Small note while I digest the posts since my last.
NOBODY from the "buy me a better meter" camp seems to have understood my point... which really should have been pretty obvious:
there is no such thing as a "noise meter" because noise is a perceptual experience.
You can buy SPL meters and very crudely weight them "A" (somebody's odd choice of one of the equal sound level curves, dunno who made the choice, eh) and you can do a "noise and number of times" weighting and you can add a "time of day" weighting, and more weightings.... until you kind of think you are reasonably near a measure more than half of your friends are willing to call "noise level."
Got the point?
I also pointed out, it is hardly feasible to do suitable human testing all the time. So ultimately you are trying to find physical measures that kind of correlate, on a good day, with the range of human responses.
I've worked on airplane takeoff noise. Ten minutes of studying the painful historical debate about airplane noise measurement and you will end up laughing or crying.
Never the less, that is where we need to go in order design and predict stuff for humans. But the physical measure is ALWAYS only a crude and untrustworthy stand-in for the perceptual measure.
NOBODY from the "buy me a better meter" camp seems to have understood my point... which really should have been pretty obvious:
there is no such thing as a "noise meter" because noise is a perceptual experience.
You can buy SPL meters and very crudely weight them "A" (somebody's odd choice of one of the equal sound level curves, dunno who made the choice, eh) and you can do a "noise and number of times" weighting and you can add a "time of day" weighting, and more weightings.... until you kind of think you are reasonably near a measure more than half of your friends are willing to call "noise level."
Got the point?
I also pointed out, it is hardly feasible to do suitable human testing all the time. So ultimately you are trying to find physical measures that kind of correlate, on a good day, with the range of human responses.
I've worked on airplane takeoff noise. Ten minutes of studying the painful historical debate about airplane noise measurement and you will end up laughing or crying.
Never the less, that is where we need to go in order design and predict stuff for humans. But the physical measure is ALWAYS only a crude and untrustworthy stand-in for the perceptual measure.
Yes actually. 😀
I came across this yesterday (see page 4 of 9, fig. 3.):
http://www.moultonlabs.com/images/pdfs/new_loudspeaker_design.pdf
Also look to page 5 of 9 figure 5.
I'm guessing they sound so "bad" because of the measurements, No? 😉
Still don't understand your point. Do you say it can't be done or do you say it's not easy? I'd agree to the latter.
Even if the measurements are great(certainly don't look bad), it just shows the bias of sighted listening. You just can't remove it. Hopefully that's what you were getting at--what I've been saying all the while.
Is that what you were saying Scott?
You should read my next couple posts after the two you quoted.
SUM, which authors?
Dan
for not wanting a negative that was a pretty negative post.
I never get how someone can slam other people work when that work is backed by proven experience, published papers and backed by other PhDs in the same field.
My only question at this point would who is SUM? You do not even have a real name or a website, You do not even build speakers for others to learn from. You do not publish papers for others to learn from ( I did read some paper you published was actually rejected, go figure!!)
Did I miss anything else?? Im pretty sure Toole does not give a crap about anything you think (you are not known in audio period!!). Toole has simply helped thousands of people that want to improve their audio experience. How do I learn? Some no name poster with incoherent ramblings online or solid fundamentals from Toole? Thats a tough one......
I suggest you start by producing a real product or a real book before you slam him again 😉
Last edited:
My impression is that you are correct when you claim you don't understand the point.
"Noise" is a human perception. There is no "natural" noise. Ever wonder why a an SPL meter works 20-20kHz and has a funny knob that selects A weighting? Ever wonder about the tech definitions of amplifier noise and why the IEC likewise has weightings tied to their concept of human annoyance? Shouldn't they use noise from DC to light?
You are simply mistaken in thinking that it can be done. It can't be done by a machine except in so far as somebody has created a machine that closely resembles human judgment(s) under all the conditions for which you'd want to use that machine. If I were given to capitalizing paragraphs, I'd emphasize this paragraph.
Without wanting to seem fatuous, that old saw about the tree falling in the forest: does it make noise?
Footnote: markus may be confusing "noise" as used in music reproduction with things like Brownian motion, random error, and various abstract uses of the word. They are valid uses in their context but here only confuse those who are allergic to human testing concepts.
Even if the measurements are great(certainly don't look bad), it just shows the bias of sighted listening. You just can't remove it. Hopefully that's what you were getting at--what I've been saying all the while.
Is that what you were saying Scott?
..
Dan
Actually no. 😱
Rather that how "well" a speaker measures with regard to amplitude on and off axis - is no guarantee of success for any particular individual.
Further, while I don't dismiss sighted bias (or any bias for that matter), there is always the question of: IF the response was, or was primarily, the result of such bias.
In other words I don't assume.
I don't assume that a person's response of "bad" sound is necessarily the product of any one thing, or even one thing more than another.. like visual bias.
In fact, such an assumption IS a bias. 😉
I've been wondering that for a long time myself, but in the end it doesn't really matter because I quit listening a long time ago.
If i interpret the polar measurements in horizontal dispersion right,
there is a maximum narrowing about 0.7 Khz and the dispersion then
gets wider and more even towards higher frequencies.
http://www.moultonlabs.com/images/pdfs/new_loudspeaker_design.pdf
(page 5 of 9 figure 5)
Even from the measurement point of view, the frequency
dependent dispersion of the B&O speaker using the
shown reflector/diffuser would in no way fit into the preferred
characteristics derived from the Toole papers.
The discontinuity in midrange dispersion is not good
and then the widening tendency towards higher frequencies
does not make things better.
And i want do admit that, even though i never did scientific
listening tests on that matter, the findings in the toole papers
concerning
- preferred on axis (and in room) response
- preferred dispersion characteristics (smooth deviation in off axis response)
meet my experience very well.
I would not expect that speaker to sound "well" especially at
voices or acoustic instruments even when looking at the
measurements only . Maybe a prejudice ... but it has an even
worse frequency dependent dispersion than a common cone & dome
2-way.
Last edited:
Was that a "slam"?
This is a "slam":
I've been re-reading Toole's Sound Reproduction ..again...
It's cr@p, it's based on cr@p, and frankly.. because of the paper milling process - it smells like cr@p. I wouldn't wipe my @ss with its pages - because well, who want's to clean up cr@p WITH cr@p?
THAT's a *slam*. 😀
Seriously though, while I like the book and mostly agree with the conclusions - much of the work surrounding those conclusions (at least from Toole and Toole & Olive) were based on a significantly questionable tests - testing that IMO would not withstand sufficient scrutiny in most legitimate scientific circles.
Moreover the editing leaves a LOT to be desired.. grammatically of course it's excellent, but several "key" concepts are significantly lacking in the clarity of their delivery to the reader. In particular about a month ago on the Geddes/Waveguides thread both Dennis/catapult and Zilch seriously misunderstood a fundamental concept Toole was *trying* to explain to his readers.
Sure, it's a "must have" because there is nothing else quite like it.. but is it really "all that". I don't think so.
..and that's not a slam. 😉
Toole did an end run around 50 years of subjectivist speculation as to the relationship between listening preferences and measurable performance, namely, the non-existence thereof.
Specifically?
Specifically?
Last edited:
Maybe you should improve your didactic skills?
"Noise" is a concept not a perception. Like there's no perception of "color". There's the perception of red, green, blue, etc. but not the perception of "color".
The rest of your post results from this false premise. Have you ever heard of color calibration? If anything you said is true, then color calibration can't exist. But it does.
I hope that you noted that those polars were normalized and the results for the listener axis shown later in the paper were very bad. Overall, I would not release a speaker like that and its certainly nothing that could be hearalded as exemplary.
Actually its used both ways, which, I believe, is part of the point. When used perceptually it has no meaning because it is only defined objectively - but its used both ways! Of course its going to be confusing.
Maybe that's why bentoronto is confused about relating perception to measurements.
http://www.diyaudio.com/forums/multi-way/103872-geddes-waveguides-454.html
4534 and 4535.. and that's IF I'm correctly interpreting both of your posts.
Hi Oliver. I agree, but I've seen worse measurements as well, and listened to them with a less negative impression. The measurements look better than I remember the speaker sounding. I'll say that. Their listening space was less than optimal however and the recorded material was run of the mill elevator Jazz stuff that I can't stand and gets plenty of play in Tokyo's Audiophile driven stereo shops.(I'm probably missing some commas in there🙂 The power response looks like it should have sounded good, but it's not enough information. I wish they hadn't normalized their polar plots--sure makes it hard for me to clearly read, but I'm no expert. Their listening room was untreated, flat shopping center walls, long and narrow and had no seating with plenty of glass display cases throughout. IOW, very wet and reflections can get the better of me. My wife says I should remove the mirrors from my view. WT.....? After all is said and done, it was an irritating experience that didn't last long. The room was probably worse than the speakers look if that's possible(it's a joke people).
It's like going to a music store and playing instruments. They always sound different when you bring them home and you can hear their tone so much better. Same goes for guitar amps--low level noise you can't hear in the store gets all too obvious at home. Too bad there isn't an objective way to know what instrument you are going to like when you are in a music store. Maybe that's the next thing for researchers to do. Where's Drs. Toole and Olive when you need them? 😕 Ha ha. (don't anyone get huffy and puffy over that comment. It is a joke.) I go to instrument shops and play for hours and hours, listening carefully through the noise, only to get home be irked. Can't tell you how many times this has happened. Unfortunate that I have only come home to a pleasant surprise once with cheap($200), old, used guitar that actually sounded even better when I got home. It is around 40 years old and the finish is a bit worn, but had a relatively new fret job and a real bone nut, compensated saddle and the best intonation I've played. Someone loved this thing and I can't blame them.
Sorry for the rambling.
Dan
It's like going to a music store and playing instruments. They always sound different when you bring them home and you can hear their tone so much better. Same goes for guitar amps--low level noise you can't hear in the store gets all too obvious at home. Too bad there isn't an objective way to know what instrument you are going to like when you are in a music store. Maybe that's the next thing for researchers to do. Where's Drs. Toole and Olive when you need them? 😕 Ha ha. (don't anyone get huffy and puffy over that comment. It is a joke.) I go to instrument shops and play for hours and hours, listening carefully through the noise, only to get home be irked. Can't tell you how many times this has happened. Unfortunate that I have only come home to a pleasant surprise once with cheap($200), old, used guitar that actually sounded even better when I got home. It is around 40 years old and the finish is a bit worn, but had a relatively new fret job and a real bone nut, compensated saddle and the best intonation I've played. Someone loved this thing and I can't blame them.
Sorry for the rambling.
Dan
Last edited:
I pointed out in my last post that "noise," of course, has meanings outside music reproduction. So what?
Within music reproduction, noise means human noise annoyance and there are alternative and competing efforts to characterize it so that the measure bears a better and better resemblance to the degree of annoyance HUMANS feel in the presence of noise.
It has no physical definition (I think that is what you incorrectly refer to as "objective") except arbitrary "standards" like the "equivalent" noise of phono pre-pre-amps. If you are big into arbitrary, just remember your measure is meant to simulate HUMAN noise annoyance and sooner or later somebody will have a better measure (for example, competing "standards" for phono pre-pre-amp noise).
As I also said a few times before, there are good reasons to seek reliable physical measures. But you shouldn't be delusional about the chicken and the egg by somehow thinking the physical ("objective") measure has any independent validity outside its correspondence to the HUMAN measures.
- Status
- Not open for further replies.