Question Should we ditch flat freq. response for the better?

AJM1981

Well-known member
a flat frequency response makes sense
- for laboratory use
- for comparison in graph data, if graph data is your main guide.

It does however not make so much sense to our ears and brains. Unless we would hear with microphones as they measure speakers.

When considering that cars, for totally other reasons are measured by shape. The shape approach of our head and how our ears and brain perceive the sound is an important variation that is often totally ignored when developing consumer products.

"Part" of a better approach of measurement taking and analysis for hifi equipment would be the following.

neumann-ku-100-binaural-dummy-head_29980_1.jpg


There are speakers considered flat by measurements, but a flat response is often not perceived as being flat.

When recording an accoustic guitar in an average studio I often notice that compared to the real recording, playback over neutral measuring monitors slightly dampens the treble. The sparkle of the real recording is tuned down.

Since it is in the recording and microphones can register its frequencies, it is not lost and speakers are able to reproduce it, with or without a little bump in treble, depending on their signature.

Another factor is aging of ears. There are tests to measure till which frequency we can still hear. Usually, depending on the conditions degrading slower or faster by age.

conclusion: A flat perception is much better for hifi compared to flat measurements for reasons mentioned above. A flat perception probably measures different on paper to a flat measurement.

It could be that human hearing is taken into account in the development proces more than in the past, but among audiophiles the flat measuring response seems still holified.
 
Last edited:

abacus

Well-known member
A good recording engineer requires flat response speakers to get an accurate reproduction of what he is recording (An alteration of the sound is the last thing needed), however most consumers (Particularly audiophiles) go for a sound that sounds nice rather than accurate (Warts and all), which is why each speaker manufacture has their own way of doing things. (One of the reasons audiophiles prefer one set of speakers over another)
If you look at most speaker/room correction curves they have a slight tilt down from bass to treble as this is what most consumers prefer. (Everyone has a different perception of sound so there will never be one size fits all, but the recording has to be as accurate as possible)
Altering the response to match failing ears is a no, as you cannot alter the response of the real world so everything would sound unnatural. (Get custom hearing aids instead)
The picture you show of the head is used if you wish to hear 360deg sound on headphones (It’s called binaural) as the recording record as the ear hears, however it sounds terrible over speakers as they are not attached to the ear.

Bill
 
When considering that cars, for totally other reasons are measured by shape. The shape approach of our head and how our ears and brain perceive the sound is an important variation that is often totally ignored when developing consumer products.
I can't see that this matter at all - your head and ears are the same shape for live and recorded sound, so it isn't a variable.
 
Funnily enough, although the idea of flat speaker response (anechoically) is often discussed, very few commercial speakers achieve it. Uplifted highs, or a smile shaped response dominate, as these add ‘extra air’ or showroom loudness, both catching the ear.

However, once at home, the room dominates, and a lot of HF is absorbed, giving a pleasant experience. Have you seen how often contributors on this forum seek greater warmth, smoothness, and similar? Only rarely is a brighter sound in demand.
 

Gray

Well-known member
Have you seen how often contributors on this forum seek greater warmth, smoothness, and similar? Only rarely is a brighter sound in demand.
Yes and when you look at some of the bare, reflective room pictures, it's no real surprise.
Warm speakers required for acoustically cold rooms, better, but no real answer.
I much prefer a neutral, detailed speaker in an acoustically deadened room.
There's good reason why speakers are tested in completely non-reflective surroundings - unfortunately unrepresentitive of many of the rooms they end up in.
 
  • Like
Reactions: AJM1981

AJM1981

Well-known member
Funnily enough, although the idea of flat speaker response (anechoically) is often discussed, very few commercial speakers achieve it. Uplifted highs, or a smile shaped response dominate, as these add ‘extra air’ or showroom loudness, both catching the ear.

However, once at home, the room dominates, and a lot of HF is absorbed, giving a pleasant experience. Have you seen how often contributors on this forum seek greater warmth, smoothness, and similar? Only rarely is a brighter sound in demand.

I wonder if the last thing mentioned is because neutral or slightly warm speakers can be brightened by adjusting the treble. Whereas a bright speaker is often not really adjustable to sound neutral. Unless a full band equalizer is involved to do fine tuning adding the BBC dip. That is rarely an option.

I owned some Madison speakers for a second set which were cheap and slightly bright. But truly mind blowing for what they are. Could easily compete with the best.I gave them away because I stumbled upon a bargain with the Wharfedale Diamond 12.0 and I didn't want to collect speakers. But I could as well have kept them.

Bright, when done right.. sounds amazing on accoustic guitar strings. :)
 
Last edited:
  • Like
Reactions: nopiano

AJM1981

Well-known member
Most records isn't flat recorded, so no matter what, 99.99 % of all music won't sound flat

It's just that som system (spakers) it will sound flatter but not 100% flat

This would be a valid point if a random recording would be the benchmark.

The key is that a flat measuring set of monitor speakers prevents that the producer, engineer etc equalizes and corrects recordings according to how the speaker measures.

Example: If you are a producer and take a speaker of an ancient midi set that has troubles in the lows and treble and you start correcting your recording according to the shortcomings of the speaker, your recording will sound completely overstretched on a speaker that measures well.

Comparible to the fact that color correction cant be done on a non calibrated screen.

The whole flat response standard is also not to make the music sound like the musicians intended, but to achieve a "similar" measurement to a studio monitor. So you hear "more or less" the same thing as to what a producer hears on its monitors.
 
Last edited:

manicm

Well-known member
This would be a valid point if a random recording would be the benchmark.

The key is that a flat measuring set of monitor speakers prevents that the producer, engineer etc equalizes and corrects recordings according to how the speaker measures.

Example: If you are a producer and take a speaker of an ancient midi set that has troubles in the lows and treble and you start correcting your recording according to the shortcomings of the speaker, your recording will sound completely overstretched on a speaker that measures well.

Comparible to the fact that color correction cant be done on a non calibrated screen.

The whole flat response standard is also not to make the music sound like the musicians intended, but to achieve a "similar" measurement to a studio monitor. So you hear "more or less" the same thing as to what a producer hears on its monitors.

I'm not so sure about this. Recording engineers will use proper monitors, for accuracy and a baseline, and then adjust and compress to suit a wide audience and range of equipment.

I haven't heard the latest mix of All Things Must Pass, but Dani Harrison definitely took some liberties and not everyone is happy with the new mix of My Sweet Lord.
 
  • Like
Reactions: Tinman1952

Tinman1952

Well-known member
I'm not so sure about this. Recording engineers will use proper monitors, for accuracy and a baseline, and then adjust and compress to suit a wide audience and range of equipment.

True. I did a recording session in the 80s and the sound engineer played back the take through a few different speakers...including a portable mono transistor radio! 🙂
 

AJM1981

Well-known member
True. I did a recording session in the 80s and the sound engineer played back the take through a few different speakers...including a portable mono transistor radio! 🙂

The idea is
When the monitors in the studio are as flat as possible, the reference is neutral. Finding a more or less flat response consumer speaker brings one "kind of" close to what a producer was listening to.

Average onsumer audio in the past was heavily colored or handicapped, depending on what system you used. This has imho heavily improved throughout the ages.

If you create a blueprint on a monitor, it should kind of do well on multiple systems.

Those test runs on small transistors are useful. Sometimes a little extra adjustment like overdubbing something can make a track sound fuller on any system up to scale. Music for allmost all "Audiophile" tracks have been through this proces. :)

Not all choices work out. Human taste can get often in the way on either side. I love ride cymbals and hihats, a friend hates them. So he would probably not like that much going on in the treble. Neither is wrong or right.
 
Last edited:
  • Like
Reactions: Tinman1952

manicm

Well-known member
'
If you create a blueprint on a monitor, it should kind of do well on multiple systems
'.

But sometimes that can be too brutal for normal listeners. Monitors are there for accuracy, but not necessarily palatable.
 

Surly Sid

Well-known member
Feb 6, 2020
274
113
2,070
Visit site
A flat frequency response tells you a speaker is accurate, not necessarily how it sounds.

I know my Harbeths are designed to have the flattest frequency response possible, and they are among the better speakers I have heard.
 

AJM1981

Well-known member
'
If you create a blueprint on a monitor, it should kind of do well on multiple systems
'.

But sometimes that can be too brutal for normal listeners. Monitors are there for accuracy, but not necessarily palatable.

That is the whole point made

Reference vs consumer audio is a whole different ball game. Though, in general.. on "speaker systems we like at this forum" the flat response is there with often minor adjustments. So both worlds connect.

On monitors you need to hear and measure everything that is going on in all frequencies. In consumer audio you have the choice of bumping or tuning down some frequencies in order to create illusions. It is not for no reason that some consumer speakers are quilified as 'for jazz, vocals, rock or electronic music". e. g. They slightly lean towards a kind of consumer that favors a kind of music and spice things up.

Little additional example: An older midi set can measure horribly and and not be enjoyable. A Klipsch La Scala system can measure bad and be really enjoyable.

However. The LaScalas are huge and their literal depth of their huge cabinet can trade off a non flat graph for a sense of musical depth, justifying trading their non flat measurements off for an illusion. But the La Scalas as any other speaker functions best on recordings in which reference monitors have been used.
 
Last edited:

TRENDING THREADS

Latest posts