. The store said I got was the first one to buy one, and took the time to provide a demo. Did not sound good on their system, but I still got it because I heard the potential of it in my setup. No regrets either. (BTW, if anyone finds info how to open one of these up to tweak it, it is much appreciated.)
Last edited: