Earl,
I don't disagree with you as much as you might assume from reading my post.
If we can all agree that no loudspeaker is perfect in all scenarios, then preferences come into play as a factor. It's rare that you can have it all.
Like political extremists that have no respect for the other extreme, objectivists and subjectivists always seem to be at cross purposes, saying the other side is always wrong, i.e., never right.
While I used the term preferences in a manner that may sound like "tastes", what I mean is the collective characteristics that bring about believability. My point about preferences wasn't about likability, but believability. For example, you may like a particular timbre of one speaker over another. However, neither may be more believable than the other, as the differences may not be that great. However, where one loudspeaker has 10x greater ability in dynamics, even though it is only 90% as good in timbre as the other speaker, it might very well be more believable over all.
I don't know how you can have absolute measurements for that, although I do applaud any attempts at finding out how.
Last edited:
Yes, I noticed their synopsis mentioned a compromise between direct and reverberent response just before I linked to it. You really should read the paper and perhaps come to your own conclusions. As I have recounted numerous times they placed a side firing dipole on top of a forward firing conventional system. Through the two speakers they could try a variety of axial and power response combinations. Flat power achieved by tilting the axial response up was way too bright. Flat power response by adding energy into the dipole was still bright (so flat power response is generally wrong). Peaks in the power response were generally audible while dips in power were innocuous. They did have one interesting case in a small room where an early reflection added distinct brightness. In the larger room the same settings were not bright (so arrival time of the added energy seem to be important.). Anyhow, worth reading.
The funny thing is that I ran into John Vanderkooy at a Toronto section meeting a few months back and mentioned how key I thought the paper was regarding speaker/room interactions. He remembered the paper but couldn't remember what they had concluded. I guess when you have written that many papers....
So if you take a more directional speaker but place it in a more lively room, Per your preference, doesn't one just counterbalance the other? If it meant that you got a longer reflection free period (ITDG?) that would likely be a good thing, but is that guaranteed? Have you measured and compared impulse responses for this approach vs. the wide dispersion, dead room counterpart?
Finally, careful with assumptions about power response being more important in a lively room. It will certainly dominate the steady state response you might measure, but that isn't necessarily the perceptual effect. The extremes of this are the usual cinema or large room PA. Measurements follow the power response and generally roll off considerably in the treble, but the systems never sound as dull as they measure.
Regards,
David
This is flipping argument for wide uniform radiation pattern upside down.
When reflection is highly dissimilar to direct sound, it forms perceptual processing stream calling for more attention. Survival value exists in making initial assessment of reflection as belonging to primary threat, so that appearence of secondary threat from direction of first threat's reflection may be made.
Yes exactly to creation of new phantom source in combination with direct sound for high level short delay. This is basic concept of fusion of sounds into perception of single source. Fusion leads to perception of source location, and timbre. When spectra differ between source and reflection timbre of direct sound is replaced by timbre of fused sound. When delay and direction of reflection grow beyond a point, fusion no longer occurs. Delay also increases as time is takes to perceive lower frequency sounds.
This leads to classic behavior of forward firing speakers in proximity to walls. Walls behave as extended waveguides for lower frequencies with omnidirectional sounds from speaker, leading to increased perceived bass.
With very omnidirectional speaker such as this:
When listened to from angle with clear view of reflection, spectral balance is unity and nearly same amplitude as direct sound. Each ear gets slightly different angle and delay time, and source appears as point on wall directly behind speaker. Timbre is indistinguishable from position in front of speaker that shades wall reflection. From this position.
Single wall acts as 180 degree waveguide.
Reflections outside of fusion zone yet not perceivable as echos contribute to sense of loudness, spaciousness, and timbrel changes.
Because when you limit the passband of each driver, the directivity is more uniform?
Being a simple country boy, I sometimes start to think why some people can't stand ambivalence and uncertainty. In life (outside audiophilia) we encounter these daily.
Why is it so difficult to accept that people disagree? Why must one try to convert the other to his side? We are not discussing of simple things in this thread.
Variation and evolution go together, there can't be one without another. Applies to techonology, politics, science, philosophy etc. Historians know many people who thought that they had found the truth... some are called gurus, some psychopats!
Why is it so difficult to accept that people disagree? Why must one try to convert the other to his side? We are not discussing of simple things in this thread.
Variation and evolution go together, there can't be one without another. Applies to techonology, politics, science, philosophy etc. Historians know many people who thought that they had found the truth... some are called gurus, some psychopats!
Well, both scenarios, additional "perceptual processing streams" from skewed reflections and tons of similar reflections means a lot of work for our little brains. Isn't the complete lack of reflections from the source what we really want in sound reproduction?
This simply isn't so; based on Toole's work, and by all accounts from people who have listened to speakers in anechoic chambers. Hearing perception evolved in highly reflective environments; be in forests, jungles, and in the ocean.
Ability to ignore what isn't relative is primary to survival with entirety of perceptual systems for all our senses.
Not sure what exactly you're referring to. Did you ever listen outside? Try it. You'll be suprised.
That's a hypothesis which sounds plausible but where's the "data" to support it?
If we forget about reproduction for a second and just consider playing music in an environment, there are a number of well accepted desirable conditions.
The room should neither be to lively nor too dead. Ideal Reverberation time grows with room size but most concert halls target 2 to 2.4 secs. They usually expand the ceiling height to achieve this. Since RT become proportional to volume it is useful to look at mean alpha. I know for cinema the range of desired target RT works out to a mean alpha of about 0.4 independant of room size. For music it is probably lower. If the room is too dead it will sound dry and volume level will be lost. Too reverberent and intimacy and articulation will be lost.
We like lots of lateral reflections as they give us a better sense of being in an involving space. Shoebox shaped halls are much preferred to fan shaped halls because they give more strength to early lateral reflections. Lateral reflections excite binaural hearing while vertical reflections confuse it.
Strong reflections must fall in a particular time window. Reflections after 50 ms degrade legibility or definition. Much later than that and they are perceived as echos which are very undesirable. In lecture halls reflections after 30 ms kill articulation or legibility of speach. Reflections before 50 ms are required to bolster strength of the music and this is a crucial requirement. We expect a certain sound volume that is appropriate for the distance from the musicians. (Technical parameter: G)
Reflected energy must follow a particular spectrum. Bass energy should be louder than midrange. Treble energy lower than midrange but not too low.
I think these characteristics scale over many room sizes, from the concert hall to the jazz club to the nice sounding home listening room. If they are a given, then the question becomes what the best reproduction system (2 channel or?) will be in terms of speaker design, room design and system layout.
Who has the answer?
David S
Why "forget about reproduction for a second" when this is the topic??? If you define the room as a part of the decoding process then we're lost. Recodings are created in all kind of rooms and they are played back in all kind of rooms. There is no standardized set of acoustical parameters for these rooms. The result is arbitrary and all discussion about speaker parameters becomes moot.
Taking comments out of context for mindless criticism is pedantry at it's forum finest.
This is an interesting (by and large) discussion, but the efforts to discredit various observations grow old and don't add much.
Having interest as a hobbyist, I acknowledge the main participants expertise here, but it behooves those carrying the bulk of the conversational weight to maintain the high ground.
Do I think that will happen? Probably not, but one can always hope!
John L
This is an interesting (by and large) discussion, but the efforts to discredit various observations grow old and don't add much.
Having interest as a hobbyist, I acknowledge the main participants expertise here, but it behooves those carrying the bulk of the conversational weight to maintain the high ground.
Do I think that will happen? Probably not, but one can always hope!
John L
Last edited:
My battle is against those that would have us shelve all measurements because "they aren't what I hear". I simply cannot abide by that point of view, so I sometimes go to an extreme to prevent it. As I said there are times when one needs to calibrate their measurements to reality, but this should not be based on simple preference, nor on any ones singular opinion. All measurements must be shown to have a correlation to something perceptive or what good are they? I have done that on several occasions and discarded some measurements that I used to swear by as simply irrelevant. I have added others. It is an ongoing process and I fell that as I get more and more "relevant" objective measures, my designs get better and better. I don't chase my tail or follow dead ends.
If you can't measure something, like "believability" then it becomes a moving target and maybe should be left alone until it is better quantified. Perhaps, and I think this happens more times than not, current measures are "as good as it gets." My money will always be on the objective as the more stable and if some things get missed, then lets look for them and quantify them. But a wholesale plunge into the depths of subjectivism is simply an easy way to drown.
Well there is that and then there is getting old and ... I forgot what I was saying ...
You see that's exactly the point, I think that it is guaranteed. Either that or I don't understand sound in rooms. I have long wanted to do a study on just this aspect, but just like John and Stan, I just don't have the energy for major research projects anymore. I do some computer simulation studies, that kind of thing, but lugging around equipment, etc. to take actual measurements in different spaces with different speakers, well that's not going to happen.
Isn't your last point precisely what John and Stan said in their paper, that the perception was both. I can buy that. The direct field is the most important, but the reverberant field does have an effect on timbre. If there is an effect then it has to scale with the ratio of the direct to reverberant energy. At no time, of course, does the reverberant energy dominate the direct - well not in anything close to a normal room.
That it varies over a narrower range is true, but it still varies. It is an improvement perhaps, but not a solution. And then there is the crossover problems which increase with more drivers. Seems to me a small gain with big problems.
It is fine to disagree and to discuss differences. It is not OK to argue for things that are known to be wrong. What a lot of people do is to argue a point not realizing that it is wrong. That's fine as long as the facts are accepted when they are presented. But often that doesn't happen. Take evolution for example. Do you realize that in the US some 40% of people still don't believe it? How is that possible? Does it make sense to argue the point when the defense is that "My pastor says it is so!" ? Or "Global warming is a farce" - really!? It is not OK to believe in things that are known to be false.
Absolutely. Did you read that because I recommended it or did you find it some other way? Yes, all audiophiles should read that book.
I think the intent of this thread (one of them) is to speculate a bit about the ideal speaker and room. Certainly Earl does that in describing a room that pairs best with his speaker design approach.
My intent is to point out that while we all have our pet theories of what an ideal speaker and room should be, it is highly possible that many of them will not satisfy the basic needs of producing (therefore reproducing) music in a space.
For example it is worth pondering that if a concert hall designer must provide wide lateral reflections for his hall to be a success, and that a pair of stereo speakers placed 50 degrees apart can't possibly create or recreate those reflections, then what quality of music reproduction can we expect?
Isn't this the bigger picture?
David S.