Equal-Loudness-Curves / Fletcher-Munson / ISO 226 / etc. - WHY do we care? - Page 3 - diyAudio
Go Back   Home > Forums > Loudspeakers > Multi-Way

Multi-Way Conventional loudspeakers with crossovers

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 25th February 2009, 12:26 AM   #21
Banned
 
Join Date: May 2008
Location: Switzerland
smellygas, we don't need to equalize to conform to one of the curves because this is a filter that is part of our hearing. We can't switch it off.
The equal-loudness contours are very important if you want to hear a recording in exact the same tonal balance no matter what volume level you're listening to. The differences below 1000Hz are NOT minor.
The problem is that you would need to know what reference level a recording was mixed at. But there's no standardized reference level in music mixing or mastering.
Another problem is that the equal-loudness contours are the mean of a lot of ears—your personal equal-loudness contour could be different. See http://www.nedo.go.jp/itd/grant-e/re...pdf/is-01e.pdf

Best, Markus
  Reply With Quote
Old 25th February 2009, 12:32 AM   #22
Pano is offline Pano  United States
diyAudio Moderator
 
Pano's Avatar
 
Join Date: Oct 2004
Location: Milliways
Blog Entries: 4
Recordings are mixed and mastered at a cetain SPL. Alas, we don't know what that is - tho we can guess.

So if you are listening at a much lower level, a little F/M boost can help restore the tonal balance to where the mastering engineer liked it. As has been mentioned above.

One thing not often mentioned is that the equal loudness curves were measured with pure tones. With multiple tones, such as noise or music, the curves are not the same. Or so they say....
  Reply With Quote
Old 25th February 2009, 12:36 AM   #23
diyAudio Member
 
Join Date: May 2008
Quote:
Originally posted by MaVo


Following is my paraphrase / interpretation of the quote. Sorry if i dont get it 100%, since i am no native english speaker, as most of you are.

I think, that ISO 226 (the Equal Loudness Countour) has to be used in the right way, while you seem to be drawing the wrong conclusions from it.

Does ISO 226 have something to do with loudspeaker design?

Yes, but you have to use the information it holds in the right way. This way is, for example, to identify the lowest frequency, one can hear at a certain level. Then you design a loudspeaker with this information, so that it doesnt struggle to produce frequencies, which you wont hear anyway, since it cannot make them loud enough, by raising the low frequency corner of said speaker.
Okay, thanks. That's definitely more clear, but I still don't understand the last paragraph. No big deal.

SG
  Reply With Quote
Old 25th February 2009, 12:40 AM   #24
Pano is offline Pano  United States
diyAudio Moderator
 
Pano's Avatar
 
Join Date: Oct 2004
Location: Milliways
Blog Entries: 4
Hey, you're lucky you got that much out of Mr. Sreten. He usually limits himself to 1 or 2 lines.
  Reply With Quote
Old 25th February 2009, 04:53 AM   #25
infinia is offline infinia  United States
diyAudio Member
 
infinia's Avatar
 
Join Date: May 2005
Location: Californication
Default on axis anachoic flat response

2 articles is simply is not proof to me, even the streophile author points to others dissent of this within his article. In my mind flatness is probably number 6 on the list of things not even mentioned.

1) We don't listen in an anachoic chamber. So room interaction varies with each design type differently, better or worse. ie offaxis dispersion and integration thereof between drivers is key.
2) Is one type more pleasing if it has a wider response and/or rolloff character but not as flat.
3) Power compression can limit the useful dynamic range.
4) Inband distortion from many sources of nonlinearities.
5) I would agree that smoothness is more important than tighter tolerance to flatness of on-axis anachoic response .
__________________
like four million tons of hydrogen exploding on the sun
like the whisper of the termites building castles in the dust
  Reply With Quote
Old 25th February 2009, 07:09 AM   #26
diyAudio Member
 
Join Date: Jun 2007
First, what do the Fletcher-Munson curves tell us about human hearing, only that at low levels we don't hear high and low notes as well.

At higher levels, we perceive more bass and more treble. Our hearing response come closer to flat as loudness increases.

Which in turn means, that you can't reproduce concert sound at low levels unless you add bass and treble boost. I have a switch on the front of my amp labeled 'Loudness' that attempts to do just that; give a bass and treble boost at low SPL levels.

Beyond what I have already described, should you try and compensate for it in your speaker design or in system equalization? Well, no, because it reflects our normal natural hearing. You have heard everything from the day you were born colored by these curves. If you alter the curve or try to counteract it, it will sound unnatural because it is unnatural.

Further, if you did try to equalize your response to counter the F-M curve, that curve is volume (SPL) dependent. At high sound levels our perceived frequency response flattens out, at low level, the midrange is emphasized. I suspect this is related to some genetic survival adaptation.

To get this so-called equalization right, any bass and treble boost at low levels would have to gradually fade out or flatten out as the sound level rose. Can you do that in standard speaker and/or passive crossover design? I don't think so.

Now I'm sure we've all experienced the sensation that such-and-such a speaker really doesn't get moving aesthetically until you have the amp turned up to a certain level. What we are probably experience is the effects illustrated by the Fletcher-Munson curves. That we really don't hear the bass and treble clearly until we have our amps up to a certain level.

Though likely that threshold is also related to being able to feel the music as well as hear it.

So, everyday in every way, the Fletcher-Munson Equal Volume Curves do influence and effect your life. But there is really nothing you can do about it, because that is natural normal hearing.

Further, any method you attempt, like a 'loudness switch' has to be variable, high boost at low volume, and gradually fading to low boost at high volume. That is difficult to implement in a speaker or passive crossover design. Though much easier in amplifier electronics.

Though in DIY speaker design, if you are aware of the range of loudness levels your speakers are likely to be listened to, you could build that bass and treble boost in to accommodate those listening volumes. Again, though, unless it is self-adjusting, it is not likely to sound right at all loudness levels. It's trick business.

For the most part, you really can't and shouldn't do anything about the F-M curves. It is worthwhile information to know, but it is like how we perceive light. The spectrum we see does not match a dogs or a snakes, but it is, none the less, perfectly natural for a human, and is how we have perceived sound since we were born. To fool with it, or counter act it, is like putting on rose colored glassed to correct you light perception. You can do it, but it won't look natural because that is not how human see. And you can try to counter act the F-M curves, beyond the standard 'loudness button', but that is not how humans naturally hear.

My advice, file it away and forget about it.

As to speaker optimized in an anechoic chamber, all the really mean is the speaker have been tested in an environment that doesn't have any outside interference. If they tried to test them in a standard room or office, or optimized them for a standard room or office, they would probably sound terrible in your room.

There is nothing magical about an anechoic chamber beyond that you can test speakers there without interference. If they are flat there, then they are truly flat. But the minute you put them in any room, that room throws the flatness completely off. So, you try an tweak and correct the room to get the best fit between 'flat' speaker and a not flat room.

Many DIY speaker designer, don't optimize speakers to be perfectly flat in the neutral environment of an anechoic chamber. They specifically design them with their specific room in mind. They design them to be flat in the room they know they will be listening to them in.

I alway thought that was a bad idea. It ties the speakers and room together too tightly. Sure it will sound fine until the wife decides to redecorate or move the furniture around. Then it is back to the drawing board.

Speaker manufacturers can't possibly know the details of your specific room, so they do their best to design speakers that are independently flat, and from that point on, making them work in your room is up to you.

Just a bunch of rambling.

Steve/bluewizard
  Reply With Quote
Old 25th February 2009, 08:16 AM   #27
Jmmlc is offline Jmmlc  France
R.I.P.
 
Join Date: Oct 2005
Default Re: Equal-Loudness-Curves / Fletcher-Munson / ISO 226 / etc. - WHY do we care?

Hello,

If a record has been mixed in order to have a given equilibrium at 85dB SPL (which is the most used mean level by sound engineers) and we want to listen to it at a lower mean SPL level, we experiment during audition a change in the tonal equilibrium (noticeably the inetnsity of the frequency seems to lower).

If we want the to obtain the same tonal equilibrium between we need to compensate for the lost of intensity and we have to use -not the ISO226 curves- but the differential loudness curves.

Such differential data is often presented as a 3D graph "Human Auditory Differential Threshold for sound intensity " see by example:
http://pages.cpsc.ucalgary.ca/~hill/...mages/dh20.jpg

Best regards from Paris, France

Jean-Michel Le Cléac'h


Quote:
Originally posted by smellygas
This might seem like a silly question, but the equal-loudness contours (originally the Fletcher-Munson curves), and now the ISO 226 standard, all basically show the actual SPL (dB) loudness of a frequency that sounds just as loud as another frequency. It also shows that this varies based on the actual listening volume. But WHO CARES? - is what I want to know.

I mean, it's been very well established that a perfectly flat SPL vs. frequency curve, measured in an anechoic chamber, is what sounds best to listeners. So when we would we care about these equal-loudness curves, or for that matter, when would we want to equalize to conform to one of them?

SG
  Reply With Quote
Old 25th February 2009, 11:01 AM   #28
MaVo is offline MaVo  Germany
diyAudio Member
 
Join Date: Jan 2006
Quote:
Originally posted by smellygas


Okay, thanks. That's definitely more clear, but I still don't understand the last paragraph. No big deal.

SG
If you look at this picture: http://upload.wikimedia.org/wikipedi...47/Lindos1.svg you see a line at the bottom, which is labeled threshold. This is the frequency dependant absolute hearing threshold. Signals below cant be heard. If you look at 20hz for example, the threshold is slightly above 70db. If you design a speaker, which is flat to 20hz but limited to less than 70db max spl at this frequency (which usually happens with small drivers or too little xmax), then this doesnt make sense, as it cant reproduce a 20hz wave loud enough for it to be heard. Instead, you should tune the speaker higher, lets say flat to 40hz, which should also give you a max spl rise.

A speaker should be able to be significantly louder than the hearing threshold. If you add about 30db, you end up with 100db maxspl requirement at 20hz, which puts alot of woofers to shame.
  Reply With Quote
Old 25th February 2009, 11:19 AM   #29
sreten is offline sreten  United Kingdom
diyAudio Member
 
Join Date: Nov 2003
Location: Brighton UK
Quote:
Originally posted by smellygas


I'm sorry, but I read what you wrote twice, and
I don't understand what you're trying to say.

SG
Hi,

The short version is (as Gedlee suggested) speakers should be flat.

But its almost a truism, understanding it means understanding
when it is not true, this then defines when it is actually true.

You cannot understand something by assuming it is always true.
(Unless of course it is, which you seem to be implying is the case.)

Another example :

Monitoring for programme distortion at artificially high SPL levels :
These system often use inverse FM, bass and treble cut, to allow
a more natural apparent tonal balance that matches the monitors
that are actually used for tonal balance, prespectives etc that are
run at lower "typical" SPL levels.

/sreten.
  Reply With Quote
Old 25th February 2009, 04:11 PM   #30
diyAudio Member
 
Join Date: May 2008
Default Re: on axis anachoic flat response

Quote:
Originally posted by infinia
2 articles is simply is not proof to me, even the streophile author points to others dissent of this within his article. In my mind flatness is probably number 6 on the list of things not even mentioned.

1) We don't listen in an anachoic chamber. So room interaction varies with each design type differently, better or worse. ie offaxis dispersion and integration thereof between drivers is key.
2) Is one type more pleasing if it has a wider response and/or rolloff character but not as flat.
3) Power compression can limit the useful dynamic range.
4) Inband distortion from many sources of nonlinearities.
5) I would agree that smoothness is more important than tighter tolerance to flatness of on-axis anachoic response .
Hey, I do appreciate your opinion. However, when someone has conducted a well-designed empirical study that is published in a reputable scientific journal, and the results are very clear, I tend to believe it. When these types of experiments are repeated by other independent groups, and also published, and similar results are obtained, then I tend to believe the shared conclusion is true. This is called science.

If you would like to challenge the methodology or conclusions of the Toole study (2nd link) specifically, I'd be happy to discuss it with you. The paper is very elegant. You take a large selection of loudspeakers, measure their anechoic response, and you take a bunch of listeners, blind them, and ask them to rate the speakers. There is a very clear rank order of speakers that listeners preferred. There is also a very clear trend in the frequency response curves that show that flat and smooth is preferable. The author already accounted for the effect of low frequency cutoff (your #2). Furthermore, they also looked at the response of speakers a) in a room and b) off-axis with various weights. It's a classic paper. You should actually read it.

SG
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Fletcher Munson Vs Robertson Dadson inertial Multi-Way 4 26th March 2009 10:52 AM
Derive Triode Curves from Pentode Curves? dsavitsk Tubes / Valves 3 14th May 2007 06:10 PM
Equal-Loudness Curves for low frequencies? Sjoerd v L Subwoofers 4 9th June 2005 10:22 PM
Calling Dr. Fletcher! Dr. Munson! Apparent 20 dB Discrepancy! kelticwizard Multi-Way 10 18th November 2004 01:59 PM


New To Site? Need Help?

All times are GMT. The time now is 09:23 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2