is good question . Flat response with minimal as possible deviation mean no or minimal coloration to sound. this is necessary condition to Hi End BUT NOT enough . there are different technical issue . Conclusion--- You have to choose Loudspeaker with Flat response and DO A LISTENING test , If both conditions passed, go ahead.
There are no good reasons for an electronic component to not have flat frequency response. Loudspeakers have a different set of issues revolving around expected room interactions. Flat frequency response at the listener position is not desirable. It's sounds tipped up in the treble and unnatural. Most listeners would prefer a gently downward sloping frequency response at the listening position.
"Flat" means a component or speaker is not likely to color the sound on its own. The problem is that there are many different components in the system...and then there’s the room itself...and your own hearing, your mood, etc.
The problem is not so much the individual component or speaker. The problem is in trying to get all the different freq. curves (each with their own slight variations) to work together and add up to one final in-room response that you can live with...flat or otherwise.
The big lie is to think that because you’ve bought a component or a speaker with a "flat" response (or in fact any stated response) that therefore you will be assured that it will remain so in your system and in your room. It isn’t so. The room may dominate, other components may dominate. Bad interaction with certain product combinations may dominate. Even things like a lack of power-conditioning may dominate. All you have bought is a component that will ’allow’ for a flat response under certain conditions (sometimes those certain conditions are specified and sometimes they are not), but the flatness is not guaranteed in your setup, only on the test equipment used to test that piece of gear. But, "most" gear is designed to be used successfully in "most" rooms and with "most" other gear.
A truly flat response may actually be audibly (subjectively) desirable, but almost always only with an Extremely well-executed system of rather expensive, high quality and full-range gear. In fact, it might surprise a lot of people just how much attention to detail and how much money must be spent to subjectively feel safe from any real concern with frequency response issues altogether. For the rest of us, we will need to consider it - whether a little or a lot.
You need to check out the Fletcher Munson curve. Our brains perceive frequencies at different levels, depending on the volume of the sound. Most people do not prefer a flat response. I set a room with equal volume in all frequencies about 30 years ago. It sounded very "flat," no pun intended.
Of course without a flat response any attempt at neutrality has gone out of the window. So it makes sense that all loudspeakers/ headphones should aim to have a flat response as possible as long it's not at the expense of dynamics.
Some like me, felt that too many British loudspeakers in the past with a reasonably flat response tended to sound too polite. For me that's too great a price to pay. You may feel different.
On the other hand too great a deviation from flat gives the speaker an obvious character on everything you play and you will hear this constantly.
It doesn't have to be ruler flat, the main thing is to avoid obvious spikes. Especially if you are mixing/ broadcasting/ or in any other pro use. Pro's do tend to stick with what they know.
For the rest of us some form of tone controls / equalizer is always desirable to compensate for different rooms or recordings ect.
As usual with loudspeakers, nothing is close to perfect. The best you can do, as ever in audio, is to pick your own set of compromises.
Flat freq response for what and for whom? An anechoic chamber? A high end audio show room? Tom, Dick or Harry’s room. Every room is different. And for whose ears? give me a break!
A flat frequency response means all frequencies are reproduced at the same amplitude. Even if that was somehow possible to achieve in a listening room, that would remotely resemble what live music sounds like. I think a better use of this measurement in as far as comparing speakers is concerned, is how wide of the frequency spectrum the speaker drivers can cover and not if there are ripples within that response.
There are some other important technical parameters to consider. Dynamic range, sensitivity, response to square wave, distortion. Frequency response is probably fourth or fifth down the list.
With multiple drivers the crossover from the tweeter to the mid/bass is critical. Often its hardly ever smooth enough. Too many issues with dispersion, plus the mid starts to struggle as all drive units tend to break up badly at their extremes.
With treble units its usually at the low end where a lot of male voices are.
@elizabeth , I think you are absolutely, and without any doubt in my mind, 1000% correct!!!
Easier said than done sometimes maybe...but not impossible (if you don't confine yourself only to the most immediate possible causes and are persistent). And that will pay ever bigger dividends the further one can take it - even if someone decides to take it past the point where they feel it may already be 'good enough' - always plenty of unrealized rewards to be mined there.
In a room no system is measuring flat you have to go to heroic efforts to even get close. Aiming for flat from a design perspective is so end user has a known base to work from still many designs do not aim for flat response adding more variables to system synergy. Since massive variables exist every audio system sounds and would measure differently depending on these variations.
Loudness curves - Yes, this makes it important to listen to a speaker at your normal listening volumes, or get one that is flat and use a loudness compensation like Yamaha has or Dennon used to. A speaker tuned for 65 dB listening is going to have more bass and more treble than one designed for 90 dB. Some of the Dynaudio speakers are great examples of this.
Imaging - Wilson, among others, has taken advantage of using a dip at 2.4kHz or so which gives a perception of enhanced imaging, at the loss of some information. Not all wilsons, this seems to have fallen out of favor recently.
Detail - By using a ragged frequency response some speakers can appear to enhanced detail. B&W 802D and GE Triton 5 have done this, and bowled JA over. I can’t stand them, but I am not buying your speakers.
The GE Triton 1 offered a nasty sounding AMT with a very exaggerated upper octave. To me that was painful, but apparently older listeners love them.
My last point is, your hearing is unique and changes over time, so regardless of a technical ideal, or a particular reviewer's take, you need to buy speakers that are ideal for you. If you know you have hearing deficiencies in certain frequencies, maybe that is something you should look for when buying, speakers which naturally have more in that area. Nothing wrong with that. :)
This IS a very complex and interesting topic. Have a studio that is tuned carefully to flat response and a listening room that is catered towards dense orchestral sources.
Find that the studio is crystal clear but in a direct way that many listeners would find unappealing -- especially over time.
In the listening room have tuned the entire system and room with REW and JRiver so that it is flat from 16 to 30k within +/- 5db. This is my listening preference and can easily hear each orchestral section (or soloist) in the soundstage.
However, if I boost the bass around 5db more I grant you that age and other considerations mentioned above makes the overall sound more exciting. And boosting the treble around 3db increases the presence and makes all sound closer to you (this works if, as Elizabeth noted above, your treble response is without noise and distortion).
Should you be doing correction, keep in mind that impulse response is just as important as frequency EQ in making the soundstage come alive.
An excellent topic for a thread and one that many will have different, but equally good, responses.
The quality of treble is often a function of room acoustics. Try throwing pillows and blankets on the floor between and behind the speakers as an experiment. You may be awesomely surprised at how much this enhances sound, despite the first reflection mantras dominating the topic of room acoustics.
This discussion begs this question: would two different speakers with identical frequency response curves sound the same in your room? I'm betting the answer is no.
Yes. I’ve said that. Components should be designed to be as flat as possible to be faithful to the original recordings. From that point, all hell breaks loose with rooms, our ear/brain function, personal taste, probably sunspots for all I know. Without flat component response, we have no standard for comparison. What’s to keep products from varying from one to the next? What good would it do to audition something if the example you buy is different?
Recording engineers don't record anything flat first there are really no flat microphones Actually we use our ears to get the overall sound we prefer It is never flat. Our control rooms were rolled off top and bottom forcing the engineers to boost base and treble. A perfectly flat recording in a flat room sounds terrible. Forget about frequency response and listen to what you like Alan
If you want to hear it the way it was recorded and intended to be listen to, then flat is best. If you don’t like it, buy a equaliser and do what you consider to be the best. http://www.schiit.com/products/loki
This discussion begs this question: would two different speakers with identical frequency response curves sound the same in your room? I'm betting the answer is no.
There are so many other variables. Impulse response, compression, dispersion (power response) and distortion among a few of them.
ESL's have often really mediocre FR but their lack of reflections make up for it in detail and imaging.
I second your thoughts on experimentting with room acoustics.
I designed my listening room with an acoustician so it’s a good sounding room. But I also have some leeway in adjusting the reflectivity af various points. I have a shag rug (yeah, baby!) and the ceiling is a drop down structure covered in felt with various portions of heavy absorption areas hidden behind. So just those two elements alone tend to reduce a lot of room reflection hash etc for a tmbrally smooth sound.
But I also have a variety of curtains gatherered in the room corners that I can pull across any specific area of the wall to kill other upper frequency reflection points, or alternatively liven the sound by allowing more room reflection. It’s great having this flexibility (and a recently added diffuser has increased the control over my room acoustics).
As to your suggestion for people to try dampening behind the speakers I can attest to that as well. Several feet behind my main stereo speakers is a wall sized projection screen with a 4 way velvet masking system. This means I can make the back wall behind the speakers very reflective (velvet curtains wide open revealing reflective screen surface) or completely covered in velvet, or anything in between. Only adjusting the reflectivity of the back wall really has an effect on the room sound. Too much and it becomes too dead. But a judicious screen size setting behind my speakers leaves a really nice balance. The more I close the velvet the more the ambience tone and reverb comes only from the recording. Opening the curtains makes the sound more open and lively.
kosst_amojan I completely agree with Elizabeth and Erik. It's real easy to get nasty treble when it's bouncing all over a reflective room and that jazz needs tamed.
There are a few subjects being discussed here: Flat response from Equipment and speakers Flat In room response. If you think about it, when the recording is done, are they done in perfect rooms? Carnegie Hall, Studios, Restaurants, Church's. Every recording will reflect the room that it was recorded in, then our room has an effect again on what we hear. The only way to hear the recording as it was done is with a flat room. Ideally, all recordings would be done in a perfect anechoic chamber and our rooms would be the same, but its tough, shoving a symphony in a chamber. Next, for Equipment, I do believe that it is important for our equipment to be producing a flat response. That is the only way to faithfully reproduce the recording, assuming that we can have a perfect room. I believe that it was @kalali that posed the question, would 2 different speakers with identical response curves sound the same in the same room. I can answer that, I have built many speakers and have come EXTREMELY close in duplicating response curves, but I can tell you for a fact, not opinion that different driver materials sound different and on crossovers, even using the same slopes and same type of compensations or no compensation that crossover parts still sound differently from each other. Overall, my contention is Absolutely, you need to come as close as possible to achieving a flat response curve on your electronics, in your speakers and in your room. If that is achieved. It is then that you get as close as possible to creating that all allusive "Live" sound that we all work so hard to achieve. Plus when things are acoustically flat, it does help hear so many other design presentation differences in amps, pre's, cables etc.
Look around; you'll see that even ruler flat response bears little correlation between the perceived quality/acceptance of a speaker by the listener.
I have handled enough speakers which deviated in some respects which others might consider unacceptable, yet they can be made to sound beguiling. People get hung up on the search for flat response, just as they� get hung up on the search for "The One," the perfect speaker. Imo, fairly fruitless efforts.
I have heard plenty of systems which were purportedly "dialed in" to be flat response. Meh; they were pretty "flat" all right.
The market has just not responded that well to it. While Vandersteen and Thiel have ardent followers they also have not dominated the high end speaker market. Neither have single driver designs.
@douglas_schroeder Hi Doug, although I do agree that there are several speakers that are nice to listen to that don’t necessarily come close to that ruler flat acoustical challenge. I have to disagree on the idea that seeking and building toward that flat response a fairly fruitless effort. All Amplifiers and Pre Amplifiers come VERY close in a flat response, but as you know, One can and will sound much different than another. The flat response is no more than the idea that everything starts with a level playing field, but yes, you are absolutely correct that a totally flat response curve does not at all guarantee excellent results and/or a system can be satisfying with all kind of things being out of wack. @ieales As far as the Phase discussion, there are a ton of threads on Phase (not to be confused with phase angles), but phasing affects imaging and sound stage much more than it does frequency curves. Tim
If your gear measures perfectly flat on-axis in an anechoic chamber, and it has ideal off-axis (steadily and evenly sloping the further you go), then it will sound good in a “typical room”.
Now, the room does play a large role, if you are in an untreated room that has tile floors and large windows, you may want a speaker that has a recessed treble.
Handyman has it right check out The Fletcher Munson Equal Loudness article. This is one reason you see a "loudness" switch or adjustment on some gear. Frequency response changes with volume.
The performance space affects the sound on the way to the recorder. Your room affects the sound that comes to your ears.
Having amplification that has an output that is exactly the same as the input except for variable volume, just gives you the chance to hear something like what you would have experienced at the performance - except for your other gear and your listening room, which is unlikely to be totally passive.
But if you can at least eliminate one source of colouration, I'd view that as a positive move.
Speakers need to be "voiced," i.e. you have to listen to them. I talked to a speaker designer years ago who said that if you had the ability to adjust the various sonic parameters of a speaker from your listening spot (including cabinet materials, crossover settings, speaker materials), the results you would wind up with from simply listening to music would be far from flat.
I think most music lovers and audiophiles can agree that there is no such thing as “flat” anything - especially speakers. Although, it is probably a good thing not to have prominant peaks or dips in a speakers’ response. That said, I believe that the key ingrediant to great sound is the room itself. Assuming your equipment is well matched, and you are at a certain price point, you can get more bang for buck in room tweaking than in component changes. I have 2 systems that I listen to frequently. The one with superior room acoustics sounds better, no matter which equipment I use in it.
Come on people. Speakers’ frequency response is almost always specified along the lines of 40 Hz to 16 kHz on axis + - 2dB. Their sensitivity is almost always specified along lines of 92 dB for 1 Watt at 1 meter. Just like stereo cartridges are usually specified along the lines of 15 Hz to 20 kHz +- 1dB.
For everyone, geoff is accurate, speakers publish their spec's with a frequency response and a variance within that response. Exceptionally flat speakers are normally rated Plus or Minus 1 db. Anything that is typically more that plus or minus 3db is normally not considered audiophile, although that certainly is not cast in stone. Today, Many speakers are rated at a flat 2.83v input rather than 1 watt 1 meter. For accuracy, you need an accurate impedance, an accurate response curve and an accurate sensitivity rating.
A further word on J Gordon Holt. He conducted most of his listening evaluations utilizing master tapes which he made of acoustic classical performances. And he seemed to posses a very good auditory memory. That allowed him to make value judgements on how well given components came to reproducing those live creations. How many reviewers today do anything similar?
I offer this since many readers may be too young to have much appreciation for Holt and what he meant to audio reviewing. He cared about electrical performance, but his ears were the final arbiter.
Speakers posting their frequency response often are challenged by measurement (for example, every speaker reviewed in Stereophile), with peaks and valleys here and there all over their range. This might not mean much if the designer listened to them, which is generally the case. Also the frequency response noted by manufacturers is often utterly incorrect...I have a pair of very coherent, great sounding Silverine Preludes, which list the low end as -3db at 38hz, when at that point they're more likely -10db at least. My Klipsch Heresy IIIs are rated to around 58hz and that's pretty much exactly right...my subs take over at that point so it's all good.
IMO, phase has been ignored because so few people have heard a properly phased system. Designers who don't understand the importance choose to ignore it and the media play along. Few systems are phase correct to begin with and those that are rarely setup correctly. Very few people who buy HiFi ever hear unamplified music in a good hall and thus have no reference to an acoustic space.
The Absolute Sound recently reviewed a time-aligned 2-way loudspeaker with 1st order crossovers. Unless the drivers are perfect, it is simply not possible to achieve minimal phase error with that design.
I think a bass hump was engineered in over the years simply to make bookshelf speakers seem more alive and warm. However, if it's not in there, and more accuracy is involved in the design, you can utilize subs to make up the difference. The bottom line with this stuff is for a speaker to deliver a coherent and reasonably accurate sound that listeners actually like.
You must have a verified phone number and physical address in order to post in the Audiogon Forums. Please return to Audiogon.com and complete this step. If you have any questions please contact Support.
