Ok first here is direct link to the comparison as it'll be easier to refer to it:
YouTube
I agree with you about that but in this case i don't think this is what happen.
For the first song we agree, the sm80 seems better on a lot of parameters, more 'direct' sounding, frequency response seems closer to the original song, there is too a strong fantom image which seems to be missing from the Q7 and the frequency response of D & B seems not true.
In the second example here again i think we agree there is a difference between the two but it is much less obvious than in the first example.
The third example i wasn't really sure if there was a difference between both set and to be honest i didn't really understood what was the point.
At first as i said i was really astonished to ear more or less the same thing that what i experience in my room with my own loudspeakers so i thoughts it was a typical case of comparaison of a point source and a spaced multidriver loudspeaker and the way they interact with the room.
Then i thoughts that i already played on (as a dj) and used ( for PA as engineer) products from D&B and they were really, REALLY good sounding system ( circa 2003/2005 and not a line array kind).
So i was worried about the brand having evolved in a bad way. Then i decided to listen to the example again and this time i expanded the picture on my screen and spent a little time looking at it.
Then it striked me: both set of speaker are located relatively high on pole and listening position ( desk location) is relatively away from the loudspeakers located way off axis on horizontal plane and vertical plane. This is not the way i set up things but i've already seen this kind of setup used.
From my point of view if the loudspeakers are high pass filtered to high ( beyond 80 hz) it may create an issue with the sub hence i dislike this kind of setup. But it is a situation you may encounter in case of PA.
As the records are binaural i thought the differences were from the way a coax interact with room as it spread a coherent and smooth power response into the room and i suposed the Q7 was a typical MT setup with the typical destructive interference pattern.
I checked the Q7 datasheet to be sure about my point and discovered these are not MT but some variation on the MTM theme.
MTM are a way to have a kind of point source behavior so the issue came from something else. Then i took a look at polar plot and here is the answer i think:
Sm80 are coax nested into a waveguide so they offer a 80* conical drirectivity.
The Q7 are 75* horizontal plane and 40* vertical plane. This is a somewhat meaningful diference and looking again at the picture you can see both set of speakers point toward the rear wall without any angle ( tilt) toward the listening position.
Not really an issue with sm80 ( they'll loose a bit of high freq but otherwise the tonality won't change this much off axis) it can clearly be with the Q7 as they cannot garantee a consistent sound when used without angle toward listening position.
But this is not the only thing at play in the first example:
The q7 is a variation around the MTM principle and in vertical plane this kind of config may have inconsistent behavior with lobing and variation of covering angle with freq with narrow/wide variation on relatively small frequency area.
And if you look at Q7 datasheet for polar map this is exactly what we see p6 in fig 6:
Q7 loudspeaker | d&b audiotechnik
There is a 'hole' in the vertical coverage between 300hz and 500hz approximately then a 'rise' of freq between 550 and 750hz.
So here is what i think happen in the first example: the Q7 isn't tilted toward listener as it is expected to be in use AND the listening position is probably located at an angle which correspond to the 'hole'/'rise' in the vertical coverage.
One of the two issue only might have been acceptable but both at the same time is too much hence the weird frequency response.
In the second example you find the sm80 'shooty' but in fact i think this is the Q7 which is not enough mid rangy focused. The fact that there is no voices involved in the track makes it less obvious than in the first example.
Lastly the last examples doesn't exhibit as much difference as the first ones. I initially concluded this was because the open space then i spoted the mic which this time was located in the vertical 'window' of correct use for both speakers...
So here i think the difference is all about vertical coverage of both speakers and listening position.
This is not really what i expected when i first listened to the example and thought the difference came from reflected sound ( early reflection) being not 'coherent' enough with the Q7 versus the one from SM80 and coloring the resulting sound at listening position ( which might play a role too but not as important as i thought in this case).
What do you think about it? Makes sense to you?
In the end i find the comparaison unfair to Q7 as they are not used the way they should and this may be the reason of difference.
Last thing i don't explain is the fantom image being not present in the q7 first example. As you stated earlier it may be a Freq range issue but i'm not sure...
Last edited:
Thank you, that sounds like a good setup. inspired by this thread i reinstalled a pair of sb16-coax drivers i had on shelf in some test cabs and trying to work some crossovers for them, these drivers are really budget and used tweeter and middriver are not a very good match in my book so it is a bit tricky to get a system response from them that i like. i really would like to try seas and some eminence 10" coaxs out.
you write that the your tannoys has recessed high freq when used in the far field, i have the same problem in my listening room, when listening in the near field i can hear deep into the recording and the tonal balance is just right, but as soon i listen further away the tonal balance gets thick and somewhat dull, i have tried many things to compansate and there is just no good cure other then go back and listening in the near field. i guess this must be an room issue, my room has quite a lot of absorbtion, a hand clap sounds very dry and almost choked, but i have still some room resonances that bothers. how is your room treated, much of absorbtion? i wonder how amar bose 901 could sound like, with lots of reflections with good tonal balance might work, or not
you write that the your tannoys has recessed high freq when used in the far field, i have the same problem in my listening room, when listening in the near field i can hear deep into the recording and the tonal balance is just right, but as soon i listen further away the tonal balance gets thick and somewhat dull, i have tried many things to compansate and there is just no good cure other then go back and listening in the near field. i guess this must be an room issue, my room has quite a lot of absorbtion, a hand clap sounds very dry and almost choked, but i have still some room resonances that bothers. how is your room treated, much of absorbtion? i wonder how amar bose 901 could sound like, with lots of reflections with good tonal balance might work, or not
looking nice, i see you now using a non coax setup, may i ask why you did the change?
krivium said:
sure, it makes perfect sense. i have no experience with any of those speakers, i have almost none experience working with pa, so i can not tell if the recording is represantative for each speaker. but is it not strange if the dude who maked the video did not take care of any lobing issues if he wanted to do an accurate comparsion? or maybe i misunderstood the video and he just wanted to highlight the differencies?
The focus for that project was the woofer, which had accelerometer feedback. The original speakers were smaller, tweeter mounted in front of the woofer with a 2" dome midrange on top:
Internet sourced picture
The questions "how to make the feedback system work in a larger internal volume box (with similar outer dimensions)" and "can imaging be improved" were what started this project.
The coaxials were used because they could be crossed over lower and the integrated tweeter made it so there was no need for a seperate tweeter in front of the woofer, or the speaker to be any bigger.
The imaging was improved and the feedback system was adjusted with succes so the questions were answered. I just wasn't particularly fond of these Kef units so that's why the boxes were put away eventually. There isn't a particular reason why I wouldn't try another coax unit though. I would like to hear those SB ones one day.
i wonder how amar bose 901 could sound like, with lots of reflections with good tonal balance might work, or not
Hi Celef, i would like to have a pair of each possible kind ( design philosophy) of loudspeakers in my room!
Ime this is the only way to really know what suit your taste in your room. But not really convenient if you live with other human being...
I agree lobing free and constant smooth power response is addictive. Doesn't make it the 'silver bullet' ( there is always compromise) or 'cure it all' answer but if you like the presentation this is addictive.
Have you changed the width of baffle from your previous experiment with the driver? If yes and if you remember the outcome i bet they are less 'pin point' regarding stereo with your wide baffle. Even less if you could use a large radius roundover instead of edge.
Ps: have you seen my pm?
yes these baffles are wider then before and the edges are beveled. last time i tried these drivers out i was not convinced, i could not get the drivers and the xo to work as i wanted. that is still true to some degree but now with a slightly different goal in mind the system zum is quite nice. i need more listening but some tracks sounds just amazing.
sorry no, i have not recieved any pm.
sorry no, i have not recieved any pm.
I mean the speakers not being equidistant from the listening position, mainly
Ah, ok.
That was weird when he talked about the right speaker being virtually hinged on the left one.
I feel like when the crossover is all wrong, moving the speaker in any direction or angle wouldn't nescessarily improve it.
My old neighbour had a set of B&W's with yellow cone midranges. First time I heard them, I asked him if they were wired out of phase. He checked, but said they were ok. Few years later I saw a frequency response chart, and wonder if the large dip in the amplitude response could have been responsible for the overly spatial sound they had. They sounded very 'deep' but not realisticly sized.
Wonder if that could even be possible if both sides were identically filtered, or that he had been fiddling with (one of) them.
That was weird when he talked about the right speaker being virtually hinged on the left one.
I feel like when the crossover is all wrong, moving the speaker in any direction or angle wouldn't nescessarily improve it.
My old neighbour had a set of B&W's with yellow cone midranges. First time I heard them, I asked him if they were wired out of phase. He checked, but said they were ok. Few years later I saw a frequency response chart, and wonder if the large dip in the amplitude response could have been responsible for the overly spatial sound they had. They sounded very 'deep' but not realisticly sized.
Wonder if that could even be possible if both sides were identically filtered, or that he had been fiddling with (one of) them.
Hmm I would think that if you adjust the setup for low frequency, you could be way off in the mids because of the difference in wavelength/group delay step up towards the low end. The units are all in one box so it's difficult to move one without the other....
Something else I can't bend my head around. What about c-t-c spacing. Rule of thumb asks for a certain maximum distance at crossover frequency, but does that mean the whole thing just collapses above this frequency? I haven't seen this rule of thumb specify a type of slope... I would think that above the crossover frequency (when the rule of thumb has been applied minitiously) there would be a power response irragularity especially in the vertical plane? Like the 'crossover region', the area between the driver centers itself is beaming?
Something else I can't bend my head around. What about c-t-c spacing. Rule of thumb asks for a certain maximum distance at crossover frequency, but does that mean the whole thing just collapses above this frequency? I haven't seen this rule of thumb specify a type of slope... I would think that above the crossover frequency (when the rule of thumb has been applied minitiously) there would be a power response irragularity especially in the vertical plane? Like the 'crossover region', the area between the driver centers itself is beaming?
I haven't tried the speaker setup method, but few things cought my attention:
1. obviously, speaker positioning has to be right to get most of the speakers (be it any setup method).
2. One aspect of great speaker setup is the depth of field (how ever one perceives it)
3. optimal speaker positioning depends on your speakers and the room
4. the method is done by listening! You move speakers, and you like the results! or not.
Anyway, here are my thoughts about the method.
Initial step, listening the bass: what happens in the method, is you listen to a certain song that has these repeating bass notes. Pull out one speaker from the front wall, little by little, and find position(s) where the bass notes seem equally loud at the listening position. This is to change the SBIR in regards the front wall and room modes in general. Now, change a song which is in another key and the bass notes won't probably be even loud anymore But, this is not the goal I think. When you get The song bass notes even, your brain has easier time during rest of the speaker setup, which is done by using the same song! Now, foundation stays still, easier to listen what else is happening in the perceived soundfield. Hence this is done first. The bass is the contrast you then listen the singer against from now on.
The next steps in the method are to get the phantom center to the center of the listening position and to get the focus to the singer to be right (not too wide, not too small, realistic height, just how you like it).
To get the phantom center to be center the listeners hearing condition and the room will affect the speaker positioning and thus the setup ends up being asymmetric by default? At least my ears have very different condition, and have no dedicated listening room. I could imagine brain is doing some balancing / compensation between trying to figure out who is singing, and sometimes it has to work harder. I've been thinking recently, that all kinds of issues with the playback that give harder time for the brain to interpret the sound will give some sort of listening fatigue and take away from the perception.
Anyway, a good setup might be dictated by room layout, a spouse, limited floor space or any other aspect than the actual sound and thats ok What ever helps you enjoy the music.
The most important thing is to realize you have to use the ear as ultimate measure of the system performance (tape measure is more accurate than ear, but is used to measure distance not qualities of sound). Be the end result what ever, it is judged by the listener to be the optimal, and that is the key.
Lack depth? do it again, or use some other speaker positioning method or revert to the tried and true old speaker position. If still lack of depth it is the music, the speaker or the room that doesn't let the depth happen, or conflict between possibilities of stereo playback and the expectation of it.
1. obviously, speaker positioning has to be right to get most of the speakers (be it any setup method).
2. One aspect of great speaker setup is the depth of field (how ever one perceives it)
3. optimal speaker positioning depends on your speakers and the room
4. the method is done by listening! You move speakers, and you like the results! or not.
Anyway, here are my thoughts about the method.
Initial step, listening the bass: what happens in the method, is you listen to a certain song that has these repeating bass notes. Pull out one speaker from the front wall, little by little, and find position(s) where the bass notes seem equally loud at the listening position. This is to change the SBIR in regards the front wall and room modes in general. Now, change a song which is in another key and the bass notes won't probably be even loud anymore
But, this is not the goal I think. When you get The song bass notes even, your brain has easier time during rest of the speaker setup, which is done by using the same song! Now, foundation stays still, easier to listen what else is happening in the perceived soundfield. Hence this is done first. The bass is the contrast you then listen the singer against from now on. The next steps in the method are to get the phantom center to the center of the listening position and to get the focus to the singer to be right (not too wide, not too small, realistic height, just how you like it).
To get the phantom center to be center the listeners hearing condition and the room will affect the speaker positioning and thus the setup ends up being asymmetric by default? At least my ears have very different condition, and have no dedicated listening room. I could imagine brain is doing some balancing / compensation between trying to figure out who is singing, and sometimes it has to work harder. I've been thinking recently, that all kinds of issues with the playback that give harder time for the brain to interpret the sound will give some sort of listening fatigue and take away from the perception.
Anyway, a good setup might be dictated by room layout, a spouse, limited floor space or any other aspect than the actual sound and thats ok
What ever helps you enjoy the music.The most important thing is to realize you have to use the ear as ultimate measure of the system performance (tape measure is more accurate than ear, but is used to measure distance not qualities of sound). Be the end result what ever, it is judged by the listener to be the optimal, and that is the key.
Lack depth? do it again, or use some other speaker positioning method or revert to the tried and true old speaker position. If still lack of depth it is the music, the speaker or the room that doesn't let the depth happen, or conflict between possibilities of stereo playback and the expectation of it.
Last edited:
This is not a rule of thumb: the 1/4 wavelength is the limit for ACOUSTIC coupling of two source. This mean the two source behave like one from an acoustic point ( both wavefronts merge into one coherent).
The whole thing doesn't collapse abruptly once you are above (it is progressive) and you can have acceptable results up to 1/3 wavelength. Once you go beyond that the vertical coverage start to narrow up to a 39* coverage at one wavelength ctc.
The type of slope will have an effect on lobing, the steeper the less destructive interference pattern between drivers hence lobing.
Yes there will be power response anomaly in the vertical plane: take a look at fig6 of Q7 datasheet previously linked this is what to expect from an mtm with not to steep xover slope.
Robbintip,
Have you heard about Horbach-Keele filter method? It is based on vertically spaced drivers and the use of FIR filter particular profile to control vertical directivity.
It is spot on your questions.
You can download the papers here:
AES Papers -- Official website of D.B.Keele
Scroll down to #42 and download both papers.
Fig3 in part 2 will give you an idea of behavior of spaced source relative to wavelength. For info in d'apolitto mtm the spacing is 1/3 wavelength. You can see why in fig3, this is the limit for quasi omni radiating pattern.
Have you heard about Horbach-Keele filter method? It is based on vertically spaced drivers and the use of FIR filter particular profile to control vertical directivity.
It is spot on your questions.
You can download the papers here:
AES Papers -- Official website of D.B.Keele
Scroll down to #42 and download both papers.
Fig3 in part 2 will give you an idea of behavior of spaced source relative to wavelength. For info in d'apolitto mtm the spacing is 1/3 wavelength. You can see why in fig3, this is the limit for quasi omni radiating pattern.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.