I will disagree with that in as a matter of degree. "All"? No. But on the very best systems I've heard there was very little need or desire for tone control or constant level riding - as on most systems. That would be on the great majority of recordings, but of course not all.
(I include the room in the system.)
Don't go off on me here Davey. 🙂 But that is what I see in the PDF that was linked. That is what SL said. That was his presentation!!! That is the point of contraversy.
Here is the figure from the presentation:
An externally hosted image should be here but it was not working when we last tested it.
What does it say? Eq arrived at by listening. Then it shows "Flat W and T setting", V 0 (-1.8dB), V 1 (-4dB) and V 2 (-3.2dB).
But you tell me there is more to the story. Ok, I went to his site and looked at the new transfer functions and then overlaid them with the old ones. They are SL's measurements. I'm not showing the entire transfer functions but here for obvious reasons, but here is the pertanent difference.
An externally hosted image should be here but it was not working when we last tested it.
This is a far cry form what he is presenting in his paper and the idea the flat is not correct. In fact other than the correction to the midrange response the change in the tweeter response looks more like a simple correction for the rising response of the T25, something that was address in the NaO II from day one by the component L4 in the passive tweeter network, and similarly addressed in my all active version.
An externally hosted image should be here but it was not working when we last tested it.
You know Davey, we have had some interesting discussions over the years both public and private, but you have always been a straight shooter. And I thank you for insisting that I look at the details here. But I just don't know what to make of this.
Last edited:
John,
Yes, your overlay plot looks correct. And you'll notice it's the same as the plot I posted in message #388. 🙂 (Which, apparently, nobody looked at.)
SL's description of this whole matter is confusing.....as I've said previously. But you have to read his comments literally and not read anything into them......as many have been doing on this thread. 🙂
I agree there's an apparent conflict between the theory he's disseminated in the PDF's/presentations and the actual implementation in the Orion-3. You have to read his Orion-3 page very closely.
Or, the whole "flat is not correct" concept is just a big subterfuge to effect a correction/change in the Orion system response? 🙂
Cheers,
Dave.
Yes, your overlay plot looks correct. And you'll notice it's the same as the plot I posted in message #388. 🙂 (Which, apparently, nobody looked at.)
SL's description of this whole matter is confusing.....as I've said previously. But you have to read his comments literally and not read anything into them......as many have been doing on this thread. 🙂
I agree there's an apparent conflict between the theory he's disseminated in the PDF's/presentations and the actual implementation in the Orion-3. You have to read his Orion-3 page very closely.
Or, the whole "flat is not correct" concept is just a big subterfuge to effect a correction/change in the Orion system response? 🙂
Cheers,
Dave.
Last edited:
I honestly want to see the truth in that. I just can't find any technical basis that would support it. Maybe my ears aren't that good anymore.
But what about?
I remember a time several years ago when I popped into a Hi-Fi store and listened to some $5000 speakers. they sounded pretty good. Maybe a bit thin. Later that day a friend invited me over to see slides of a trip he went on, with music as a backdrop. He had four speakers sitting on the floor away from major room boundaries, cross-hooked from a stereo source. These speakers sounded way better to me than the $5000 speakers I heard earlier that day. I asked what they were. They were Radio Shack 3 way... I took it as one more demo of how room acoustics dramatically affect the outcome of any speaker.
Linkwitz is IMO one of the best speaker engineers in the world. His honesty about tweeking the Orion-3 appears to confuse many. People are always trying to simplify things by closing their minds to the vast array of variables at play in any real world system.
When I got heavily into photography at age 12, I remember being frustrated by how film didn't compensate, pixel by pixel, for contrast/lighting changes, like the human eye-brain mechanism does. I had to learn to compensate for the limitations of the reproduction process. With audio, mics have certain idiosyncrocies that the ear-brain mech doesn't have. Also, the recording environment has all the same acoustic problems that exist in a listening room. If you try to record acoustic music or anything where you know what it is supposed to sound like during reproduction, you'll notice that it's pretty much impossible to get it exactly right. You do the best you can, which IMO can be plenty enjoyable. But pretending that a system calibrated to be "flat" should be as good as it can get, seems pretty naive and closed minded to me. It's never that simple. If it seems to be, then it was probably by luck, that the acoustics of the listening room complimented all else, or the listener just doesn't realize how much better it could sound with further adjustment of tone.
I remember a time several years ago when I popped into a Hi-Fi store and listened to some $5000 speakers. they sounded pretty good. Maybe a bit thin. Later that day a friend invited me over to see slides of a trip he went on, with music as a backdrop. He had four speakers sitting on the floor away from major room boundaries, cross-hooked from a stereo source. These speakers sounded way better to me than the $5000 speakers I heard earlier that day. I asked what they were. They were Radio Shack 3 way... I took it as one more demo of how room acoustics dramatically affect the outcome of any speaker.
Linkwitz is IMO one of the best speaker engineers in the world. His honesty about tweeking the Orion-3 appears to confuse many. People are always trying to simplify things by closing their minds to the vast array of variables at play in any real world system.
When I got heavily into photography at age 12, I remember being frustrated by how film didn't compensate, pixel by pixel, for contrast/lighting changes, like the human eye-brain mechanism does. I had to learn to compensate for the limitations of the reproduction process. With audio, mics have certain idiosyncrocies that the ear-brain mech doesn't have. Also, the recording environment has all the same acoustic problems that exist in a listening room. If you try to record acoustic music or anything where you know what it is supposed to sound like during reproduction, you'll notice that it's pretty much impossible to get it exactly right. You do the best you can, which IMO can be plenty enjoyable. But pretending that a system calibrated to be "flat" should be as good as it can get, seems pretty naive and closed minded to me. It's never that simple. If it seems to be, then it was probably by luck, that the acoustics of the listening room complimented all else, or the listener just doesn't realize how much better it could sound with further adjustment of tone.
You said it, not me.
But SL does have this need to justify things rather than just saying it sounds better this way.
Simple answers generally indicate simple (mis)understanding, or simple mistakes . . .
My own experience tells me that indeed "flat is not correct". But the HRTF theory accounts (at best) for only part of the reason. There's the room, and the difference between our "norma" listening position and microphone placement, and the room, and listening levels ("Fletchre-Munson" compensation), and the room, and the vagaries of perceptual accomodation and adaptation, and, did I mention the room?
There's a lot of "exploring" left to do . . .
You said it, not me.
It's called being a good student. Having the desire to understand the "why". That's why he's one of the best. 🙁
OK, so for those of you who have follow the research, read the papers, studied the studies.
Is the flattest speaker the preferred speaker in an anechoic or nearly dead room?
If it is, is it still preferred in more normally reverberant rooms?
How does it measure in-room as opposed to anechoic conditions?
Is the flattest speaker the preferred speaker in an anechoic or nearly dead room?
If it is, is it still preferred in more normally reverberant rooms?
How does it measure in-room as opposed to anechoic conditions?
And this is because Orion with rear tweeter probably never was flat because it would have been too bright (and this is not a disadvantage of Orion, it is a technical implication). So for Orion, the DSS is an improved response shaping (along with other mods) compared to just cranking the tweeter level down.
But the DSS is applicable to speakers that are at least similar to Orion like Gainphile's and mine and you can apply it to a flat response. If you prefer -3.2dB or -3.0 is another story.
So in general, I totally agree with Gainphile. This is the data that is being presented to non-Orion owners and it seems you need a pinch of salt with it.
And the thread (being a question) is perfectly valid. Pluto does not use the DSS AFAIK but SL stated when asked that it is not flat (anymore ?). And there are many other speakers that are not flat from bookshelf to tower. Having said that, maybe Orion 3.2 in its entirety with a monotonically falling response makes it even more similar now to Pluto. Who knows.
And yes, it is definitively confusing what has been published.
Pano:
1 Yes, except in the bass, where it is a case of 'adjust level to taste', although bass is also preferred without peaks and dips. CAVEAT: 'flattest' is ill-defined; a flat on-axis speaker is not preferred if it has a non-smooth off-axis response.
2 Yes.
3 For a typical well executed forward firing speaker in a typical listening room, the ungated response falls smootholy from low to high frequency, excluding the bass zone which is highly variable with room dimensions, speaker positioning etc.
1 Yes, except in the bass, where it is a case of 'adjust level to taste', although bass is also preferred without peaks and dips. CAVEAT: 'flattest' is ill-defined; a flat on-axis speaker is not preferred if it has a non-smooth off-axis response.
2 Yes.
3 For a typical well executed forward firing speaker in a typical listening room, the ungated response falls smootholy from low to high frequency, excluding the bass zone which is highly variable with room dimensions, speaker positioning etc.
Theoryetrically, IMO, you could argue that flat response in an anechoic chamber is the right default goal. But there's still the reality of ear frequency response changing with level, and how it's intended environment is likely to affect what's heard at the listener position.
Reverb from room acoustics is always frequency selective. Each room has it's own "signature". Generalizations about the effects of room acoustics and how they color sound (resonances, comb filter effects from reflections, different psycho-acoustic effects depending on how delayed a given reflection is or how long a given resonance rings) can be misleading. If the room caused a substantial resonance at one frequency, it could be argued that you want to cut stimulation energy at that frequency more than a cal'd mic with pink noise might suggest.
I'd say that the anechoic chamber is a good reference, especially for testing or verifying the electro-mechanics of the speaker itself. It's pretty rare though that a speaker will be used in such an acoustical environment. Linkwitz, in his realtively large listening room, recommends "normal" amounts of room reflections rather than a completely dead room. I generally agree. I feel that although it could be argued to be a step away from pure "fidelity", room reflections can put some life and 3-D nature back into a reproduction that has been made stale by too much technical perfection mixed with the idiosyncrocies of recording with a microphone. Mics don't usually hear the same way as the human ear.
When I design a speaker, one of my first questions is how far from significant room boundaries is it likely to be? Once it's done and in, I'd want to electronically EQ down any resonances in the lower mids and bass, and maybe pump up the lowest bass so the acoustic response is flat to below 30HZ, if the speaker is capable of soing that well. I would not try to pull up cancellations in the frequency response at the listener position because it would make things worse elsewhere in the room, and it would make them more audible since it can't really be successfully done. Generally, I'd use high resolution EQ in the bass and lower midrange freqs. to reduce anything that looks like a resonance, and more gradual EQ at the higher freqs. to get a variation of reverse Fletcher-Munson that works for me.
Linkwitz is not always the best communicator. But who is? With an Open Baffle speaker, you really have no choice but to play some kind of game, so to speak, with how room acoustics are likely to influence the sound. That rear wave will add at the listeners ear with roughly 6-10mS of delay, which is a recipe for a comb filter effect. But where cancellations will occur is totally dependant on speaker placement in the acoustical environment, and the angle you're at relative to the straight fire. If I understood correctly, Linkwitz put the 3kHZ dip in so the 45 degree off axis response would be more even, so the overall room response including room reflections would feel more balanced. An O.B. speaker is definitely a challenging project, but the way it brings to life the individuation between instruments, and puts a 3_D nature to any embedded reverbs, gives me a thrill... Once bitten... But they are no good in small rooms where the rear output reflections bouncing off the front wall will be less than about 6mS IMO. That delay time has undesirable psycho-acoustic effects.
Theoretically we can argue anything, but there is no need to argue because EMPIRICAL evidence clearly points to the conclusion that flat anechoic FR is the right default goal, although it is not sufficient by itself; there are other important parameters too.
Agree with this in the bass region but largely not a factor above about 300Hz.
Totally agree, it would be a big mistake (inconsistent with psychoacoustic preferences) to set an anechoic environment as one's listening goal.
Agree, providing the measurement is at the listening position.
Agree outside of the bass region, but in the bass it is desirable to correct for cancellations provided your speaker and amp can cope.
Don't try to reverse FM, it is a mistake. Go for flat gated FR at the higher frequencies.
Last edited:
Really? In anechoic conditions? Are you sure? Why would the off axis have much, if any effect there? Or do you mean something else?
So was it found that an anechoic flat response speaker had a falling response in the typical listening room? That's gotta depend on the room and power response of the speaker, but there was a trend? Any more data on that, or can you point me to it? It's very interesting.
Sorry, I misread your question. Off-axis sound doesn't have an effect in anechoic conditions. I thought you were asking if the flattest anechoic speaker is the preferred speaker. I am not aware of preference ratings inside an anechoic chamber. It would be a strange thing to do, choosing between yuk and yuk.
Agreed. Like I said, typical speaker, typical room.
Hmmm
Because of the variability of ear sensitivity at high frequencies from person to person, I would default to a flat response as a starting point, and then using variable slope tone controls (not shelving) I would tailor it to my own taste; what sounds right to my ear. If I'm doing it for someone else, I'd set tone for a gradual rolloff starting around 400HZ, and try a 2-3dB/oct. rolloff. Just a personal opinion based on my experience. No technical reason that I know of, other than maybe it seems that most rooms tend to reinforce the frequencies below about 400HZ a bit more.
Because of the variability of ear sensitivity at high frequencies from person to person, I would default to a flat response as a starting point, and then using variable slope tone controls (not shelving) I would tailor it to my own taste; what sounds right to my ear. If I'm doing it for someone else, I'd set tone for a gradual rolloff starting around 400HZ, and try a 2-3dB/oct. rolloff. Just a personal opinion based on my experience. No technical reason that I know of, other than maybe it seems that most rooms tend to reinforce the frequencies below about 400HZ a bit more.
Well yes, it might be. But I was just wondering if a certain response curve was found to be preferred in an anechoic (or very dead) space, how would those same speakers measure in a typical listening room?
Though it may not be pleasant to listen under anechoic conditions, at least would we be sure that the room had no influence. Headphones are mostly immune to room effects, of course, but they have other problems.
Certainly it's more complex than that, but I was hoping for at least a baseline to start with.
- Status
- Not open for further replies.