The soundboard of a grand piano has a very high modal
density even at low frequencies.
Sound is radiated in a rather decorrelated manner
in many directions. Even in rather small and
reverberant places a piano may sound well
and balanced - unlike a tuba e.g. ...
(I once listened to a traditional brass band playing in
a garage ... oh my)
That decorrelated/diffuse soundfield of the piano
is often destroyed, when playing the recorded sound
using conventional speakers in the usual reverberant
listening room, even if the recording was able to preserve
some of that characteristics.
The early reflections are highly correlated to the direct
sound, and artefacts like combing are quite common with
usual speakers.
The characteristic piano sound does not like those artefacts ...
presumably because its nature is being virtually free from
those in the original sound.
It is not necessary to simulate "the wavefront" but it is highly
desirable to mitigate or circumvent those speaker/room artefacts.
Bending wave planars ( sorry for the repetition...🙂 ) can transport
piano sound in a very "realistic" manner.
They do nothing to restore the "original wavefront", they just
mitigate reflection and combing artefacts.
Piano sound is an example which benefits highly, maybe more
than any other instrumental sound.
Cheers
Last edited:
Sorry, I forgot to say "thank you" to Dave.
Anyway, I posted this before, but it really pertains to what LineArray and Dr. Geddes and every control room designer have been saying and pretty much the opposite of what Mr. Linkwitz and Deward say. Remember what Gov. Schwarzenegger said referring to climate change,'if I go to 100 doctors and 98 say the same thing, I'm going to listen to the 98.' Of course they have the goal of undoing the recording process through their listening room reverberation(don't ask me how) and HRTF compensation to get back to the original event.
Stravinsky and I have a different view. Here's what he thought about stereo vs. the real deal--"How can we continue to prefer an inferior reality (a concert hall) to ideal stereophony?" Good thing he didn't hear a surround sound control room! When you look at the RT60/30 of mine and Pano's rooms this start to make more sense--it could also hint at a need for 'HRTF' compensation. There are a lot of problems to overcome trying to undo the recording. Interested parties should read this and the links. audio blog: Tightening The Loudspeaker, Recording and Room Connection Might put you in mind of a different goal.
Dan
Anyway, I posted this before, but it really pertains to what LineArray and Dr. Geddes and every control room designer have been saying and pretty much the opposite of what Mr. Linkwitz and Deward say. Remember what Gov. Schwarzenegger said referring to climate change,'if I go to 100 doctors and 98 say the same thing, I'm going to listen to the 98.' Of course they have the goal of undoing the recording process through their listening room reverberation(don't ask me how) and HRTF compensation to get back to the original event.
Stravinsky and I have a different view. Here's what he thought about stereo vs. the real deal--"How can we continue to prefer an inferior reality (a concert hall) to ideal stereophony?" Good thing he didn't hear a surround sound control room! When you look at the RT60/30 of mine and Pano's rooms this start to make more sense--it could also hint at a need for 'HRTF' compensation. There are a lot of problems to overcome trying to undo the recording. Interested parties should read this and the links. audio blog: Tightening The Loudspeaker, Recording and Room Connection Might put you in mind of a different goal. Dan
Same . . . but some (small) number of "good enough". Good enough to pass for a bad seat in a marginal house, anyway, or listening from back stage. What I've never heard is a convincing choral recording, or even a convincing human voice. I've heard plenty of "rock" recordings where the sound in my listening room satisfactorily replicates the (amplified) sound in the hall, but I have *never* heard a recording of a soprano that sounds like she really sounds standing ten or fifteen feet away on stage or in the rehearsal room. Not even close.
I don't really have words to describe the difference . . . you just know . . .
Right. As in science research, hard to make a proof of "no difference" because any ham-fisted researcher can "detect" no difference simply by being inept.
Likewise with the sound of music. What is a good test?
For starters, complex things like orchestras hard to have a sound in mind for, let alone remember. While radio announcers are just people talking. A recording of my toaster popping toast up or my wife's violin are familiar in the way a tuba is not.
Some truly impressive recordings of big Mahler symphonies have flute solos. Sometimes the flute is recorded very realistically. But then.... is that little breathy sounds I hear? Yes. Isn't that realistic!!!! Great recording!!!!
Of course in a concert hall you never hear those artificially close-mic'ed breathy sounds. But in a recording, it sounds wonderful even, as we might incorrectly say,.... realistic!
C'mon, how many times, if ever, have you been busy doing something down the hallway or even drunk and then looked up and said "... gosh, somebody must be playing a guitar in my music room." I always know when it is my wife practicing her violin in the music room and not a recording of Lang Lang or Ling Ling playing the violin*. Mistake any kind of drum recording for real?
*Just a little joke. Neither Lang Lang nor Ling Ling are any good on the violin. But you know what I mean.
Last edited:
also:
quoted from: Polar response
the practical answer to all this questions and impressions is to be found here for allmost three years now:
http://www.diyaudio.com/forums/multi-way/121385-loudspeakers-room-system.html
best,
graaf
I'd start by checking if the hills are alive. 😉
Darn close, yes. At least twice. RMAF (Denver) and T.H.E Show (Vegas). Was so unsure I had to go down the hall to check it out. "Is it, or isn't it?" Both times were big coax systems.
I've heard that occasional sax or trumpet coming out of a room that made me go investigate. And I know a pair of speakers that can do that on chamber music. A full orchestra? That's very difficult and very rare.
You might have a glance at my article here: Capitol Audiofest 2010 | Hifi Zine Scroll down near the end to the photo of the band and read the second paragraph under that. I heard recorded and live jazz in that room. No, it was NOT the same.
1. Soooo, you were "down the hall", eh.
2. Happens in our lives as often as we "mistake a wife for a hat-rack" - without being too serious with the neurological disfunction reference or random luck, still of all the mega-dollar demos in Washington, only one hit it lucky for you, perhaps by luck and down the hall.
3. My belief is that tweeter power is a main shortcoming of home systems and as much as I dislike treble horns, they do have power (4 Dennesen ESL upper tweeter panels per side also has good power when driven by 340 watts per side). I wonder if a lot of recordings have too much treble and we tweak it down at home to sound flat.
2. Happens in our lives as often as we "mistake a wife for a hat-rack" - without being too serious with the neurological disfunction reference or random luck, still of all the mega-dollar demos in Washington, only one hit it lucky for you, perhaps by luck and down the hall.
3. My belief is that tweeter power is a main shortcoming of home systems and as much as I dislike treble horns, they do have power (4 Dennesen ESL upper tweeter panels per side also has good power when driven by 340 watts per side). I wonder if a lot of recordings have too much treble and we tweak it down at home to sound flat.
Last edited:
I was kinda wondering the same after listening to tweeters all day then sitting a few feet from live cymbals in one of the same rooms.
One of the "Hey! Is that really someone playing sax?" moments did come at the D.C. show. And it was one of the smallest systems there.
In the room it probably would not have fooled me - but it was darn good. Not many do that. The designer seemed to have a knack with crossovers.
One of the "Hey! Is that really someone playing sax?" moments did come at the D.C. show. And it was one of the smallest systems there.
In the room it probably would not have fooled me - but it was darn good. Not many do that. The designer seemed to have a knack with crossovers.
See my earlier comment about how, in a real performance, the oboe *isn't* "front and center". And this is a generalizable criticism of even the "purist" two-mics-over-conductor recordings, which capture the strings in a way you'd *never* hear otherwise (and is typically "too bright").
Of course a version of this has also been done in performance . . . "reduced complement", for example, where each section is played by one instrument or sung by one voice. Same music, but a very different "sound" from full ensemble. And works are commonly "arranged", or re-written with different instruments providing the bass, or carrying the melodic line(s). For "studio produced" music the mix is part of the band (and if you want a "breathy" flute you can have it). But for "high fidelity" the mix is part of the reproduction chain (and if it's a quartet "up close" you should hear the breath, if it's an orchestra on stage you should not). It's a distinction that is lost on those who have no experience of live performance or its reproduction.
"flat is incorrect" --> flat compared to what?
Just for kicks i carefuly mesured a swept sine with my speakers to see just what settings i prefer.
Well, the plot is rather bumpy, specialy the low tunes under 100 hz.
for fun i managed to set up an EQ software to make sure the responce is flat.
The electret mic i used to mesure has a calibration datasheet, so supposedly i did get close to what one would call flat.
I have listened to the setup for over an hour now.
I find it less realistic then my custom setup with the dips and peaks.
Actualy, it may be having dips and peaks according to the mic.
but not according to my senses.
In the end i think its my own senses i find more important.
I done some tests with signal generator, for my ears the freqvency range does indeed seem flat. My ears tell me its flat.
My conclusion is, flat IS correct, but is a relative term.
Flat for my ears may be horrid for yours.
Question is, when it comes to DIY audio, what is the goal we try to reach?
Say...
Lets just take a recroding.
So should my speakers replicate the recording, with the goal to maintain a flat responce according to an electret microphone?
in that case the characteristics of my hearing would alter this.
example, i drive a tractor, so my hearing is not quite as good in amny freqvency ranges. I would hear the recording from my speakers with my own hearing that is not perfect.
But what gives if the responce is flat according to my hearing?
This is where i think the magic happens.
I can again hear CRISP and CLEAR highs, like if my ears and hearing would function just as when i was young.
Imagine, if i would go to a theater, even given the best accustics, my personal sonic experience would still be poisoned by my hearing.
On the otherhand, with my settings -that are absolute not flat according to the mesurement microphone- i can lessen the effects of my hearing, and experience the recording as IF i had better hearing.
So, if someone would ask me, i would say "flat" is correct, depending on what we base "flat" on.
Recordings suufer from soo many problems. And there are just too many sonic events none will ever be able to replicate 1:1.
But, listening to a recording under some circuimstances can be better than the live act.
Wierd, huh ?
//notice, the above is my personal opinion, while it does contain questionable parts, i would think arguments would lead nowhere, as from person to person the sonic experience associated to individuals own hearing characteristics will warry.//
There are hearing characteristics that are quite simular for everyone.
For exmaple, the freqvency range. Or the loss of sensitivity to high freqvency sound as we age. I think this is the reason why a "flat" responce may not be correct.
Just for kicks i carefuly mesured a swept sine with my speakers to see just what settings i prefer.
Well, the plot is rather bumpy, specialy the low tunes under 100 hz.
for fun i managed to set up an EQ software to make sure the responce is flat.
The electret mic i used to mesure has a calibration datasheet, so supposedly i did get close to what one would call flat.
I have listened to the setup for over an hour now.
I find it less realistic then my custom setup with the dips and peaks.
Actualy, it may be having dips and peaks according to the mic.
but not according to my senses.
In the end i think its my own senses i find more important.
I done some tests with signal generator, for my ears the freqvency range does indeed seem flat. My ears tell me its flat.
My conclusion is, flat IS correct, but is a relative term.
Flat for my ears may be horrid for yours.
Question is, when it comes to DIY audio, what is the goal we try to reach?
Say...
Lets just take a recroding.
So should my speakers replicate the recording, with the goal to maintain a flat responce according to an electret microphone?
in that case the characteristics of my hearing would alter this.
example, i drive a tractor, so my hearing is not quite as good in amny freqvency ranges. I would hear the recording from my speakers with my own hearing that is not perfect.
But what gives if the responce is flat according to my hearing?
This is where i think the magic happens.
I can again hear CRISP and CLEAR highs, like if my ears and hearing would function just as when i was young.
Imagine, if i would go to a theater, even given the best accustics, my personal sonic experience would still be poisoned by my hearing.
On the otherhand, with my settings -that are absolute not flat according to the mesurement microphone- i can lessen the effects of my hearing, and experience the recording as IF i had better hearing.
So, if someone would ask me, i would say "flat" is correct, depending on what we base "flat" on.
Recordings suufer from soo many problems. And there are just too many sonic events none will ever be able to replicate 1:1.
But, listening to a recording under some circuimstances can be better than the live act.
Wierd, huh ?
//notice, the above is my personal opinion, while it does contain questionable parts, i would think arguments would lead nowhere, as from person to person the sonic experience associated to individuals own hearing characteristics will warry.//
There are hearing characteristics that are quite simular for everyone.
For exmaple, the freqvency range. Or the loss of sensitivity to high freqvency sound as we age. I think this is the reason why a "flat" responce may not be correct.
So... we've got to ask, did you EQ to flat response in room including a substantial amount of room resonances? In my experience that does sound pretty bad. (i.e. lifeless, not real, no impact)
The premise here is that you should start with "Flat" frequency response in a traditional sense (i.e. +- one or two dB using outdoor measurements or an anechoic chamber), then you would "shape" the response to have a downward tilt of 2 to 10db from bottom (20Hz) to top (20kHz).
There are others (perhaps including yourself) that don't believe this offers any significant benefit.
Some, including myself, believe there is some benefit to this "shaping" but most (including myself) would still admit it would be preferable to "Fix" anomalies in your room rather than trying to EQ the response of the speaker to provide a flat in room response.
The specifics of exactly how you should "shape" the response are the subject of this great debate but there are some common elements as illustrated by Dr. Earl, Sigfried Linkwitz, John K and many others. But I don't think anyone here would advocate that you would actually try to EQ the response to be perfectly "Flat" in your room, although the acoustic character of your room may have an effect on how much "shaping" may be desirable.
A few thoughts
Flat is correct for engineering and recording because it makes a flat line on a graph and means the system produces all frequencies with equal intensity, in proportion when reproducing.
Flat was the goal until CD’s came along.
In order to convince the masses that they needed to switch over and buy a new music collection, it was necessary to demonstrate CD’s sound better, more clear and this was why they always were mastered so much brighter than records.
Anyone listening A vs B could hear a difference.
People have forgotten about that change in taste and now to compensate, we roll off the high end.
Also, while we have tweeter that go higher than before, the focus tends to be on one meter measurements while non-constant directivity can cause the frequency response to change with distance even outdoors.
Best,
Tom Danley
Flat is correct for engineering and recording because it makes a flat line on a graph and means the system produces all frequencies with equal intensity, in proportion when reproducing.
Flat was the goal until CD’s came along.
In order to convince the masses that they needed to switch over and buy a new music collection, it was necessary to demonstrate CD’s sound better, more clear and this was why they always were mastered so much brighter than records.
Anyone listening A vs B could hear a difference.
People have forgotten about that change in taste and now to compensate, we roll off the high end.
Also, while we have tweeter that go higher than before, the focus tends to be on one meter measurements while non-constant directivity can cause the frequency response to change with distance even outdoors.
Best,
Tom Danley
This debate is kind of like telling everyone to they should Jazz instead of Rock n' Roll or listen to Modern Country instead of Classical. Its purely a subjective back and forth. Good guidelines from the people you mentioned but how someone likes their setup EQed is simply a choice.
They can go out and learn about the guidelines behind in room response curves and so far I think Sean Olive, Toole, etc have done the best studies and created the best books, online resources. They have shown through many controlled listening tests that a tilted in room response is preferred. This sort of curve is a good start because it follows what the Fletcher Munson curve tells us.
In the end though people need to have the ability to control the response in room the way the see fit. I have control from 10Hz to 20KHz and I will change the curve depending on the content.
you are verry right there, actualy the only way to get close toperfectly flat would be with a full digital speaker array setup.
there are allso reflections that would be hard to get rid of.
there would be just way too many things that would prevent flat responce, but from cca. 80 cm i did mesure more or less flat responce for my test.
anyways as you sayd, i found it dead, and lacking any impact. sortha like if it would come from behind a curtain.
as it seems we do share our view at one point, EQ is there to adjust things to personal preference.
Even ignoring equal-loudness contours (which we shouldn't do), there is no agreed definition even for loudness from a given set of speakers in a given location and for a mic or mics in specific location(s). Depends on the stimulus - sustained sine-waves or whatever vast choice of alternatives. And a whole bunch of somewhat arbitrary definitions.*
So let's not get all thrilled about something that is not real.
Since the goal is music reproduction, a panel of judges listening to some class of music is the criterion for flat. Or if there were a generally agreed equivalent synthetic stimulus to music.
*Yes, you can design and try to sell speakers which by virtue of cannily chosen design choices will sound pretty good in a lot of rooms.
So let's not get all thrilled about something that is not real.
Since the goal is music reproduction, a panel of judges listening to some class of music is the criterion for flat. Or if there were a generally agreed equivalent synthetic stimulus to music.
*Yes, you can design and try to sell speakers which by virtue of cannily chosen design choices will sound pretty good in a lot of rooms.
Last edited:
Is the problem mostly the room?
I just got back from the local multiplex cinema and found no problem with the subjective frequency response or tonal balance. It's just the JBL stuff everyone else has. But of course the room is big and was built for the purpose. It's not perfect (I've heard better) but it's not bad.
How are the cinema installers doing it?
I just got back from the local multiplex cinema and found no problem with the subjective frequency response or tonal balance. It's just the JBL stuff everyone else has. But of course the room is big and was built for the purpose. It's not perfect (I've heard better) but it's not bad.
How are the cinema installers doing it?
Hi Tom - its been a long time! Good points about CD mastering.
Any discussion of "flat" has to take into account the difference between the direct field and the reverberant field. There is no justification for having a non-(near) flat direct field, but the reveberent field is another issue. One must also consider the use of single point microphone measurements as they are almost useless.
A flat direct field is certainly the goal (albeit some HF falloff of a few dB seems to be justified, but its not clear why). A faster falloff of the reverberenat field seems to also be desired, but in any case a smooth falloff is essential - no peaks or dips. Adjusting to "taste" seems ludicrous to me.
Any discussion of "flat" has to take into account the difference between the direct field and the reverberant field. There is no justification for having a non-(near) flat direct field, but the reveberent field is another issue. One must also consider the use of single point microphone measurements as they are almost useless.
A flat direct field is certainly the goal (albeit some HF falloff of a few dB seems to be justified, but its not clear why). A faster falloff of the reverberenat field seems to also be desired, but in any case a smooth falloff is essential - no peaks or dips. Adjusting to "taste" seems ludicrous to me.
Last edited:
Yes, flat direct field would be important if people could tune-out the rest of the sound, which obviously they can't do. Which I hope puts that idea to rest. But canny designers try to get Basic Requirement #1 to be pretty flat because it is a reasonable starting foundation although it seems to fail at the extremes.
You can achieve a reasonable loudness compass. And I believe that some day you can achieve a tweakless installation by testing with mics, if you had the right methods including a sound stimulus resembling the kind of music you want to play. Every speaker will sound right in some room but need tweaking in all others.
We are blessed with very forgiving perception which, as Helmholtz argued, activiy constructs a coherent and determinedly stable universe for us. For example, it papers over a whole world of artfully created but nonsensical sound in movies. We can say "that's a very enjoyable gamelan band" but it is another thing to say "that sounds like a gamelan band in my music room".
Last edited:
- Status
- Not open for further replies.