What you showed must also apply to the crossover range of a speaker. It has opened my eyes and it also shows why Tannoy and Kef bother with the dual concentric idea plus why many woofers and tweeters have a flat side to enable the closest possible placement.
Good video and I can see what Danny is saying here about why multiple tweeters may not necessarily work well for additional output. One thing that Paul McGowen of PS Audio said, and his old friend Arnie Nudell did with the Infinity IRS IV speakers is use multiple tweeters in a line array and did the same with the midrange drivers and woofers so no matter whether you are sitting or standing, the sound won't change, but if you were to raise the volume, multiple tweeters can mean they have to work less hard to produce the treble at higher volumes, so they never get shouty like some speakers so when you crank up the volume so as you crank the volume, the tweeters in a line array setup never get hard sounding and thus fatiguing at higher volumes. At least that's what Paul says anyway. Another thing I think also causes some listener fatigue is using one overly bright tweeters or too small a tweeter for a given design so they have to work harder to produce volume at higher levels., but metal domes tend to be bright compared to soft domes say. But what do I know, I don't design speakers but do use some vintage ADS L810 large bookshelf speakers that use a 3/4" soft dome tweeter, a 2" soft dome midrange driver and 2 8" woofers in an acoustic suspended speaker cabinet that are the best I've had in my system and probably made some of the biggest improvements in all the other parts, outside of the cartridge and perhaps the phono stage.
I almost replied with "...but what about Tekton Designs" and then I remembered that they have one tweeter in their tweeter array that's playing frequencies above the 1/4 or 1/2 wavelength to keep the comb filtering at bay. It's a very cool design using tweeters with super low Fs as an low-mass array.
I have 7 speakers in my home theater (other than subwoofers). All of them have tweeters. How come those tweeters are not significantly cancelling out each other. I can only hear a subtle difference by changing my position. I have also used software to see my freq. response at different seating positions. I am just curious that how having tweeters in totally different locations and obviously not synced exactly in time are so forgiving. I understand that at any one spot I may be listening to just one speaker dominant in that area but I have tried by turning on and off different speakers and playing around with their positioning. Any help would be appreciated.
Great video, just like all the others. My question though. Your test setup places the tweeters approx 7" to 8" apart. If the tweeters were smaller, like Tymphany D26NC56-06 or Dayton Audio ND20FA-6 for instance and placed touching, one mounted above the other on the same baffle. Would the combing effect be nearly zero as the center distance is only 1.5" to 2" ? I'm considering this as I need approx 95 DB to 96 DB @1W/1M. I'm not too keen on waveguides as I would like good polar dispersion if possible. Many thanks. Richard UK.
You don't even need fancy test equipment to hear the effects. Use white noise, pink noise, or hiss, which you can find doing a search on UA-cam. Play it through two tweeters, or full range drivers, and move your head up and down or side to side.
I have the Sony SSCS towers that I bought for rear surrounds. They have a "super tweeter". Though I got them for rear surrounds , of course I've listened to them in two channel. I didn't notice any weird stuff going on in the highs. They lack bass but are not overly bright. To be honest, they don't really do anything that I noticed. I wonder if I disconnected them if I would notice a change.
Greshman Acoustics in Canada did exactly this. They took couple of their book shelf speaker (Grand Studio) and put them on top of each other, selling it as their Grand Studio II model. Apparently they adjusted the bottom tweeter’s phase to resolve the issues that’s been discussed here. Any thoughts???
I'd imagine if you do dual tweeters with ones that have no baffle of their own and small neo magnets, and inset both into the baffle in a custom waveguide with the domes very close to one another, this would be minimized
@@mattseaton5832 Its true, but any interference wiggles will be small enough that in a normal listening environment at normal listening distances in a typical room, they will nearly vanish with respect to the much larger room issues. The only time this might be an issue IMO is nearfield. Just my two cents
Looking forward to seeing some other manufacturers in this series! It’s a cool alternative to the PS Audio’s “Ask Paul” series, but it would be great if NRD was able to feature some other companies and gear.
Line arrays are very different. They act more like a single continuous driver. If you pulled out all of the tweeters except for two of them then the response will be just like what I showed in the video. But while two of them may be causing a cancellation others that are playing are not. Their distances are different. What you get is +/-2db ripples that move around as height changes. Dome tweeters or drivers that are spaced too far apart still have a mess of a response though.
It's all about understanding the tradeoffs. Line arrays do truly amazing and beneficial things in the realm if acoustical physics that are literally impossible for a single point-source tweeter.
Line array is different because it is not a multiple point source, but a continuous seamless panel, or line, but could be sensitive to vertical position in theory. Some professional JBL line arrays are curved to where the top and bottom of the array is the same distance to the general target listening position. Someone should test one.
Tried to comment before but it didn't take. I'm a little late to this post. First off, thanks Danny and Ron for your contributions! I have a pair of Rectilinear IIIs (the tall ones) from, I think, '69 or '70. The speaker cabinet is 18" wide by 35" tall by 12" deep and has a 12" woofer, a 5" whizzer cone midrange, two 2 1/2" tweeters and two 2" super tweeters. All four tweeters on these speakers are spaced apart: one tweeter is just above the woofer, almost vertically center-baffle and horizontally off to the side, the other is higher, next to the midrange, about 9" from the top of the baffle, on the opposite side. The two super tweeters are at opposite corners, top and bottom. Watching this video has me concerned about this configuration. I’ve not had these measured in this 1-watt/1-meter way and wouldn’t have the equipment or know-how to do so, but I’d be curious to see how they would measure. I do like the way these speakers sound, very natural, wide soundstage, good imaging, good bass down to maybe 40hz. As I try to understand comb filtering and determine whether or not I can hear this with the Rec IIIs, I am curious to get your take on the Rectilinear configuration and whether or not you’re familiar with these speakers.
Well, without question if one is delayed in time just slightly compared to the other then they will directly cancel each other out and cause some holes in the response.
The differences between the two speakers probably has to do partially with the geometry of the stand, diffraction around the differing geometries, and the ceiling vs floor relative to the tweeters and woofers.
Question: It makes a lot of sense I am just wondering, you are using 16 neo3 tweeters in your line source, and it clearly works? Can you use 2 x ne03, 3 x, 4 x or is there some math to it?
Danny, will the comb filtering effect come into play if the tweeters are next to each other on the horizontal plan instead of the vertical plan? Thanks, Rick
Interesting. Does this apply to bipole speakers with tweeters facing different directions? I have a pair of Def Tech SR9080's & I noticed when I put my receiver in "all stereo" mode where every speaker is playing the same thing, the sound becomes muffled like I'm losing my hearing as a I move my head slightly back & forth.
Does this "comb filtering" issue also apply to car audio as i have seen many people install multiple tweeters and mid-range speakers on just one door panel.. I myself have never done it , but I'm just curious because I've have heard or read something that was lightly discussed on that subject on a forum, i believe, but never got to deep into researching on why it wasn't a good thing ....lmk . Thanks, Jose
Dynaudio puts two tweeters on many of their top line speakers. And Tekton puts even seven or more. But as I understand only one of those tweeters runs at an upper end of a tweeter range.
@@garygranato9164 They are not wired together but via different crossover filters. And for Tekton 6-tweeter array that could be series-parallel connection.
On my Tektons, the 6 tweeters of the outer ring are wired pairs in parallel and those pairs are wired to each other out of phase. the wire from crossover is wired to the positive of the first tweeter of the first pair. The negative of the last tweeter of the last pair returns to the crossover
That's a very good demo of the phenomenon of interference, big thanks...I have a mono guitar amp that has two piezo tweeters 12" apart, I'd like to replace them with some better dynamic tweeters, would you suggest to use only one tweeter instead of two?
Love to see this methodology applied to the Tekton Impacts. In theory, the nearfield response should have dips all over the place. Room placement may flatten them out somewhat, but it's going to be a hit or miss proposition.
@@Newrecordday2013 Interesting. I am only mildly surprised,. Empirical evidence should always trump theory. But if those are near-field measurements from multiple spots, I wonder how Tekton dealt with it in the design. Then, of course, there are the real-world in-room issues that can swamp out any near-field observations. Your own sound demo recordings show marked differences between near-field and in-room results. Keep up the good work. It's thought-provoking.
@@Newrecordday2013 Exactly, because there are exceptions to almost every "rule" in loudspeaker engineering and audio in general. This is why these blanket statements shouldn't be propagated without extreme caution. For almost every blanket statement one can make about speaker engineering, an exception can be found.
What I am illustrating and showing here doesn't really apply to the Tekton designs. They use only a single tweeter to cover the whole upper range. The rest of the tweeters are actually playing upper mid-range areas only.
A tall ribbon tweeter (about 8” tall) has a very similar response when tested in the same way.... of course a 6’ one has less if near the middle and on axis - but it is still there
Why doesn't this cancellation propagate from left channel tweeter and the right channel tweeter? Why do they have to be "on the same baffle" for this cancellation to occur?
Great presentation. YOu could alo look at it thinking that the wavelength of a 20khz sound if about 1.7cm, tiny compated to a 17m wavelength for a 20hz signal. What would you get if you had 3 tweeters, th one firing forward and two aimed off to the sides to widen the listening area?
My only experience of a 2-tweeter speaker is the Tandberg Studio Monitor which are great sounding speakers.. I wonder if I simply wasn't hearing it (it was a while ago) or if Tandberg did something interesting. Anyone know? They were used as studio monitors in a good few places.
Very interesting and informative. What audio software do you use? I am not really fond of the software I am currently using. Do you think most standard laptop audio cards are accurate enough to get good results when measuring and testing? Or should I use something like a USB preamp from Behringer or Presonus etc? I probably need phantom power for the reference mic. Can you give some advice on the best way to set up for measuring?
I have been a speaker builder for many years and have tried many materials and volumes of stuffing in bass ansd midrange enclosers whats your thought gareth in the uk
So the big question is why doesn't this effect the woofer? There are tons of speakers that use several woofers in a tower arrangement. also, Back in the 70's there was a company that used two phenolic ring tweeters mounted at two different angles to expand the soundstage of their speakers. They were made by Studio Lab and the model number was SL 110 and they were fairly successful at the time. They sounded great for using phenolic ring tweeters, which by today's standards aren't that good. Ny ear I could hear any loss of signal throughout the sound stage and seems to have a very wide sweet spot. You can find pictures of them on google images so you can see what I'm talking about. The question is then, how did they get away with using two tweeters? Was it because they were facing different direction?
So Danny, for a mono playback on a modern hifi system (including modern stereo cut mono albums) should we be killing one speaker instead of either letting it play in “stereo” or summing the channels with a mono switch? I made a jumper box where I can kill either L or R, or jump LR, but that’s the pre power amp. I feel like the better solution is having a dedicated mono power amp with a mono speaker in the center of the room?
Your speakers are probably 6-8 ft. apart which could affect the midrange output but not the treble. If you can get a tight center image from a mono recording, you have probably minimized the comb filtering.
Whether you drop out a channel or not could depend on the speaker placement and distance that they are from each other compared to you. And the number of recordings made that way are limited. Go with what works best (or sounds best) for you and your application.
You’re only measuring at one location. There will be both constructive and destructive interference throughout the room. You are showing the destructive interference case.
Check the wavelength at 3kHz. www.soundoctor.com/freq.htm So when you cross your mids to the tweeter you'll want the acoustic centers of the mid to the tweeter to be less than the wavelength at the crossover point. That's a good rule of thumb.
@@dannyrichie9743 isn't it usually a good idea to have the center of the mids and highs as close as physically possible anyway? Or is there an optimal distance depending on the crossover point?
@@dakken74 Yes, the closer they are together the easier it is for them to blend, and it will minimize cancellation patterns in the off axis response. See the rule of thumb posted above your post. An example would be if the acoustic centers are 6.5" apart then the crossing over point needs to be less than 2kHz. If the acoustic centers are 5" apart then they need to cross below 2.6kHz.
I currently have front speakers set up in this exact configuration. I stacked them for the exact reason Danny mentions. I had no idea that I would introduce this issue. Luckily, these speakers were part of a 5-speaker set for my home theater, all of which are matching except that the front L and R speakers are bi-amped. I will experiment with turning the bi-amped tweeter off to see if I can hear the difference. I did have the mid-point of the tweeters at ear level to set both tweeters the same distance to my ear. Danny, I use software called ScreenCast-O-Matic (screencast-o-matic.com) to make training videos for software that I write. It captures my computer screen while I operate the software. I like that you show the software in this video and would like to see more of that.
They will absolutely cause the same things seen in the video. And be sure to tune in next week when I show what happens when a super tweeter is added to a full range or wide band driver.
SPC design also, being a dipole configuration, you won’t get the sort of comb filtering that using two tweeters in the same polarity on same face of the baffle brings.
@@RennieAsh In all fairness measurements are the only way to prove whether your right or wrong. You can turn any smart phone into a very accurate microphone for measuring here is the link. www.parts-express.com/dayton-audio-imm-6-calibrated-measurement-microphone-for-tablets-iphone-ipad-and-android--390-810
What about something like JBL’s VRX932LA, which has 3 tweeters mounted in each box on a mild curve in a line array. The tweeters receive -3, 0, 3 db signal difference. Looking at their literature it seems a single speaker box has a lot of high frequency roll off in the vertical bandwidth but when you mount 3 in an array suddenly the curve above 1khz is fairy flat. Can you explain how they are achieving this? www.jblpro.com/ProductAttachments/VRX932LA-1_FINAL.PDF Also how would this affect the horizontal as they do not show a plot with multiple speakers
Wow, that's pretty enlightening. Having never really given any thought to how they would interact with each other, it's pretty dramatic. I do have a question that's not necessarily related to this situation. As we get older, (I'm darkening 60's door) our (read my) ability to hear higher frequencies is dropping off. If I'm looking at a frequency response and it's not looking good let's say beyond the 10k-12k Hz, do I really need to be concerned about it?
That's what I was thinking but one of Tectons videos reveals only the centre driver is being used as a tweeter the rest are being used as midrange and at a distance and upper frequency where comb filtering shouldn't occur.
Yep comb filtering sure looks bad. Long video kept saying show you & you'll see. Does he actually think he's the only one who knows that comb filtering looks bad? How many people can hear comb filtering? Some of the greatest & most knowledgeable loudspeaker designers of all time put multiple tweeters on their loudspeakers. It works. Has this guy never heard of line array?
Certainly I am not the only one that knows about or understands comb filtering, but it is a subject that a lot of hobbyist don't understand very well. How many people can hear it? I think everyone can hear it especially when it causes large holes in the output. Have I ever heard a line array? :-) Maybe you'd like to drop in on our Facebook page and scroll through some of the pictures of speakers that I have designed: facebook.com/pg/grresearch/photos/?ref=page_internal
@Nat N Those LS-6's and LS-9's were great speakers. I designed the drivers for those and we only had several thousand made. Once they were all sold out they were gone. I didn't have any more made. I do however offer some kits now that exceed their performance in multiple areas. So if you still want to build a kit that takes you to the top level of design and performance then I can get you taken care of.
@Nat N Check this kit out: gr-research.com/nx-treme.aspx This one uses a planar magnetic tweeter in a waveguide with a low crossover point. The mids are better than the older LS series designs. The baffle is smaller. Imaging is a lot better. It still uses multiple woofers to handle the lower end (very fast) and is full open baffle from top to bottom. And the drivers are physically and acoustically aligned on both sides of the speaker (front and back). They are incredible.
Seriously, go listen to a Stirling Broadcast LS3/6 and claim it's a bad sounding speaker with a straight face. But I suppose designers the likes of Derek Hughes have no clue what they're doing. 😂
If both tweeters are both covering the same range then what is illustrated in this video will be present without question. But if you like the result then happy listening.
Absolutely fabulous video Danny. Comb filtering explained in layman's terms!
What you showed must also apply to the crossover range of a speaker.
It has opened my eyes and it also shows why Tannoy and Kef bother with the dual concentric idea plus why many woofers and tweeters have a flat side to enable the closest possible placement.
Good video and I can see what Danny is saying here about why multiple tweeters may not necessarily work well for additional output.
One thing that Paul McGowen of PS Audio said, and his old friend Arnie Nudell did with the Infinity IRS IV speakers is use multiple tweeters in a line array and did the same with the midrange drivers and woofers so no matter whether you are sitting or standing, the sound won't change, but if you were to raise the volume, multiple tweeters can mean they have to work less hard to produce the treble at higher volumes, so they never get shouty like some speakers so when you crank up the volume so as you crank the volume, the tweeters in a line array setup never get hard sounding and thus fatiguing at higher volumes.
At least that's what Paul says anyway.
Another thing I think also causes some listener fatigue is using one overly bright tweeters or too small a tweeter for a given design so they have to work harder to produce volume at higher levels., but metal domes tend to be bright compared to soft domes say.
But what do I know, I don't design speakers but do use some vintage ADS L810 large bookshelf speakers that use a 3/4" soft dome tweeter, a 2" soft dome midrange driver and 2 8" woofers in an acoustic suspended speaker cabinet that are the best I've had in my system and probably made some of the biggest improvements in all the other parts, outside of the cartridge and perhaps the phono stage.
I almost replied with "...but what about Tekton Designs" and then I remembered that they have one tweeter in their tweeter array that's playing frequencies above the 1/4 or 1/2 wavelength to keep the comb filtering at bay. It's a very cool design using tweeters with super low Fs as an low-mass array.
I have 7 speakers in my home theater (other than subwoofers). All of them have tweeters. How come those tweeters are not significantly cancelling out each other. I can only hear a subtle difference by changing my position. I have also used software to see my freq. response at different seating positions. I am just curious that how having tweeters in totally different locations and obviously not synced exactly in time are so forgiving. I understand that at any one spot I may be listening to just one speaker dominant in that area but I have tried by turning on and off different speakers and playing around with their positioning. Any help would be appreciated.
Wow, thank you for explaining this so it is easy to understand. I totally agree with you. One tweeter is better; your measurements proved it.
Great video, just like all the others.
My question though. Your test setup places the tweeters approx 7" to 8" apart. If the tweeters were smaller, like Tymphany D26NC56-06 or Dayton Audio ND20FA-6 for instance and placed touching, one mounted above the other on the same baffle. Would the combing effect be nearly zero as the center distance is only 1.5" to 2" ? I'm considering this as I need approx 95 DB to 96 DB @1W/1M. I'm not too keen on waveguides as I would like good polar dispersion if possible. Many thanks. Richard UK.
you will still get comb filtering but only above 10k hz
You don't even need fancy test equipment to hear the effects. Use white noise, pink noise, or hiss, which you can find doing a search on UA-cam. Play it through two tweeters, or full range drivers, and move your head up and down or side to side.
You are correct.
Dave Rat has a good demo of the effect; ua-cam.com/video/VHjdh-Vka-g/v-deo.html
How about the Tannoy and Townsend add on super tweeters? they are very expensive and form companies who know stuff.....
Make yourself a present this christmas and buy a larger monitor! Love your videos Danny!
How do you explain line source speaker, where you have multiple tweeters and mid ranges?
I have the Sony SSCS towers that I bought for rear surrounds. They have a "super tweeter". Though I got them for rear surrounds , of course I've listened to them in two channel. I didn't notice any weird stuff going on in the highs. They lack bass but are not overly bright. To be honest, they don't really do anything that I noticed. I wonder if I disconnected them if I would notice a change.
Greshman Acoustics in Canada did exactly this. They took couple of their book shelf speaker (Grand Studio) and put them on top of each other, selling it as their Grand Studio II model. Apparently they adjusted the bottom tweeter’s phase to resolve the issues that’s been discussed here. Any thoughts???
they must adjust the place of the listener too . then it works .
The Beatles sounded pretty good, four tweeters.
I'd imagine if you do dual tweeters with ones that have no baffle of their own and small neo magnets, and inset both into the baffle in a custom waveguide with the domes very close to one another, this would be minimized
At 10khz the wavelength is about 3.5 cm, maybe 1.5"
Any separation larger than this will cause these interferences
@@mattseaton5832 Its true, but any interference wiggles will be small enough that in a normal listening environment at normal listening distances in a typical room, they will nearly vanish with respect to the much larger room issues. The only time this might be an issue IMO is nearfield. Just my two cents
Looking forward to seeing some other manufacturers in this series! It’s a cool alternative to the PS Audio’s “Ask Paul” series, but it would be great if NRD was able to feature some other companies and gear.
Great video.. very informative!
I wonder why line arrays are so praised? Do additional drivers cancel out (fill in) the cancellations :)
Line arrays are very different. They act more like a single continuous driver. If you pulled out all of the tweeters except for two of them then the response will be just like what I showed in the video. But while two of them may be causing a cancellation others that are playing are not. Their distances are different. What you get is +/-2db ripples that move around as height changes. Dome tweeters or drivers that are spaced too far apart still have a mess of a response though.
It's all about understanding the tradeoffs. Line arrays do truly amazing and beneficial things in the realm if acoustical physics that are literally impossible for a single point-source tweeter.
Line array is different because it is not a multiple point source, but a continuous seamless panel, or line, but could be sensitive to vertical position in theory. Some professional JBL line arrays are curved to where the top and bottom of the array is the same distance to the general target listening position. Someone should test one.
Tried to comment before but it didn't take. I'm a little late to this post. First off, thanks Danny and Ron for your contributions! I have a pair of Rectilinear IIIs (the tall ones) from, I think, '69 or '70. The speaker cabinet is 18" wide by 35" tall by 12" deep and has a 12" woofer, a 5" whizzer cone midrange, two 2 1/2" tweeters and two 2" super tweeters. All four tweeters on these speakers are spaced apart: one tweeter is just above the woofer, almost vertically center-baffle and horizontally off to the side, the other is higher, next to the midrange, about 9" from the top of the baffle, on the opposite side. The two super tweeters are at opposite corners, top and bottom. Watching this video has me concerned about this configuration. I’ve not had these measured in this 1-watt/1-meter way and wouldn’t have the equipment or know-how to do so, but I’d be curious to see how they would measure. I do like the way these speakers sound, very natural, wide soundstage, good imaging, good bass down to maybe 40hz. As I try to understand comb filtering and determine whether or not I can hear this with the Rec IIIs, I am curious to get your take on the Rectilinear configuration and whether or not you’re familiar with these speakers.
Well, without question if one is delayed in time just slightly compared to the other then they will directly cancel each other out and cause some holes in the response.
The differences between the two speakers probably has to do partially with the geometry of the stand, diffraction around the differing geometries, and the ceiling vs floor relative to the tweeters and woofers.
Some of the greatest speakers in the world use multiple tweeters, such as the Infinity IRS V.
Hey Tom, Line Source is a different animal all together, not the same thing as the examples being given in this video. Thanks for commenting!
Yes line sources are not the same thing as just two tweeters like this.
@@Newrecordday2013 The Von Schweikert Ultra 11 uses two tweeters. I've heard it sounds OK.
Question:
It makes a lot of sense I am just wondering, you are using 16 neo3 tweeters in your line source, and it clearly works?
Can you use 2 x ne03, 3 x, 4 x or is there some math to it?
best combfiltering explained ever! thanks Danny!
Another great video. I would love to see the ungated differences...something I will do myself I think. Keep 'em coming Danny. Cheers.
Danny, will the comb filtering effect come into play if the tweeters are next to each other on the horizontal plan instead of the vertical plan?
Thanks, Rick
Interesting. Does this apply to bipole speakers with tweeters facing different directions? I have a pair of Def Tech SR9080's & I noticed when I put my receiver in "all stereo" mode where every speaker is playing the same thing, the sound becomes muffled like I'm losing my hearing as a I move my head slightly back & forth.
Does this "comb filtering" issue also apply to car audio as i have seen many people install multiple tweeters and mid-range speakers on just one door panel.. I myself have never done it , but I'm just curious because I've have heard or read something that was lightly discussed on that subject on a forum, i believe, but never got to deep into researching on why it wasn't a good thing ....lmk . Thanks, Jose
Thanks you saved me a ton of money.
Dynaudio puts two tweeters on many of their top line speakers. And Tekton puts even seven or more. But as I understand only one of those tweeters runs at an upper end of a tweeter range.
+1. & how are they wired ?? series or parallel ??
@@garygranato9164 They are not wired together but via different crossover filters. And for Tekton 6-tweeter array that could be series-parallel connection.
On my Tektons, the 6 tweeters of the outer ring are wired pairs in parallel and those pairs are wired to each other out of phase. the wire from crossover is wired to the positive of the first tweeter of the first pair. The negative of the last tweeter of the last pair returns to the crossover
That's a very good demo of the phenomenon of interference, big thanks...I have a mono guitar amp that has two piezo tweeters 12" apart, I'd like to replace them with some better dynamic tweeters, would you suggest to use only one tweeter instead of two?
Love to see this methodology applied to the Tekton Impacts. In theory, the nearfield response should have dips all over the place. Room placement may flatten them out somewhat, but it's going to be a hit or miss proposition.
Hi Stan, I have measured the impacts and was unable to find any evidence of comb filtering issues.
@@Newrecordday2013 Interesting. I am only mildly surprised,. Empirical evidence should always trump theory. But if those are near-field measurements from multiple spots, I wonder how Tekton dealt with it in the design. Then, of course, there are the real-world in-room issues that can swamp out any near-field observations. Your own sound demo recordings show marked differences between near-field and in-room results. Keep up the good work. It's thought-provoking.
@@Newrecordday2013
Exactly, because there are exceptions to almost every "rule" in loudspeaker engineering and audio in general. This is why these blanket statements shouldn't be propagated without extreme caution.
For almost every blanket statement one can make about speaker engineering, an exception can be found.
What I am illustrating and showing here doesn't really apply to the Tekton designs. They use only a single tweeter to cover the whole upper range. The rest of the tweeters are actually playing upper mid-range areas only.
@@dannyrichie9743 Thanks for clarifying, Danny. That was not obvious from a simple visual impression.
A tall ribbon tweeter (about 8” tall) has a very similar response when tested in the same way.... of course a 6’ one has less if near the middle and on axis - but it is still there
Why doesn't this cancellation propagate from left channel tweeter and the right channel tweeter? Why do they have to be "on the same baffle" for this cancellation to occur?
It can, but those are typically different signals (stereo). The room reflections are often a much larger issue.
@Nat N No there is still comb filtering but because the distances are greater the comb filtering can take place in lower frequency ranges.
Great video. Thanks guys!
Thank you for sharing this vital knowledge and testresults.that means I am going to use a horn in my future diy speakers ;) have a nice day !
Great presentation. YOu could alo look at it thinking that the wavelength of a 20khz sound if about 1.7cm, tiny compated to a 17m wavelength for a 20hz signal.
What would you get if you had 3 tweeters, th one firing forward and two aimed off to the sides to widen the listening area?
My only experience of a 2-tweeter speaker is the Tandberg Studio Monitor which are great sounding speakers.. I wonder if I simply wasn't hearing it (it was a while ago) or if Tandberg did something interesting. Anyone know? They were used as studio monitors in a good few places.
Gr-research website won't load ..timed out
Very interesting and informative. What audio software do you use? I am not really fond of the software I am currently using. Do you think most standard laptop audio cards are accurate enough to get good results when measuring and testing? Or should I use something like a USB preamp from Behringer or Presonus etc? I probably need phantom power for the reference mic. Can you give some advice on the best way to set up for measuring?
I have been a speaker builder for many years and have tried many materials and volumes of stuffing in bass ansd midrange enclosers whats your thought gareth in the uk
Would have been good to actually hear the results as well as seeing graphs to further your argument/statement.
what about rear mounted tweeters ?
They can be used to good effect to bring out some of those spatial cues that can improve imaging.
@@dannyrichie9743 Thanks for your reply, and your informative videos
Isn't two drivers only gaining 3db as opposed to one, and not 6db?
I accidently added to dual channel speakers in my car and now i have double the tweeters
So the big question is why doesn't this effect the woofer? There are tons of speakers that use several woofers in a tower arrangement.
also, Back in the 70's there was a company that used two phenolic ring tweeters mounted at two different angles to expand the soundstage of their speakers. They were made by Studio Lab and the model number was SL 110 and they were fairly successful at the time. They sounded great for using phenolic ring tweeters, which by today's standards aren't that good. Ny ear I could hear any loss of signal throughout the sound stage and seems to have a very wide sweet spot. You can find pictures of them on google images so you can see what I'm talking about. The question is then, how did they get away with using two tweeters? Was it because they were facing different direction?
Wavelenght. Low hz == very long waves, so the woofers do not have this issue.
Great work. Thank you.
So Danny, for a mono playback on a modern hifi system (including modern stereo cut mono albums) should we be killing one speaker instead of either letting it play in “stereo” or summing the channels with a mono switch? I made a jumper box where I can kill either L or R, or jump LR, but that’s the pre power amp. I feel like the better solution is having a dedicated mono power amp with a mono speaker in the center of the room?
Your speakers are probably 6-8 ft. apart which could affect the midrange output but not the treble. If you can get a tight center image from a mono recording, you have probably minimized the comb filtering.
Whether you drop out a channel or not could depend on the speaker placement and distance that they are from each other compared to you. And the number of recordings made that way are limited. Go with what works best (or sounds best) for you and your application.
in a line source --- u have many identical sources --- where is the comb?
You’re only measuring at one location. There will be both constructive and destructive interference throughout the room. You are showing the destructive interference case.
"Hold my beer" - Tekton
I'd be interested in seeing a binaural or ORTF result at distance in an open environment. Just curious...
Wonderful. You talk about it like a professor working on a Large Hadron Collider about fluctuations in a quantum field :)
I love it!
Nothing like it you clown
c-to-c between the tweater is clearly to big for tweeters.. you can push the drop over 15khz with smaller c-to-c
In the vertical off axis the cancellation will still be fairly extensive and well into the 3 to 4kHz range.
If I use dual 3 inch mids per tower should I be worried about comb filtering? The range will be 200-3000 hertz.
Check the wavelength at 3kHz. www.soundoctor.com/freq.htm So when you cross your mids to the tweeter you'll want the acoustic centers of the mid to the tweeter to be less than the wavelength at the crossover point. That's a good rule of thumb.
@@dannyrichie9743 isn't it usually a good idea to have the center of the mids and highs as close as physically possible anyway? Or is there an optimal distance depending on the crossover point?
@@dakken74 Yes, the closer they are together the easier it is for them to blend, and it will minimize cancellation patterns in the off axis response. See the rule of thumb posted above your post. An example would be if the acoustic centers are 6.5" apart then the crossing over point needs to be less than 2kHz. If the acoustic centers are 5" apart then they need to cross below 2.6kHz.
I currently have front speakers set up in this exact configuration. I stacked them for the exact reason Danny mentions. I had no idea that I would introduce this issue. Luckily, these speakers were part of a 5-speaker set for my home theater, all of which are matching except that the front L and R speakers are bi-amped. I will experiment with turning the bi-amped tweeter off to see if I can hear the difference. I did have the mid-point of the tweeters at ear level to set both tweeters the same distance to my ear.
Danny, I use software called ScreenCast-O-Matic (screencast-o-matic.com) to make training videos for software that I write. It captures my computer screen while I operate the software. I like that you show the software in this video and would like to see more of that.
And did you notice how bad it sounded? Nope. Cause it didn't.
Sometimes you dont know how bad something sounds until you hear how good it can sound.
I disconnected my bi-amped tweeter and the soundstage and imaging improved. Thanks for the free upgrade!
@@Newrecordday2013 Good is subjective. You could just as accurately say how different it could sound.
Or, "I disconnected my bi-amped tweeter and the soundstage and imaging improved. Thanks for the free upgrade!"
Hi Danny, I see that Aurion has a tunable pair of ribbon tweeters for $299 to add on top of existing floor speakers. What are your thoughts?
They will absolutely cause the same things seen in the video. And be sure to tune in next week when I show what happens when a super tweeter is added to a full range or wide band driver.
Thanks Danny... You may have saved me $300!
You’d probably be better off pointing them upwards or rear wards.
That’s two bad, I’m using too tweeters per side in dipole as an experiment and it’s working! :)
show us the measurements.
SPC design I don’t have measurements, but that’s not a requisite for it to work.
SPC design also, being a dipole configuration, you won’t get the sort of comb filtering that using two tweeters in the same polarity on same face of the baffle brings.
@@RennieAsh In all fairness measurements are the only way to prove whether your right or wrong. You can turn any smart phone into a very accurate microphone for measuring here is the link. www.parts-express.com/dayton-audio-imm-6-calibrated-measurement-microphone-for-tablets-iphone-ipad-and-android--390-810
Tekton Designs uses up to *15* tweeters on one speaker baffle. Hmmm.
What about something like JBL’s VRX932LA, which has 3 tweeters mounted in each box on a mild curve in a line array. The tweeters receive -3, 0, 3 db signal difference. Looking at their literature it seems a single speaker box has a lot of high frequency roll off in the vertical bandwidth but when you mount 3 in an array suddenly the curve above 1khz is fairy flat. Can you explain how they are achieving this? www.jblpro.com/ProductAttachments/VRX932LA-1_FINAL.PDF
Also how would this affect the horizontal as they do not show a plot with multiple speakers
Wow, that's pretty enlightening. Having never really given any thought to how they would interact with each other, it's pretty dramatic. I do have a question that's not necessarily related to this situation. As we get older, (I'm darkening 60's door) our (read my) ability to hear higher frequencies is dropping off. If I'm looking at a frequency response and it's not looking good let's say beyond the 10k-12k Hz, do I really need to be concerned about it?
goddamn physics.
Lot of noise in the background...
tekton double impact speakers have 7 tweeters , and so far not a single bad review from anyone ?
The Tekton approach doesn’t apply here. They are not crossing at the same frequencies.
Sony has 3 tweeters ... Tekton has ... heck of a lot! Interesting ...
That's what I was thinking but one of Tectons videos reveals only the centre driver is being used as a tweeter the rest are being used as midrange and at a distance and upper frequency where comb filtering shouldn't occur.
This seems to be explained by wave interference
Aperion
Yep comb filtering sure looks bad.
Long video kept saying show you & you'll see. Does he actually think he's the only one who knows that comb filtering looks bad? How many people can hear comb filtering? Some of the greatest & most knowledgeable loudspeaker designers of all time put multiple tweeters on their loudspeakers. It works. Has this guy never heard of line array?
Line Source is a different animal all together, not the same thing as the examples being given in this video. Thanks for commenting!
Certainly I am not the only one that knows about or understands comb filtering, but it is a subject that a lot of hobbyist don't understand very well. How many people can hear it? I think everyone can hear it especially when it causes large holes in the output. Have I ever heard a line array? :-) Maybe you'd like to drop in on our Facebook page and scroll through some of the pictures of speakers that I have designed: facebook.com/pg/grresearch/photos/?ref=page_internal
@Nat N Those LS-6's and LS-9's were great speakers. I designed the drivers for those and we only had several thousand made. Once they were all sold out they were gone. I didn't have any more made. I do however offer some kits now that exceed their performance in multiple areas. So if you still want to build a kit that takes you to the top level of design and performance then I can get you taken care of.
@Nat N Check this kit out: gr-research.com/nx-treme.aspx This one uses a planar magnetic tweeter in a waveguide with a low crossover point. The mids are better than the older LS series designs. The baffle is smaller. Imaging is a lot better. It still uses multiple woofers to handle the lower end (very fast) and is full open baffle from top to bottom. And the drivers are physically and acoustically aligned on both sides of the speaker (front and back). They are incredible.
Seriously, go listen to a Stirling Broadcast LS3/6 and claim it's a bad sounding speaker with a straight face. But I suppose designers the likes of Derek Hughes have no clue what they're doing. 😂
If both tweeters are both covering the same range then what is illustrated in this video will be present without question. But if you like the result then happy listening.