Flat Frequency Response

This is a very controversial subject. Many would resent this idea. However this should not be and some examples will be provided to further enlighten the matter.

1st example: -

Sound as known consists of a fundamental frequency and a series of harmonics. Harmonics are multiples of the fundamental.

e.g. if fundamental frequency is 100Hz, then 1st harmonic is 200Hz, while 2nd harmonic is 300Hz, 4th harmonic is 400Hz and so on. There are effectively unlimited harmonics until it's decay energy dies completely.

Normally the fundamental frequency has the strongest level, where the 1st harmonic (a natural by-product of the fundamental) has a weaker strength (cause it's an effect), while the 2nd harmonic has even weaker strength than the 1st harmonic and so on. Normally this is the case, but not always. Anyway on a graph they looks like: -

This is the real sound represented on a "Frequency Domain" graph. The word "real" is stressed because any attempt to create a sound / fundamental frequency, will automatically yield / produce harmonics in acoustics domain (not electrical domain, not normally anyway).

Although most electronics (electrical domain) also has this parasitic form / problem (not a problem really), [anyway] but it is only the technology at this writing in time is not up to the challenge yet, which is cost effectively enough for all consumer, professional researchers / producers of music, albums, movies even. 

Back to this topic.

Assuming the sound is properly captured by the microphone (flat frequency response) and properly stored (flat frequency response). When played back on a flat frequency response system it will sound exactly the same.

Theoretically true. However in reality not true, because the microphone only captures sound in 2 dimensional terms (strength and time, but not spherical motion / directions of the sound wave) [some times refer to as phase] {but not as the phase, that we know} <it's a different world out there>

Coupled with the effect of the question, was the microphone near to source or far. 

In the near case, surrounding environment's conditions (room acoustics) may be ignored. 

If far, then the microphone would have captured the room's response as well. Please be aware at this point, the original sound is now joint with it's siblings (reflections or multiples of itself at a latter time, coming back from all directions of the room and most even reflected countless times through hundreds of directions) 

Side notes : - as mentioned sound is an extremely complicated subject that many do not consider.

Now, assuming the recording was done using the near technique and room response is not of concern, then it is save to assume that the sound reproduction through the play back system will be perfect.

Again, this is only partially true. The reason that it will not be perfect is because when the sound leaves the loudspeaker and enters into atmosphere (acoustic energy) it will travel outwards in all 360°, what scientifically term as 4πSpace (read as 4 pie space). This outward style is completely DIFFERENT from the original real sound's outward energy pattern.

This time, the outward pattern is of the loudspeaker's natural energy pattern, not the real sound's energy pattern.

Due to this, as the sound reaches the ears and the reflections from the walls that reaches the ears, will not have the effect as the real sound having it's reflections off the wall.

To a certain extend, if everything is controlled properly, no doubly the loudspeaker's direct frontal sound output will be 100% as the real thing, but in reality, the human brains listen to sounds in time domain, which opens a window of approximately 35 ms [milliseconds] (depends on person to person) and sends it for processing.

The human brain does not receive every 35 ms segment, instead it sends it in a analogous method, which is unlimited overlapping of the sound wave in time domain but each moving forward away as time goes away.

35 ms may sound extremely short, but consider it is continuous analogously and infinitely, then it is a very very long time. As long as we live and as long as there is new sounds coming.

Please also consider the fact that sounds lingers in the room no less than half a second which is at least 500 ms. During this time, there would be numerous sounds constantly changing as we speak.

Plus the reflections in the room all going to both ears from all weird directions.

In a typical room of 3 meters by 3.5 meters by 2.5 meters ceiling. The reflections of 500ms distance would have traveled: -

Speed of sound = 343.3 m/s @ sea level @ 20º Celsius temperature [ignoring humidity for now] (m/s = meters per second)

If sound travel 500 ms, that would yield a distance of (500 ms / 1000 ms) x 343.4 = 171.7 meters of distance

Now imagine the small room, one single reflection is 171.7 meters long, how many points / walls have it reflected?

Now consider the sound outwards from the loudspeaker (or any source) is 360°, 4πSpace remember, how MUCH reflections are there now? what are the probability of one wave affecting the other at this time on point from that direction of reflections and the next time in point from another directions.

Complicated? yes, it's the reality.

Don't forget the loudspeaker's direct distance to the ear is only some 3 meters, therefore there is a constant 171.7 meters of reflections constantly interfering with the new on coming sound.

During this time, sounds (more specifically frequency / wave) that has strength 10dB stronger than the reflections will not be affected, while as the direct source strengths is lower, it is further subject to interference by countless unlimited reflections.

OK - the above was based on the recording made where the microphone is extremely near to the source.

Now consider, if the microphone was far and that it captured the recording's environment, be it a studio, concert hall or whatever.

If it is played back into the room, the sound would have to / been affected twice before reaching the ear. Isn't this a shame?

Most would have realized by now that allot of recordings, where in a single song have both near-field and far-field microphone techniques. So how will it be audiophile / perfect.

Until the recording technology changes, there is no perfection to pursue at this time. So, don't waste $$$ on branded goods. It's not worth it. Some have spend millions on audio system, they could enrol for an acoustics degree and learn the truth, or sponsor it to others to study, or sponsor it to university to do research. Wouldn't that make a better world.

As described above, this is why flat frequency response system does not appear to sound perfect.

Coupled with the fact that 99.99999999999999% of the people in this world do not realized these issues and misjudged the system.

Therefore, the readers are urge not to believe salesmen, they are not qualified acousticians. The readers are urge not to believe magazines as they are also non-scientifical in their review / judging methods.

Some would jump to the conclusion to fight it / overcome the natural disaster described above / overcome it using some simple estimates of the weaknesses in the system and counter attack it in it's weak acoustics domain. However this is not possible since there are too many variables in too many time points, windows and directions to counter attack.

Please, don't even try it. If you study enough acoustics, you'll know why.

That raps up the 1st example.

Hehehe, sorry it has to be so long, that's the reality. Now coming back to read it, I'm not so sure if I know what I was talking about. hahahahahahhaa

Nah, I know every single word I'm talking about, just didn't realize it is so densely packed. This page contains my 13 years journey in sound studies. It is a short summary. OK, long summary, but hey, it beats the stack of books all the acoustician / audionist ever produced

Now, 2nd example.

hahahaha, and you thought it's finished hahahahahahha

Since this world cannot achieve 100% real reproduction, due to natural causes (laws of acoustics), therefore no one have a chance to listen to the true performance of a flat frequency response reproduction system. The conclusion calls for everyone not to believe any other person who says flat frequency response is not good. These people do not know what they are talking about.

next to Audio Crossovers

back to The Concept

Oh, yeah, if you don't think flat is good, then what is good?

If flat is not to be used as reference, then why use anything at all? mind as well use no frequency response, where there is no sound at all?