My dogma is flat. It needs some air....

[Branwell]

Seems a given that because of the way we hear relative to SPL ( Fletcher Munsen / Robertson Dadson ), combined with playback technology and the listening rooms effect on reproduced sound, recorded music played through speakers with a flat measured frequency response is almost guaranteed to sound wrong.

This raises a question. Is there a way we can insert EQ to get reliable accuracy?

If we knew the environment, the SPL, the FR etc of the mastering speakers, and how the mastering engineer was perceiving sound the day he did the mastering, and a system existed that could correct for our room, speakers and how we are perceiving on the day, perhaps, but…..

Where does this leave us?

Seems to me we are left with self reliance and educated guessing…..

First off, I think we need to ditch the dogma that a flat FR is a correct FR. Once we get that conditioning and its effects out of our heads, I think we can become more open to the possibility of embracing technology that allows us to adjust a speakers FR on the fly as we listen to music and rely on our ears, insted of a graph, to tell us when it sounds right and most enjoyable.

A side benefit of pumping some air into the dogma is wider driver choice.
As part of the selection process, many people give driver FR flatness a high priority. If flatness has little to do with good sound, it opens the doors to a wider choice of drivers.

Any thoughts on this?

branwell

[Kal Rubinson]

Seems a given that because of the way we hear relative to SPL ( Fletcher Munsen / Robertson Dadson ), combined with playback technology and the listening rooms effect on reproduced sound, recorded music played through speakers with a flat measured frequency response is almost guaranteed to sound wrong.

Sorry. What they show is that human perceptual frequency responses are level-dependant. (Any other variation in spectral response is a fixed variable for both live and reproduced sound and, therefore, needs no compensation.) If the playback is at the same level as the original, then flat response is correct and appropriate. If one wants or needs to compensate for a lower SPL of the system, then it has to be dynamically matched to level changes and no static EQ will do.

I just play it as loud as necessary.

Kal

[fjhuerta]

Ugh.

I hate it when I think about this. Why? Because I want a reference, and flat is as good as it gets.

But I believe you are right, Branwell. Yesterday I was playing with my new speakers, which are pretty flat. I'm used to the sound of my Energy speakers, which have a bit of a loudness curve into them.

Yep, my "flat" speakers still don't do it for me. I can hear every last detail, but unless I raise the volume, I feel something isn't right. There's a point where everything comes together, and the speakers sing.

It's just that I believe the Energy speakers are correct quite a bit more often, and at various sound levels.

I'd suppose a loudness button would work here, but sadly, my (audiophile-correct) preamp doesn't have any. Or tone controls. Or anything like that.

Meanwhile, I look at my speakers and wonder if I should add just a tiny bit of extra treble.

Who said perfection was an easy goal?

[cjd]

What're your current speakers?

Your ears can take a couple weeks to a couple months to forget they have to clamp down and filter high frequency noise from using speakers that are tuned "hot"

Depending on the drivers you have, there could be more going on.

C

[Dennis H]

I stopped by SL's page to see how the progress was going with the Orion+. The latest is he's running the front tweeter at full volume and cutting 3dB from the rear tweeter. Anyway I noticed an interesting comment about the whole loudness thing. He says adding the rear tweeter with a live wall behind has made has made his recordings less sensitive to needing to be played back at the 'correct' volume. Now the volume control acts more like a 'distance' control, like it would sound if you were walking closer to the stage.

http://www.linkwitzlab.com/orion++.htm

[Branwell]

Hey Kal,

For level matching alone to work, would the listeners system not have to match the acoustical particulars of the system, room included, that was used in mastering the recording?

branwell

[AJINFLA]

played through speakers with a flat measured frequency response is almost guaranteed to sound wrong.

What exactly does this mean? A speaker system is fed an input signal and outputs it in 3 dimensions. What frequency response are you referring to?

First off, I think we need to ditch the dogma that a flat FR is a correct FR.

We do? So what type of signal chain should we use? A non-flat CD player feeding a non-flat preamplifier feeding a non-flat amplifier. Should it be cumulative or is just one enough? Based on what?

But I believe you are right, Branwell. Yesterday I was playing with my new speakers, which are pretty flat.

The top of my dining table is pretty flat. What is "pretty flat" about your speakers?
Thanks in advance.

cheers,

AJ

[fjhuerta]

The top of my dining table is pretty flat. What is "pretty flat" about your speakers?
Thanks in advance.

cheers,

AJ

Flat FR from about 200 to 15 or so KHz. But you knew that, didn't you? :B

[cjd]

Hmm, flat measured how?

[Kal Rubinson]

Hey Kal,
For level matching alone to work, would the listeners system not have to match the acoustical particulars of the system, room included, that was used in mastering the recording? branwell

That's not what I said. I said that the EQ has to be different for each level.

Kal

[Kal Rubinson]

Ugh.

I hate it when I think about this. Why? Because I want a reference, and flat is as good as it gets.

But I believe you are right, Branwell. Yesterday I was playing with my new speakers, which are pretty flat. I'm used to the sound of my Energy speakers, which have a bit of a loudness curve into them.

Yep, my "flat" speakers still don't do it for me. I can hear every last detail, but unless I raise the volume, I feel something isn't right. There's a point where everything comes together, and the speakers sing.

It's just that I believe the Energy speakers are correct quite a bit more often, and at various sound levels.

I'd suppose a loudness button would work here, but sadly, my (audiophile-correct) preamp doesn't have any. Or tone controls. Or anything like that.

Meanwhile, I look at my speakers and wonder if I should add just a tiny bit of extra treble.

Who said perfection was an easy goal?

What you are talking about are matters of taste and adaptation and they have little to do with the issues of flatness and accuracy.

Kal

[fjhuerta]

Hmm, flat measured how?

15° off-axis.

[fjhuerta]

What you are talking about are matters of taste and adaptation and they have little to do with the issues of flatness and accuracy.

Kal

Kal, something surprised me from your comment. I don't know if you share their opinions, but I unsubscribed from Stereophile because some reviewers (whose names elude me at the moment) usually loved gear that John Atkinson demonstrated to be simply awful.

An example that pops into my mind is the review of the Tetra 505 loudspeaker. I think that was the exact same moment I knew I had to build speakers myself - that I could beat those things by a mile for 1/6th their price. Their measurements looked worse than any loudspeaker I've ever built (and that's saying something - you should see the stuff I was coming out with 3 years ago!).

The thing is, some reviewers at Stereophile seem to value many things other than "accuracy" or "flatness". I'd suppose you are not one of them - still, it was kinda funny to see a comment like yours coming from someone working at Stereophile (other than Mr. Atkinson, of course!).

[joecarrow]

You know, the fancier digital equalizers such as the DEQ2496 can give you dynamic equalization; they change the frequency response based on input level.

Just a quick check of the specs showed me that they only have three dynamic EQ spots available, so they're probably just used to limit excursion at the low end of driver bandpasses as drive level increases.

A fancy parametric EQ with the ability to save lots of presets might be useful. Also, if you use your PC as your audio source you can get the Shibatch parametric equalizer to use with Winamp for free, and create several "equal loudness" EQ curves for different listening levels.

It's true that an accurate speaker will sound kind of dim and lifeless at low SPL due to your perception of sound. If you go through Linkwitz's description of the Pheonix project, I believe that a circuit to address this issue is described.

[AJINFLA]

Flat FR from about 200 to 15 or so KHz. But you knew that, didn't you? :B

No, I didn't.
Once again: A speaker system is fed an input signal and outputs it in 3 dimensions.
Do you understand what this means in relation to a single point in space (15 deg off axis in what direction at what distance)? What relevance does "measured flat" have at a single point in space? Are these headphones or loudspeaker systems in rooms? Do you listen elevated outdoors or in an anechoic chamber?

cheers,

AJ

[Dennis H]

I've never liked the sound of 'loudness compensation'. Back in the day, I had a preamp where the loudness EQ control was a pot rather than a button so you could dial in however much you wanted. Never used it, it just didn't sound right. People may disagree with SL's philosophy of speaker design but I think he has the right idea about loudness comp. At a live concert, walking farther from the stage will make the SPL go down but the frequency content doesn't really change. So, however you get there, your speakers/room should duplicate that effect. Turning down the volume should sound like you're moving farther from the stage while still preserving the basic location of the instruments in the soundstage.

[fjhuerta]

No, I didn't.
Once again: A speaker system is fed an input signal and outputs it in 3 dimensions.
Do you understand what this means in relation to a single point in space (15 deg off axis in what direction at what distance)? What relevance does "measured flat" have at a single point in space? Are these headphones or loudspeaker systems in rooms? Do you listen elevated outdoors or in an anechoic chamber?

cheers,

AJ

So...

Uh... no... I see your point, but I fail to see your conclusion. If I understand correctly, you are implying to flatness isn't a goal, and that a completely flat system can still be wrong?

If so, then what are the variables we have to optimize in order to have an enjoyable speaker system?

[fjhuerta]

You know, the fancier digital equalizers such as the DEQ2496 can give you dynamic equalization; they change the frequency response based on input level.

Just a quick check of the specs showed me that they only have three dynamic EQ spots available, so they're probably just used to limit excursion at the low end of driver bandpasses as drive level increases.

A fancy parametric EQ with the ability to save lots of presets might be useful. Also, if you use your PC as your audio source you can get the Shibatch parametric equalizer to use with Winamp for free, and create several "equal loudness" EQ curves for different listening levels.

It's true that an accurate speaker will sound kind of dim and lifeless at low SPL due to your perception of sound. If you go through Linkwitz's description of the Pheonix project, I believe that a circuit to address this issue is described.

I've been reading SL's site (specifically, the part you mention).

He advocates a 4 dB "BBC dip" (IIRC), from 1 - 3 KHz, since he says microphones pick up diffuse signals and direct ones just the same, while the human ear doesn't. And that the BBC dip helps in this regard.

I may try this. One thing he said (you can play your system louder without so much strain) makes a lot of sense. As they are, my (flat on a single point in space, from 200Hz - 16 KHz, 2800 meters above sea level, cloudy day, 10° C, etc.) speakers can get pretty strident in the midrange with saxophones and similar instruments.

[Kal Rubinson]

Kal, something surprised me from your comment. I don't know if you share their opinions, but I unsubscribed from Stereophile because some reviewers (whose names elude me at the moment) usually loved gear that John Atkinson demonstrated to be simply awful.
-SNIP-
The thing is, some reviewers at Stereophile seem to value many things other than "accuracy" or "flatness". I'd suppose you are not one of them - still, it was kinda funny to see a comment like yours coming from someone working at Stereophile (other than Mr. Atkinson, of course!).

My opinions are my own as are those of other reviewers.

Kal

[Kal Rubinson]

I've never liked the sound of 'loudness compensation'. Back in the day, I had a preamp where the loudness EQ control was a pot rather than a button so you could dial in however much you wanted. Never used it, it just didn't sound right. People may disagree with SL's philosophy of speaker design but I think he has the right idea about loudness comp. At a live concert, walking farther from the stage will make the SPL go down but the frequency content doesn't really change.

I generally agree with you but the frequency content does change subtly, depending on the local conditions.

So, however you get there, your speakers/room should duplicate that effect. Turning down the volume should sound like you're moving farther from the stage while still preserving the basic location of the instruments in the soundstage.

That's the ideal but the room's acoustics will influence the effect.

Kal

[joecarrow]

These are actually two different phenomena- the BBC dip has to do with the in-room response of acoustic instruments, the combination of direct and reflected sound, not with listening volume. The "loudness" issue is your ear's varying sensitivity to sound at different frequencies. Either or both of these can help or harm how your speakers sound to you, and are worth experimenting with.

[ThomasW]

I liked some of the original Yamaha gear. The higher the volume control was turned, the lower the amount of loudness compensation.

[wkhanna]

Flat is an accepted, real world constant to most of us, I believe. It’s sort of like the speed of light is a constant, and used as a baseline for so many theories. Theory can be and is, in many cases subjective in motivation. Just as one’s own perception of sound is subjective.

As a novice, I’m just getting my head wrapped around the concept of FR as it relates to the realistic reproduction of sound. So please, don’t run over my dogma with your karma. (Insert smiley face here)

[Branwell]

Hi Joe,
There might well be no connection, but the “BBC” dip and the curves developed by FM certainly seem to have similarities in the midrange. Have a look. http://en.wikipedia.org/wiki/Fletcher-Munson_curves

Hi Javier,
Related to “strident” sound.

I had a problem that drove me nuts for a time and it was that some recordings sounded as you put it, strident.

In investigating this, I found that in my speakers that use 1” domes with low crossover points, say less than 2.5K, the tendency for stridency was higher than the same speakers using higher crossover points.

I also noticed that my speakers capable of over 120db ( big drivers and compression tweeters ) almost never sound strident.

Looking further, I setup two speakers using a DEQX….
One speaker used a good 1” dome crossed at 1800hz. The other used the same mid, but crossed to a compression tweeter in a wave guide.

I put on a song that can sound strident and noticed that when it did sound strident ( with the speaker with the dome tweeter ), the sound lost detail and sounded compressed.
Switching to the other speaker, I noticed that where the dome had become harsh, the compression tweeter did not. I noticed that while tonally similar, there was no loss of detail and no compression.

To further the test ( using a cheap dome this time ), I hooked up a 500 watt amp and played a recording that tended not to sound strident. No problem for the speakers.
Next I put on a recording that does sound strident, matched average SPL, and within a minute, the tweeter smoked.

The conclusion I drew from this, and I am probably over looking something significant here, but the conclusion I came to was that there is something about strident sounding recordings that just catch dome tweeters wrong. Perhaps there is something in the recordings that resembles a distorted signal which push domes over the edge.

I realize none of this would stand up to any sort of scientific scrutiny, and I accept that. I’m not a scientist :B

branwell

[cjd]

So...

Uh... no... I see your point, but I fail to see your conclusion. If I understand correctly, you are implying to flatness isn't a goal, and that a completely flat system can still be wrong?

I don't believe this is what he is getting at. You keep harping on "flat" so he's trying to point out that there IS no "flat" the way you're using the term. Move your head two inches and response can change quite a bit (particularly bass response in-room). Or it may not.

I get dramatically different results based only on whether I'm sitting up straight or leaning back in my chair (leaning back both brings my head closer to the head-rest and at some point brings my head a bit lower). Quite a simple procedure.

When you say "a completely flat system can still be wrong" you're somehow wrapping up EVERY possible measurement in that comment. And as soon as you don't take everything, then the answer is "yes, absolutely, a flat measurement done such and such a way and ignoring all others CAN be completely wrong"

A very obvious means of getting such a result is looking only at one meter measurements perpendicular to the tweeter plane... a classic "on axis" measurement. You can put drivers together to measure perfectly flat in this plane using negative summing. Looks real super. Power response, measurements on about any other axis, all that stuff, is HORRIBLE and it'll likely sound quite poor as well.

FWIW I find "BBC" dip to sound slightly false.

C

[AJINFLA]

So...

Uh... no... I see your point, but I fail to see your conclusion. If I understand correctly, you are implying to flatness isn't a goal, and that a completely flat system can still be wrong?

If so, then what are the variables we have to optimize in order to have an enjoyable speaker system?

What conclusion have I made? I am pointing out to you, Branwell and a great many others who repeat this, that you are placing a microphone at a spherical coordinate x,y,z in your loudspeakers radiated soundfield, seeing a "flat response" on a screen and concluding that "Flat speakers don't sound optimal". Try some other coordinates to see just how flat your flat speaker is.

cheers,

AJ

[AJINFLA]

FWIW I find "BBC" dip to sound slightly false.

IMHO it is simply compensation for excessive radiation in the midband from uncontrolled directivity at XO. DI chaos. Reducing power in this region is a poor solution. IMHO .

cheers,

AJ

[Branwell]

Hi,
There are a lot of variables to speaker building, and one of the most cherished by many is the flat frequency response. It is this belief that I feel is the downfall of many potentially great systems.

As has been pointed out: Flat, flat where? Flat on one axes is often not flat on another and how it all sums to the ears perceived sound at different SLP levels just makes matters worse.

I believe that if we disregard flat and simply built what sounds good to us the listener, we will end up with more enjoyable systems and that’s was the spirit of the original post was meant to convey.

branwell

[cotdt]

the recording engineers use studio monitors which are "flat". so to hear the sound as intended, you would need flat speakers. they are what is accurate as defined by what recording engineers want us to hear. but sometimes a little response tweaking or even adding distortion can make intruments and voices more realistic, since it would correct for poor recording techniques, but this is song/movie specific.

[Andy_G]

speakers can get pretty strident in the midrange with saxophones and similar instruments.

this is generally what real brass instruments do... so where is the problem. ;-))

yes , it can be a bit too much, but it is real. !

[cjd]

I believe that if we disregard flat and simply built what sounds good to us the listener, we will end up with more enjoyable systems and that’s was the spirit of the original post was meant to convey.

Oh sure, fine. Yeah, let's do that. Um. Yeah. Oh, you want more people than just YOu to listen? Hmmm, too bad it makes me want to run away crying because the top end is way to hot and the midrange is mushy and... EVERY speaker I have heard that was made "so it sounded good" and without use of measurements was NOT a good sounding speaker. Or rather, they could sound pretty good but they felt "wrong" - they didn't let me get into the music.

Measuring speakers is one of the trickiest parts of doing goo ddesign work. It's too easy to not really understand what you're doing, to miss something.

this is generally what real brass instruments do... so where is the problem. ;-))

yes , it can be a bit too much, but it is real. !

Andy, you're asking too much. You want people to make a system that sounds REAL?! Nooo, you have to have those silky smooth female vocals so Ms. Krall sounds like she's right there in your living room, the bass player just there, the drummer right over there, everything nicely laid out and so...

Yeah. Most people don't have a farking clue what things actually sound like. They've never HEARD a live female jazz vocalist and how harsh that voice can be when pushed. They've never heard a live trumpet blaring into their face from a hundred meters out.

Folks, just go take your floppy coned mucked response with added distortion just where you like it from your carefully selected glowing amplification and stop trying to stir up crap.

31 years as a musician (classically trained) has taught me a thing or two about listening. One thing I have noticed is that what I hear that sounds good is ALSO flat when measured, assuming it is measured WELL. I can not always put my finger on exactly what is wrong by listening (I've not been doing speaker design nearly long enough) but sometimes I do. And I have always managed to figure it out when looking at good measurements.

Your ears don't lie, but your brain is VERY good at it, and often lies to you so you feel better. Oh yeah. And your ears can only talk to your brain, so the brain very often ignores their information.

You'd think it was a government or something up there in our heads.

C

[Evil Twin]

Oh sure, fine. Yeah, let's do that. Um. Yeah. Oh, you want more people than just YOu to listen? Hmmm, too bad it makes me want to run away crying because the top end is way to hot and the midrange is mushy and... EVERY speaker I have heard that was made "so it sounded good" and without use of measurements was NOT a good sounding speaker. Or rather, they could sound pretty good but they felt "wrong" - they didn't let me get into the music.

Measuring speakers is one of the trickiest parts of doing goo ddesign work. It's too easy to not really understand what you're doing, to miss something.



Andy, you're asking too much. You want people to make a system that sounds REAL?! Nooo, you have to have those silky smooth female vocals so Ms. Krall sounds like she's right there in your living room, the bass player just there, the drummer right over there, everything nicely laid out and so...

Yeah. Most people don't have a farking clue what things actually sound like. They've never HEARD a live female jazz vocalist and how harsh that voice can be when pushed. They've never heard a live trumpet blaring into their face from a hundred meters out.

Folks, just go take your floppy coned mucked response with added distortion just where you like it from your carefully selected glowing amplification and stop trying to stir up crap.

31 years as a musician (classically trained) has taught me a thing or two about listening. One thing I have noticed is that what I hear that sounds good is ALSO flat when measured, assuming it is measured WELL. I can not always put my finger on exactly what is wrong by listening (I've not been doing speaker design nearly long enough) but sometimes I do. And I have always managed to figure it out when looking at good measurements.

Your ears don't lie, but your brain is VERY good at it, and often lies to you so you feel better. Oh yeah. And your ears can only talk to your brain, so the brain very often ignores their information.

You'd think it was a government or something up there in our heads.

C

Impressive... most impressive. Perhaps you are a Jedi, Chris- the force of will is clearly strong in you.

Clearly you are not a pupil, but have become a Master. I could not have said it better myself, especially as Forum decorum prevents me from expressing myself with my natural vehemance.

[speedle]

I liked some of the original Yamaha gear. The higher the volume control was turned, the lower the amount of loudness compensation.

Sherwood stuff did that too, in the mid '70s IIRC. My speakers were too efficient and my room too small to use it, but it did vary.

[fjhuerta]

Oh sure, fine. Yeah, let's do that. Um. Yeah. Oh, you want more people than just YOu to listen? Hmmm, too bad it makes me want to run away crying because the top end is way to hot and the midrange is mushy and... EVERY speaker I have heard that was made "so it sounded good" and without use of measurements was NOT a good sounding speaker. Or rather, they could sound pretty good but they felt "wrong" - they didn't let me get into the music.

Measuring speakers is one of the trickiest parts of doing goo ddesign work. It's too easy to not really understand what you're doing, to miss something.



Andy, you're asking too much. You want people to make a system that sounds REAL?! Nooo, you have to have those silky smooth female vocals so Ms. Krall sounds like she's right there in your living room, the bass player just there, the drummer right over there, everything nicely laid out and so...

Yeah. Most people don't have a farking clue what things actually sound like. They've never HEARD a live female jazz vocalist and how harsh that voice can be when pushed. They've never heard a live trumpet blaring into their face from a hundred meters out.

Folks, just go take your floppy coned mucked response with added distortion just where you like it from your carefully selected glowing amplification and stop trying to stir up crap.

31 years as a musician (classically trained) has taught me a thing or two about listening. One thing I have noticed is that what I hear that sounds good is ALSO flat when measured, assuming it is measured WELL. I can not always put my finger on exactly what is wrong by listening (I've not been doing speaker design nearly long enough) but sometimes I do. And I have always managed to figure it out when looking at good measurements.

Your ears don't lie, but your brain is VERY good at it, and often lies to you so you feel better. Oh yeah. And your ears can only talk to your brain, so the brain very often ignores their information.

You'd think it was a government or something up there in our heads.

C

Wow, so much hate and anger! Young Padawan, I now know why Evil Twin wants you by his side!

(Just kidding).

Seriously - it's a good thing we are all nerds, and that our weapons are HP-48GX calculators - otherwise, someone could get hurt.

Me? I'm still tryng to understand why things sound like they do, what makes speakers sound real, and what is right and wrong (more to the point - I'm re-learning lots of things I always thought were correct) and I post here because I'm here to learn from people who, quite obviously, know a lot more than I do (like you, C).

[cjd]

WI post here because I'm here to learn from people who, quite obviously, know a lot more than I do (like you, C).

You got that right. I mean, that's why I post here too. I'm assuming that's what you meant with the (like you, C) bit. While Evil Twin may have found it in him to agree with me, I know that in truth I am not yet a Master.

C

[crackyflipside]

I stopped by SL's page to see how the progress was going with the Orion+. The latest is he's running the front tweeter at full volume and cutting 3dB from the rear tweeter. Anyway I noticed an interesting comment about the whole loudness thing. He says adding the rear tweeter with a live wall behind has made has made his recordings less sensitive to needing to be played back at the 'correct' volume. Now the volume control acts more like a 'distance' control, like it would sound if you were walking closer to the stage.

http://www.linkwitzlab.com/orion++.htm

HEY! I have been doing this as well with my computer room. I had an extra pair of speakers which at first I used as reinforcement for the original pair of speakers.

I decided to fire the reinforcement speakers into the rear which is a corner and a very lively concrete wall. I also found at that level matching the front firing speakers with the rear firing speakers sounded awful, then I closed my eyes and started adjusting the volume of the rear speakers without moving my head a millimeter and found the spot which guess what??? it is 3db lower than the front speakers at my listening position (Actually I measured a -4db difference from the blind-volume-preference changing so i rounded up to -3db)!!!

I did this test last week and I know EXACTLY what that guy is talking about with the rear tweeter idea. Now, I am going to confess I have some crappy computer speakers but anyone who hears them says they sound great (I have tuned and tweaked the hell out of it so the only thing I have bottlenecking the system is the quality of the speakers and not anything else.

Having extensively tested the effect of having a pair of speakers reversed I can say it does cause a strange effect. When the rear firing speakers are being played exclusively, you can hear a sligh cut in the high frequency output from the speakers direction BUT there is a sense of surround and a widened soundstage (it really sounds like the high frequencies are flowing over my head and behind me; very strange) But when the front firing speakers are added into the equation (and are set 3db louder than the rear firing speakers) it centers the sound in front of you but maintains the widened sense of soundstage.

And to how the final effect sounds, as long as I am not stressing the speakers, which bottlenecks me to about 95db max, the effect sounds like an immense improvement to live recordings. The only times it sounds horrible is when the recording has one of those digital echo/reverb effects.

[fjhuerta]

You got that right. I mean, that's why I post here too. I'm assuming that's what you meant with the (like you, C) bit. While Evil Twin may have found it in him to agree with me, I know that in truth I am not yet a Master.

C

No, I meant I really respect your opinions, and have learned quite a bit from you. :T

[cjd]

No, I meant I really respect your opinions, and have learned quite a bit from you. :T

Deg nabbit, you won't let me TRY to be humble at least? Sheesh! :angel:

[AJINFLA]

As has been pointed out: Flat, flat where? Flat on one axes is often not flat on another and how it all sums to the ears perceived sound at different SLP levels just makes matters worse.

Finally, you're starting to grasp it. The frequency response is utterly meaningless since it depends on where as well as where else. Tells you practically nothing by itself.

There are a lot of variables to speaker building, and one of the most cherished by many is the flat frequency response. It is this belief that I feel is the downfall of many potentially great systems.

WTF:?:

I believe that if we disregard flat and simply built what sounds good to us the listener, we will end up with more enjoyable systems and that’s was the spirit of the original post was meant to convey.

Yes, Beranek predicts this. Why stop at FR? Why not sweeten the pot with some distortion too - if it sounds good to the listener? If inaccuracy for the sake of personal subjective enjoyment is your goal, why not? I think the so called fullrange driver crowd might agree with you wholeheartedly . Heck, lets start with a non-flat CD player and start working our way towards the speakers....

cheers,

AJ

[Mark K]

Now, now, let's not all get too hostile. Branwell is just, well, confused.
Relatively flat/smooth FR is the goal, along with uniform directivity and power response. There are legitimate reasons to make deviations from this, to address specific issues. Is there more to it than this?

[Dennis H]

Hey Branwell,

Are you still running your pro-driver system with waveguides and the DEQX? Maybe what you are hearing is what others have noted with waveguides, that the room power response doesn't match the on-axis response? Lack of "air" is a common description.

[Branwell]

Hi Dennis,

Am running pro drivers at the moment and lovin it…..

Lack of air?
Not for me. The BMS tweeters seriously outperform the domes I have including the Morel 110, SS 9900, Millennium etc. ( my room is fairly live so perhaps the tighter directivity of the wave guide is a good thing )

As to the general drift of this thread, I noticed the need to deviate from a flat on axes FR with all my systems including monopoles, dipoles, line arrays, WWMTM, WMT etc.
Just seems to me that flat doesn’t sound real…..and speaking of real, I see a fair amount of live music, and have musician friends that come over and play.

I know I don’t have all the answers and don’t even know enough to know what the right questions are to ask, but I do know when something sounds real and when it doesn’t….and interestingly, so does my dog. :B

branwell

[Dennis H]

Okay, so it's not lack of air. What specific deviations from flat (on axis) sound better to you?

[AJINFLA]

If I understand this correctly, you are positioning a mic at some distance on an axis perpendicular to the face of your loudspeaker. You then use an equalizer to flatten the response at this point x,y,z. Did you then measure horizontally and vertically to see what effect this flattening had on FR elsewhere? Can you show us these measurements for the monopoles, dipoles, line arrays, WWMTM, WMT built previously? Was there any attempt to correct the mismatch in directivity at XO between drivers in any of these? How was this accomplished?
If we can see all this and determine the radiation characteristics of the design not being severely flawed, then perhaps we can discount this possibilty and move on to purely your subjective tastes leaning towards boosted LF and HF? Is it possible that your hearing is rolled of at HF and the design must compensate for this?
The fact that you went from dipole back to monopole tells me you need at least 4.8db more bass. Perhaps this was not enough?

cheers,

AJ

[Branwell]

Hi Dennis, AJ,

I plan on playing with the system this weekend, so will jpg some graphs and post them.

For most of my systems, I use a DEQX 2.6P crossover / EQ / preamp.

To setup a speaker, I generally take the mic and set it 3 feet away level with the midrange ( in a WMT ) and about 10 degrees off axes. The DEQX then does its measurements.

After this, I select a range of crossover points that look to suit the drivers and let the DEQX create the crossovers, driver and BS compensation EQ, time delay, phase and GD corrections.

For example. With a WMTM pair that used 15” bass drivers, the Seas W18ex mids, and Seas Millennium tweeters, I had the DEQX create 3 sets of crossovers at 1600, 1800 and 2000. In this case I set the bass to mid at 200hz for starters.

Next, while listening to music from the listening position, I use the DEQX’s remote to flick through the crossover points until I find one I like.

After that, I have the system create more crossovers for the bass. I think I tried from 100 to 300 in 25hz increments in the case of the above mentioned speaker.

If in the process, the lo or high crossover sounds best, I’ll have the system create several more at higher frequencies. For example. If the original 3 consisted of 1600, 1800 and 2000, and 2000 sounded best, I’d have the system create more at 2000, 2200 and 2400 and give those a listen.

Once I have crossover points I like, over a period of a few weeks, I use the DEQX EQ to fine tune the sound to best suit the material I listen to at the levels I listen at. This might include preset for low listening, medium and for hi listening. I also have a couple preset for recordings that are very thin or overly bright.

The EQ I end up with is definitely system and driver dependant, but almost always ends up with a 1 to 1.5db lift from 400hz on down and many includes a 1 to 4db notch centered at around 2800hz..

During this process at different points, I mount the mic the listening position and use the DEQX room measuring function to see what the room is doing.

Pryor to using the DEQX, I did the same type of thing with passive components, but it took months.

On the dipole back to monopole thing….I had a dipole bass setup that used 4 x 12” XLS drivers. They sounded ok down to 45hz where I had to set a hi pass to get enough SPL. Overall I found the bass nice, but lacking in detail and overall SPL. To remedy the situation, I mounted the same drivers in a pair of sealed boxes and used EQ to deal with room modes. The result was the SPL I was after, but the drivers still lacked any real detail and visceral punch.
I replaced these with 15” prosound drivers and while giving up under 45hz, what’s above is of far better quality. To fix the low bass issue, I’m planning some 18” IB subs for spring. :T

branwell