Member
Joined 2003
Hi Paul,
Sorry, I don't believe there's a graph on SL's website that shows the overall acoustic response. I imagine it's a nice continuation of the midrange response in the first graph (Orion-3) on this page:
(A nice, gentle, down-sloping response with no abrupt changes.)
ORION-3
The 3.2.1 shelving filter was added to broaden the transition in a way that couldn't be achieved by simply lowering the tweeter level. (The shelving filter being a first-order EQ and a tweeter reduction being second-order.) Also, the center frequency of the shelf (1.8khz) is not the same as the mid/tweeter crossover frequency.
I'm surprised this update has been misunderstood by so many people. It appears the understanding was the Orion-3 target (before adding this shelving filter) was a flat, on-axis response. Then, the 1.8khz -3.2 db filter was added to create an acoustic response identical to the electrical response of said filter. That is NOT the case.
Cheers,
Dave.
Sorry, I don't believe there's a graph on SL's website that shows the overall acoustic response. I imagine it's a nice continuation of the midrange response in the first graph (Orion-3) on this page:
(A nice, gentle, down-sloping response with no abrupt changes.)
ORION-3
The 3.2.1 shelving filter was added to broaden the transition in a way that couldn't be achieved by simply lowering the tweeter level. (The shelving filter being a first-order EQ and a tweeter reduction being second-order.) Also, the center frequency of the shelf (1.8khz) is not the same as the mid/tweeter crossover frequency.
I'm surprised this update has been misunderstood by so many people. It appears the understanding was the Orion-3 target (before adding this shelving filter) was a flat, on-axis response. Then, the 1.8khz -3.2 db filter was added to create an acoustic response identical to the electrical response of said filter. That is NOT the case.
Cheers,
Dave.
It's probably because it was deferred for so long after the addition of the rear tweeter (which is what it is primarily adjusting for).
I think that's true, but the primary reason it's been misunderstood is because there's some information missing in SL's write-up regarding the actual modifications necessary for the implementation.
Gainphile,
Please take a look at the plot in post 388 and re-read all my recent messages.
It will come clear eventually.
Cheers,
Dave.
John,
Whether it sounds better or not is beside the point.....kinda. Siegfried's feeling is (whether you agree with him or not) that a flat on-axis response for the Orion system is not optimal.
The misunderstanding here (by some) is the definition of the magnitude and frequency of the "DSS shelf" implemented in the Orion ASP.
You noted earlier in this thread that a -3db shelving adjustment is not a subtle difference. Obviously that's true, but there's more to the modifications than this -3.2db shelving filter.
If a person implements a -3.2db shelf filter and applies it to some other speaker (which may or may not have flat on-axis) response it's not an evaluation of a "DSS filter" and is a complete apples/oranges comparison.
John, since you are an Orion owner you have access to the full update information and can sim the relative difference as I did. Maybe Gainphile will believe your results?
Cheers,
Dave.
Whether it sounds better or not is beside the point.....kinda.
Siegfried's feeling is (whether you agree with him or not) that a flat on-axis response for the Orion system is not optimal.The misunderstanding here (by some) is the definition of the magnitude and frequency of the "DSS shelf" implemented in the Orion ASP.
You noted earlier in this thread that a -3db shelving adjustment is not a subtle difference. Obviously that's true, but there's more to the modifications than this -3.2db shelving filter.
If a person implements a -3.2db shelf filter and applies it to some other speaker (which may or may not have flat on-axis) response it's not an evaluation of a "DSS filter" and is a complete apples/oranges comparison.
John, since you are an Orion owner you have access to the full update information and can sim the relative difference as I did. Maybe Gainphile will believe your results?
Cheers,
Dave.
John,
Whether it sounds better or not is beside the point.....kinda.
The misunderstanding here (by some) is the definition of the magnitude and frequency of the "DSS shelf" implemented in the Orion ASP.
You noted earlier in this thread that a -3db shelving adjustment is not a subtle difference. Obviously that's true, but there's more to the modifications than this -3.2db shelving filter.
If a person implements a -3.2db shelf filter and applies it to some other speaker (which may or may not have flat on-axis) response it's not an evaluation of a "DSS filter" and is a complete apples/oranges comparison.
John, since you are an Orion owner you have access to the full update information and can sim the relative difference as I did. Maybe Gainphile will believe your results?
Cheers,
Dave.
Dave,
I ditched the Orion's long ago. But which came first, a "theory" of why the Orion should not have flat on axis response and then mods to achieve that, or the observation that when the on axis response was tinkered with the speaker subjectively sounded better, with explanation to follow? I thought the original observation didn't even stem from SL, but Don B(?). I don't know because I ddn't really follow it.
Anyway, flat may indeed not be optimum for the Orion. I have previously started why I can see that, and why the idea should not necessarily apply to speakers in general. I think the tittle of this thread is, perhaps, really overstepping the idea with the implication that the arguments for the Orion play forward to all speakers. I would not have to read what anyone says to disagree with that idea.
Last edited:
There's another, newish thread that addresses the puzzling observation that a lot of people (most?, all?) who do mic-based testing, end up tweaking by ear. And I suppose the gap between design and reality is twice as big when it is between a manufacturer's mic in their factory (or their fantasy of what beautiful, large room really is worthy of their speakers) and your ears in your room.
Is it necessary for me to be pseudo-polite (if disingenuous) and add, "... present company excepted"?
Is it necessary for me to be pseudo-polite (if disingenuous) and add, "... present company excepted"?
Last edited:
John,
Yes, I assumed you were no longer interested in the Orion's and/or sold them. What I meant was, that as a purchased owner, you still have access to the proprietary modification information and could simulate the relative difference the Orion-3 implements. I guess you're not even interested in doing that. That's cool.
Anyways, your second paragraph I agree with completely. The premise of this thread is flawed to begin with...when considering other speaker systems. Then, when the actual experimentation was attempted using an incorrect shelving filter and the results where disappointing I was really rolling my eyes.
Cheers,
Dave.
Here's my addition to the discussion. It's not exactly discussing the same question, but its close.
Techtalk Speaker Building, Audio, Video, and Electronics Customer Discussion Forum From Parts-Express.com - View Single Post - Flat Response
Flat what?!
Flat is highly unlikely to be the best way. Cold, thin, brassy... It could make a good starting point.
Depending on the listening level, you should benefit from some reverse Fletcher-Munson EQ ("Loudness Compensation") if it's done right. Listening room acoustics often deliver a VERY screwed up response to the listener position. And then there's what was called "Backward Inhibition" which stated many things regarding time delayed reflection contributions dominating, replacing or substantially coloring what the listener will hear. And then there's of coarse the issue of source program material, which I find is all over the place. There are many reasons to have tone controls, and perhaps some further active EQ catering to specific room resonances, especially in the bass frequencies. The 3 dB tilt of the entire response may well by itself be an improvement as well. A tweeter level control is the wrong way to get any kind of tilt though. It will give you a shelved dropoff with a change that is abrupt enough to color how the entire spectrum sounds. You want a variable slope tone control ideally, such as the Baxandall or James (passive version of the Baxandall).
Flat is highly unlikely to be the best way. Cold, thin, brassy... It could make a good starting point.
Depending on the listening level, you should benefit from some reverse Fletcher-Munson EQ ("Loudness Compensation") if it's done right. Listening room acoustics often deliver a VERY screwed up response to the listener position. And then there's what was called "Backward Inhibition" which stated many things regarding time delayed reflection contributions dominating, replacing or substantially coloring what the listener will hear. And then there's of coarse the issue of source program material, which I find is all over the place. There are many reasons to have tone controls, and perhaps some further active EQ catering to specific room resonances, especially in the bass frequencies. The 3 dB tilt of the entire response may well by itself be an improvement as well. A tweeter level control is the wrong way to get any kind of tilt though. It will give you a shelved dropoff with a change that is abrupt enough to color how the entire spectrum sounds. You want a variable slope tone control ideally, such as the Baxandall or James (passive version of the Baxandall).
I guess it depends on how you define "flat". There are many variables at play with this.Guys, I really find it hard to believe that SL would build a non-flat speakers (prior to DSS)... is there any hard data? It conflicts with what SL documented in the website and speeches.
The closest thing I could find is Phoenix which is flat before its psychoustic dip.
I couldn't agree more with DDF's blog.
And Humdinger: I have a vintage pre-amp with one of those complicated loudness circuits. I am able to crank down the amps enough to get the loudness circuit to actually track the Fletcher-Munson-like curves in my room. My system normally plays at "11 o'clock for kind of medium loudness listening (70-80 dBA)... (not "11" as in This is Spinal Tap).Big bass boosting (which sounds flat) at low loudness and there is little compensation beyond 1 o'clock.
Yup. Works just like perceptual theory says it should.
How come everybody doesn't twig to the essential need for loudness compensation? Scared of EQ? Are listeners arrogant enough to believe that can "hear flat" at any loudness level?
And Humdinger: I have a vintage pre-amp with one of those complicated loudness circuits. I am able to crank down the amps enough to get the loudness circuit to actually track the Fletcher-Munson-like curves in my room. My system normally plays at "11 o'clock for kind of medium loudness listening (70-80 dBA)... (not "11" as in This is Spinal Tap).Big bass boosting (which sounds flat) at low loudness and there is little compensation beyond 1 o'clock.
Yup. Works just like perceptual theory says it should.
How come everybody doesn't twig to the essential need for loudness compensation? Scared of EQ? Are listeners arrogant enough to believe that can "hear flat" at any loudness level?
It seems that different CDs are mixed to different "flat" SPL levels. I usually crank it up until it sounds subjectively "flat" to me. Is that arrogant enough?
Rudolf
I think we do without realizing it. Just like in the previous post diferent CD's sound different at different play back levels. I have my set-up set up too sound best at my average listening level. It sounds the most balanced there. Is it due too F/M?? Maybe it is at least part of it. Also depends on the program material.
Rob
I guess I don't understand why you want to scale the output via a Fletcher-Munson relation. The relation is between a physical quantity (dB SPL) and a psychological quantitiy ("loudness").
If pipe organ is playing a 35 Hz note at 40 dB SPL and a flute is playing a 700 Hz note at the same 40 dB SPL. That is the pysical reality of the recording, plain and simple. No, they will not sound equally loud (F-M curves show this) and that is also plain and simple.
Why try and make them sound "equal" (or as you guys say "flat").
Now, if both instruments were now playing at 95 dB SPL, then they will sound much closer in loudness to each other (as F-M shows us they should).
I always get confused by the notion of a "loudness" button. It is not aimed in the direction of accuracy.
If pipe organ is playing a 35 Hz note at 40 dB SPL and a flute is playing a 700 Hz note at the same 40 dB SPL. That is the pysical reality of the recording, plain and simple. No, they will not sound equally loud (F-M curves show this) and that is also plain and simple.
Why try and make them sound "equal" (or as you guys say "flat").
Now, if both instruments were now playing at 95 dB SPL, then they will sound much closer in loudness to each other (as F-M shows us they should).
I always get confused by the notion of a "loudness" button. It is not aimed in the direction of accuracy.
Please excuse my term "arrogant" - perhaps a less incendiary term would be "self-confident". But the loudness curves are a reality.
Rudolf and Robn recognize the effect and do exactly what is needed to compensate. Of course, it is no simple matter to discern what is the proper level to play things at (it is not a puzzle we try to solve anywhere else I can think of... Rudolf is figuring out what is the proper level, F-M backwards, by judging when the recording is flat... yet he wasn't present to hear the original) and something of a straightjacket to play things at THAT level... rather than choose the level based on all the other usual considerations and constraints we have.
And for people who are unable or unwilling to set their playback levels at a level that somehow corresponds to what the producers of the multi-mic recording had in mind (and so no one thinks it can actually be done anyway, change it as the mics on the various sections are cranked up by the engineer), you can look for a 50 yr old pre-amp or you can try to get a parametric equalizer do the deed kind of on average.
Withtarragon: let me ask you this. When you come out of a dark movie theater and the sunlight is experienced as profoundly bright, do you say, "I don't understand why you'd experience that as bright light, after all, light is physical and being blinded is psychological?" I believe your interpretation of F-M is unique. But don't ask me for a better one because I just don't know how it works on music.
Rudolf and Robn recognize the effect and do exactly what is needed to compensate. Of course, it is no simple matter to discern what is the proper level to play things at (it is not a puzzle we try to solve anywhere else I can think of... Rudolf is figuring out what is the proper level, F-M backwards, by judging when the recording is flat... yet he wasn't present to hear the original) and something of a straightjacket to play things at THAT level... rather than choose the level based on all the other usual considerations and constraints we have.
And for people who are unable or unwilling to set their playback levels at a level that somehow corresponds to what the producers of the multi-mic recording had in mind (and so no one thinks it can actually be done anyway, change it as the mics on the various sections are cranked up by the engineer), you can look for a 50 yr old pre-amp or you can try to get a parametric equalizer do the deed kind of on average.
Withtarragon: let me ask you this. When you come out of a dark movie theater and the sunlight is experienced as profoundly bright, do you say, "I don't understand why you'd experience that as bright light, after all, light is physical and being blinded is psychological?" I believe your interpretation of F-M is unique. But don't ask me for a better one because I just don't know how it works on music.
Last edited:
It all depends on loudness, with a bit of distance thrown in. If you listen at the same loudness that you (would have) heard the original, and at the same position relative the source, then “flat” (at the speakers) is right. If, however, you listen to the original from “out in the hall”, not sitting next to the recording microphone, and listen to the reproduction eight feet away from the speakers, the highs need to be rolled off to compensate for air (transmission) loss (if you want the recording to “sound right”). No “Fletcher-Munson compensation” is required.
If, for some unfortunate reason, you cannot listen at realistic levels then it will sound like the source has moved further away (especially if you (as you should) increase the HF cut to match the greater apparent distance). You can fool yourself (a bit) with “loudness compensation” that applies some approximation of “Fletcher Munson compensation” . . . and it will probably sound “better” that way. But it will never sound “right” . . . because something in your head will always know that that’s not the way it *really* sounds, at that level, at your “normal” listening distance. Of course this is all predicated on the notion that you have some familiarity with the “real thing”. People who listen mostly to recordings commonly end their (occasional) visits to the concert hall with a comment about how “the balance is wrong”.
If, for some unfortunate reason, you cannot listen at realistic levels then it will sound like the source has moved further away (especially if you (as you should) increase the HF cut to match the greater apparent distance). You can fool yourself (a bit) with “loudness compensation” that applies some approximation of “Fletcher Munson compensation” . . . and it will probably sound “better” that way. But it will never sound “right” . . . because something in your head will always know that that’s not the way it *really* sounds, at that level, at your “normal” listening distance. Of course this is all predicated on the notion that you have some familiarity with the “real thing”. People who listen mostly to recordings commonly end their (occasional) visits to the concert hall with a comment about how “the balance is wrong”.
Dewardh: You seem to understand F-M curves, and what they are attempting to compensate for.
But I also think we are probably in agreement that it (compensation as you aptly described it) is really not aimed at accuracy. BTW, so many of these things assume the listener has some knowlege (or educated guess) at what the original spectrum of the source actually is (or was).
But I also think we are probably in agreement that it (compensation as you aptly described it) is really not aimed at accuracy. BTW, so many of these things assume the listener has some knowlege (or educated guess) at what the original spectrum of the source actually is (or was).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.