 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
209.104.141.158
In Reply to: Rear-firing tweeters : gimmick or real improvement ? posted by Lambda on November 9, 2006 at 07:45:17:
"Interference" called 1/4 length wavelength cancellations (also known as destructive cancellations, or comb-filtering).This distortion of the frequency response can make the sound "more spacious", but not more accurate.
There's already too much comb filtering from room reflections -- adding a second tweeter, when one tweeter can provide sufficient SPL, can only make the frequency response worse at the listening position.
That's a sound effect or gimmick -- not an improvement.
However you always have a spare tweeter if the front one breaks!
.
.
Richard BassNut Greene
My Stereo is MUCH BETTER than Your Stereo
Follow Ups:
First of all, if what you say about comb filtering is true, most actual musical instruments which radiate HF far more omnidirectionally than most loudspeaker systems suffer the same problem, therefore comb filtering would be more accurate, not less. Secondly, there are 10 million resonance frequencies in the audible range in most rooms, comb filtering at HF would not significantly alter that. Thirdly, most HF is the result of transient attacks, not steady state sound so that by the time the reflected hf component arrives, the direct one is gone so there is no comb filtering. Fourth, by using an array of indirect firing tweeters, the frequencies at which comb filtering occurs resulting from any one of them and the front tweeter will be different from the others and in aggregate each one will be relatively insignificant.Your conclusion is not borne out by nearly 20 years of experimenting with this technique. I have yet to find a single loudspeaker system to which it is properly applied which cannot demonstrate a marked improvement in subjective accuracy and reduction of shrillness at treble frequencies. And as evidence, I have a collection of violins which are heard live frequently and a large collection of violin recordings. These recordings and others just sound more accurate to me using this technique.
With speaker enclosures of any size I've seen used in a home, the front and rear tweeters are going to be no more than a few feet apart, which means their sounds are going to be a few milliseconds apart and WILL interfere with each other (destructive interference due to out-of phase arrivals at our ears).You can't improve the treble frequency response accuracy of a speaker by using more than one tweeter.
You CAN create a "spacious" sound effect, which some listeners may enjoy, from front + rear tweeter destructive cancellations (comb filtering)
... but that's a deterioration of frequency response accuracy, not an improvement.It's very unlikely that the recording or mastering engineer used speakers with rear tweeters to make the recording -- therefore that recording will be altered by playback using home speakers with rear tweeters (or Bose 901'swith rear drivers).
Our ears blend all sounds arriving a few milliseconds apart.
.
.
.
Richard BassNut Greene
My Stereo is MUCH BETTER than Your Stereo
"by the time the reflected hf component arrives, the direct one is gone so there is no combfiltering." = ???????????????"As you know, comb filtering occurs when the reflected sound arrives at the same point as the direct sound. If the source is a steady state tone, then there will constructive and destructive interference depending on whether it arrives in phase or out of phase and by how much. And which you get will depend on where you are measuring and at what frequency. The determinant is the geometry of the room, the location of the drivers, and their spatial radiation as a function of frequency. At any given location, there will be alternating maxima and minima forming a FR graph which resembles the teeth on a comb. BUT...if the input signal is not steady state but a transient, by the time the reflected wave arrives at a given point in front of the speaker, the direct wave from that same transient will no longer be there, hence there can be no constructive or destructive interference.
"With speaker enclosures of any size I've seen used in a home, the front and rear tweeters are going to be no more than a few feet apart, which means their sounds are going to be a few milliseconds apart and WILL interfere with each other (destructive interference due to out-of phase arrivals at our ears)."
If this were a significant factor, then the sound of an insturment at stage center which is reproduced in a two channel stereophonic sound system would have a different spectral balance (other than the fact that if the speakers aren't toed inward you may be off axis, than the instrument would have being reproduced by only one speaker or the other. The interference pattern created by using this "artificial" phantom source instead of a center speaker would cause this alteration. But it doesn't seem to happen. High pitched instruments such as picolos, violins, triangles, sound the same to me whether they are at center stage or off to one side. Is that how they sound to you too?
"You can't improve the treble frequency response accuracy of a speaker by using more than one tweeter."
You can as I've already explained but I'll recap it again. In real rooms we all live in, not anechoic chambers where speakers are tested, not listened to, there will inevitably be reflections off the walls and other room boundaries but because the treble is reproduced by a speaker which confines its dispersion to an increasingly narrow beam as frequency increases, the reflections reaching our ears from the same general direction as the speaker will not have any high frequency components. This materially alters the perceived tone of the instruments and is different from the way these same instruments propagate their sound in the real world. The solution is to compensate by providing indirect HF sound complimentary to the absorption of the room so that the reflections have the same relative spectral balance as the direct sound. This flattening of spectra of total power transfer between the speaker and the listener will result in a sound which is just as clear and much less shrill than is produced from direct radiating only loudspeakers.
"You CAN create a "spacious" sound effect, which some listeners may enjoy, from front + rear tweeter destructive cancellations (comb filtering)"
That is simply untrue. If it were, you could reproduce the same effect by merely installing a comb filter in a direct firing loudspeaker. If you have a 31 or 62 band equalizer try it and let me know what you hear. I think it doesn't work. At least one reason the sound is more spacious with additional indirect firing tweeters is that human sense of direction is most accute at high frequencies. You can move off the axis of the speaker and still have reasonably flat total power transfer meaning high frequencies continue to reach your ears. I also believe that the human ears/brain direction finding system works much like a direction finding dipole antenna. As soon as your turn you head even slightly, your brain can make a subconscous judgement about the relative size of the source depending on whether the sound gets substantially louder in one ear and softer in the other ear or not. With HF sound coming from more directions than just the speaker itself, the apparant source will be more spread out. Even if you turn your head, it becomes somewhat harder for your hearing to get a lock on the direction of the source of the sound and so it appears to be wider. BTW, this lack of sharpness of direction of the source is more like what you hear in a concert hall where the overwhelming preponderence of energy at all frquencies you hear is reflected including early reflections off the stage floor and wall behind the musicians.
"It's very unlikely that the recording or mastering engineer used speakers with rear tweeters to make the recording"
This is true. But we never consider that the best sound reproduction will be obtained by duplicating the studio's monitoring equipment. If it were, most vinyl records made in the US would sound best heard through Altec A7 Voice of the Theater speakers and most vinyl recordings made in the UK would sound best through Tannoy dual concentric monitors. Probably most cds would sound best heard through B&W 801s and similar B&W models.
"therefore that recording will be altered by playback using home speakers with rear tweeters (or Bose 901'swith rear drivers)."
If you mean they will not be reproduced in the same way as they were in the studio, you are right again. But that doesn't necessarily mean worse. It could mean better. As for Bose 901, I have written extensively elsewhere about them, why I think they are flawed, and how I altered mine changing their sound substantially. Interestingly, far from increasing the apparant size of the source, the tweeters for some reason focus them in ways much like a direct speaker (depends to a great degree on how the recording was miked) but when there are many sources playing simultaneously, they are spread out and further back typical of results using the direct/reflecting principle.
The difference is the combfilter destructive cancellations that create sharp narrow nulls in the frequency response.While these nulls are too narrow to be heard directly, they combine to create what many describe as a "more spacious" sound quality compared with the point source of a mono center speaker.
A mono center speaker is better for a center vocalist.
This was discovered in the 1930's when Bell Labs decided three speakers were the minimum for a stereo effect (left, center and right) but later no one could figure out how to put three channels on vinyl, so we got a two channel compromise instead.I have experimented with center mono versus stereo vocalists going back to the late 1960's ,long before surround sound. The most natural sounding solo vocalist will be from a mono speaker on a tall stand aimed down at the listener (simulating a singer standing and singing in front of a seated listener). Of course common sense would lead to the same conclusion.
If the sound of direct radiator speakers seems shrill, then pad down the treble output of the front tweeter, place sound absorbers at side wall reflection points and listen off-axis if there's still too much treble -- there's no need for rear tweeters firing at walls. They create early reflections off the nearby walls that can never dupicate the ambience (reverberation) in a large venue.
.
.
.
Richard BassNut Greene
My Stereo is MUCH BETTER than Your Stereo
"A mono center speaker is better for a center vocalist.
This was discovered in the 1930's when Bell Labs decided three speakers were the minimum for a stereo effect (left, center and right) but later no one could figure out how to put three channels on vinyl, so we got a two channel compromise instead."Is FR the reason for this conclusion or the ability to fix the location of instruments more solidly in space? The 1930s were a time when recordings rarely had sound beyond 5 khz. Stereophonic tape preceded stereophonic vinyl records by several years and 1/4 track was standard. Had 3 channel sound been significantly superior to 2 channel sound, it could have been an option although it would have cut recording time in half. Another option was 3 channel amplifiers with a derived center channel. Equipment manufacturers would have loved it because they could have sold 50% more speakers and amplfier channels but it never came close to being a standard (my HK A500 amp had a center channel line level output with a level control on the front panel.) So are you saying that as instruments are relocated progressively from extreme left or right where they are reproduced by only one speaker gradually to stage center where they are reproduced by both speakers, in a two channel sound system their spectral balance changes? Wouldn't this be obvious by switching the preamp to the monophonic mode and then turning the balance control to one extreme or the other? You should hear a difference in FR if your explanation is correct. I'd like to know if anyone tries it what conclusions they come to.
"If the sound of direct radiator speakers seems shrill, then pad down the treble output of the front tweeter, place sound absorbers at side wall reflection points and listen off-axis if there's still too much treble"
I tried exactly what you said for years and I couldn't get my speakers (AR9) to sound accurate to my satisfaction. The tweeters could be padded down -3db and -6db and so could the upper midrange. My 10 band equalizer could create a 15 db cut and I had bass and treble controls and filters on my Marantz 3800 preamp which could provide substantial additional cuts. Nothing worked to create a sound which was both tonally balanced where the harmonics seemed correct for a particular insturment and where the sound wasn't somehow unnaturally shrill...that is until I changed the hf radiating pattern as I described. Even then, it took a very long time to get it right.
"there's no need for rear tweeters firing at walls. They create early reflections off the nearby walls that can never dupicate the ambience (reverberation) in a large venue."
Well at least the very last part of your statement is something I can completely agree with. There is no way to duplicate the reverberation of a concert hall by altering your speaker systems. That's orders of magnitude more complicated.
Is it more difficulte to implement ? And we have talked about HF but what about the midrange ? Do you think a typical woofer gives already enough dispersion ? Thanks for your very informative posts.
If you think so, then Bose 901's are the speakers for you!
.
.
.
.
Richard BassNut Greene
My Stereo is MUCH BETTER than Your Stereo
.
...nothing turns me off more than a poor treble presentation. I could stand the absence of low bass and could even tolerate a recessed or somewhat colored midrange. But if the treble is screechy or dull, I notice it immediately and I either run screaming from the room or lose interest within seconds.
As I posted elsewhere, if the treble isn't right, nothing else works, it doesn't even seem to matter. And while a speaker may produce deep bass you can feel like a pipe organ pedal note, without accurate treble, it still won't sound like a pipe organ. Not when the organist hits other notes. I've experimented and struggled with this problem for a very long time especially the way recordings of violins are reproduced. A lot of people who post here seem very glib about it mostly repeating what they've read by others who haven't really experimented with it either. Perhaps the proof that the problem is so important is the fact that there are more types of tweeters than all other types of loudspeaker drivers combined.
The problem at least at high frequencies is that the room boundaries absorb sound selectively as a function of frequency. In general, the higher the frequency, the greater the absorption. While an omnidirectional loudspeaker array would be a definite improvement over the conventional single tweeter box speaker or even a speaker wtih just one rear firing tweeter, it won't provide the ideal result of both flat FR from direct radiation and from indirect reflections meaning flat total power response. To achieve that, the sound directed at the room boundaries must be independently adjustable so that they generate more energy at frequencies where sound is absorbed most. That means that the indirect tweeters would likely have a rising output as frequency increases. A simple way to achieve this is to place a capacitor across the front firing tweeter which will roll its HF off and then boost the treble of the entire array with an equalizer. The front tweeter will flatten out and the indirect tweeters get the boost. Aiming the indirect tweeters at the ceiling should also be tried. This is the arrangement I have for my main system, three 3/8 poly's crossed over at 6khz in parallel (91 db sensitivity) and a single 3/4" front firing tweeter set to -6db relative to 89 db sensitivity. This achieves around the 12:1 ratio I found worked best in my somewhat on the live side rooms.Most woofers are nealy omnidirectional. It's very difficult to restrict the dispersion of the bass because the wavelength is very large compared to the diameter of the woofer cone. My AR9s have a dipole arrangement, 2 side firing 12" acoustic suspension woofers. As for the midrange, the dispersion at those frequencies will substantially also affect the perceived sound. Bose 901 is an extreme case. The sound is interesting, different, but living with a modified version now, I can't say which is actually better. I'm considering designing a speaker system where all of these parameters are adjustable.
.
*
"Whoever undertakes to set himself up as a judge of truth and knowledge is shipwrecked by the laughter of the gods." - Albert Einstein
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
