How important is a flat response?

Flat frequency response is only a portion of what we strive to achieve in a high linearity system. It is quite possible to achieve something that one finds to be "quite enjoyable" without being anywhere near "flat". This is where personal preference comes into play. Some people like a specific sound and tailor their system to achieve that whereas others want the system to try its' best at reproducing what is on the recording without introducing a lot of its' own colourations. This is the difference between a "music lover" and an "audiophile". Due to the amount of highly coloured speakers being marketed and devoured by consumers, i would say that most people, even many that consider themselves to be "audiophiles", fall more into the "music lover" camp. This is not meant to be a "put-down" to "music lovers", only to show that there is a difference. Sean
>

Sean, If the reason we have stereo systems is to be able to listen to music and if we agree that we would like it to be as realistic as possible why would we want "flat frequence response" at the listening position? Most recordings are close multi miked - the end result of having FFR would be the equivilent of having your ear a few feet or in some case inches from the instrument(s) or voice, something that never happens live and sounds most unrealistic (have you ever heard a singers sibilance in an unmiked live performance?). Not at all realistic! The only exception would on those recording where the mikes are placed at an appropriate distance or very simple techniques are used. Any system set up will be static so some recording will sound more realistic on some systems and not on others. I would tune a system to sound as realistic as possible based primarily on the types of recordings I listen to most. FWIW while I accept your use of the terms "audiophile" and "music lover" I think it is perhaps more descriptive to say "audiophile" and "technophile". :-)

You can't recreate a wide variety of instruments with high levels of accuracy when a component ( speakers for example ) introduce a very specific tonal imbalance into the equation. Some instruments may sound relatively natural, but others, due to their tonal balance and harmonic structures possibly being altered, will never sound "right".

As such, a neutral response assures us that our system is running like a computer i.e. "GIGO" ( Garbage In, Garbage Out ). If you have a good recording, you hear what is recorded. If you have a garbage recording, you hear what is recorded. Otherwise, you don't hear what is recorded, but what the system's colourations and distortions have added to the recording. On top of that, you'll hear this same colouration / distortion on every recording that you listen to. While you might like that specific sonic signature, i sincerely doubt that it would be "accurate" in terms of what one would hear at any given recording session or live performance. It might sound "pleasant" or "enjoyable", but accurate would be a long shot.

Having said that, i don't think that the problem of flat frequency response is what you are concerned with Newbee. I think that the approach that your mentioning here is more of an attempt to "band aid" poor mic'ing and recording techniques as used by most engineers and limited bandwidth in the recording chain. This type of system is probably going to sound "warm and smooth", lacking leading edge information and "bite". If we were to view this on a scope, the front of a square wave would show limited rise time with rounded edges. Some would refer to this type of system presentation as "sugar coating", which helps "sweeten" bad recordings. While this can sound very "pleasant", it simply isn't "accurate" to what the music or recording "should" sound like. There's nothing wrong with this, but chances are, a system built like this will work better on specific genres of music and individual recordings than it does universally. If one doesn't listen to a wide variety of musical stylings and is happy with that type of limited presentation, there's nothing wrong with this. Having said that, this is where we get into the divergences between "music lover" and "audiophile". For lack of better terms "audiophiles" want to preserve the music as recorded and "music lovers" want every recording to be "pleasant". Different outlooks with different goals. That's why comparing notes on various components / systems can be so wide-ranging i.e. people are looking and listening from different perspectives.

As a side note, when i recorded "demo's" for a few local bands, i placed mic's nearfield to the instruments and / or amps and then two mics out away from the "stage". Those two "audience" mics were what the main mix consisted of and i could use the individual instrument mics to highlight certain instruments as needed. This gave an overall "natural" presentation as one would hear from being in the audience in the nearfield. At the same time, it allowed us the flexibility to alter / highlight individual parts of the presentation as we saw fit. I was able to retain the "musical presentation" of the group as a whole while still being able to perform technical changes without detracting / distracting from that presentation.

The "funny" thing about this is that after working with a few bands and making recordings like this, several of them came back to compliment me months / years later. After working in "Professional" studios, they were happier with the sonic results that we had achieved recording "on the fly" in the warehouse that i used. Yes, the "Pro Studio" was able to achieve a higher "gloss factor", but the music lacked the proper feel and presentation that the band was all about / wanted to deliver. Most of the differences probably had to do with heavy handed production techniques, phenomenal amounts of compression, poor microphone techniques and being in a non-friendly environment ( affecting the mind-set and performance of the performers ).

All of these were things that i tried to avoid as i knew them to be detrimental to what i and these specific bands were trying to achieve. That is, i wanted to capture the sound of the band, not turn them and the recording into something that they weren't. Using a system that isn't "flat" does much the same thing as what the heavy-handed recording engineers did i.e. added "distortions" to the music and presentation that wasn't really there to begin with. If one likes that given distortion characteristic(s), it is their system and recordings to do with as they please.

As mentioned above though, flat frequency response is only a small portion of what we hear. High levels of linearity in the frequency domain combined with high levels of linearity in the time domain ( transient response ) are what we should be striving for. At the same time, we should be pushing the recording industry to get their act together. While we can build our systems to achieve the first two thirds of the equation, i think that we will always be limited by the quality of the source recordings. That is, so long as "engineers" are doing the recording and "audiophiles / music lovers" have no input into the procedure. Sean
>

Sean wrote:
As such, a neutral response assures us that our system is running like a computer i.e. "GIGO" ( Garbage In, Garbage Out ). If you have a good recording, you hear what is recorded. If you have a garbage recording, you hear what is recorded. Otherwise, you don't hear what is recorded, but what the system's colourations and distortions have added to the recording. On top of that, you'll hear this same colouration / distortion on every recording that you listen to. While you might like that specific sonic signature, i sincerely doubt that it would be "accurate" in terms of what one would hear at any given recording session or live performance. It might sound "pleasant" or "enjoyable", but accurate would be a long shot.

This is what I experienced using Paul Lam's LW-1 Passive Controller. It is the only way one can accurtely test equipment or to see what is happening when designing something.

My speakers are very good also. There sensitivity to what is happening up stream is very helpful when trying to design cables.

Just thinking!

A flat measured response is sometimes fools gold. The output of a loudspeaker in a static input of, say, 85db, or 90db, is one indicator. Yet, when the dynamic signal of real music is applied, there is no emperical proof that the 'dynamic' output will remain anywhere near that first measurement.
The only (fundamental) test is listening.
I have designed speakers which measure reasonably flat, insofar as the static input goes, with tragic results with dynamic input.
Now, you could argue that multiple input volumes starting at very soft, to very loud with sweeps; then overlay the results to compare. That could give some meaningful data, but its still not the 'real' test.
Speaker design is truly 'art and science' and transcends science, and goes into the artform arena. It takes a special person who can do both, then be objective with the results. That is why there are many loudspeakers that are 'technically' correct, yet leave us wanting musically.
So, "How important...."
Somewhat, however....
Best,