EQ away comb filter?
So that's what it looks like with 2 drivers. It's the same with two microphones as well. In fact, when you see someone on at a podium with two microphones, a smart engineer is likely only using one of them to get rid of these interference effects.
When the placement of two drivers is small compared to the wavelength, then they are working together and cannot cancel each other out (when fed the same signal). When they are further apart, the difference in distance to the ear allows the two signals to arrive 1/2 wavelength or more apart in time.
The constructive and destructive interference results in points where the signal is twice as loud and points where it cancels completely to silence (for a given frequency). These points are clustered into "lobes" of higher and lower sound pressure levels.
So that's the real reason why the answer is "no" to your question. If some areas have a signal at 2x loudness and others are at 0x loudness, how can you boost or cut a signal to compensate? At the 0x spots, an infinite boost won't help, meanwhile the 2x spots just get louder.
Now, one method of EQ to eliminate the comb filter effects IS available. A complete cutoff (shelf filter) above the critical frequency. This will kill the comb filter effects.
At that point, you need to either live without those frequencies, or supply them in another form. Both methods are employed, as is just leaving the comb filter effects in the soundfield by ignoring them.
Many concert systems roll off early, even as low as 8kHz to get rid of the annoying comb filter effects. One member here added a single supertweeter to add the extra highest frequencies back in.
see http://www.diyaudio.com/forums/show...e=25&highlight="apex+jr+tweeter"&pagenumber=3
The is a pretty decent solution for home listening. The technical problem with it is that the mix of the two types (line source and point source) results in the balance of sound received being different at different distances. (If you get up close the single driver gets louder and as you get further away the line source stays louder and the point source gets quieter faster).
For a given listening distance, you can just adjust the level of the super tweeter for pretty acceptable results.
If you have a somewhat small vertical listening zone, and a fixed number of tweeters, you can try packing the tweeters really tightly vertically at ear level. This will limit the distance of the nearfield zone, so you'd want to be careful of this.
The ideal line source array would be a continuous full range driver that goes from the floor to the ceiling (and touching the highly reflective ceiling being more important than extending all of the way to the less reflective floor).
It would also be infinitely narrow (horizontally).
Understanding how departures from the ideal will affect the sound will allow you to make something listenable.
Any departure that involves a placement error of less than 1/2 wavelength of the highest frequency being driven can be considered an non-issue.
Limiting the listening area to a zone of accurate sound can allow you to make many compromises as well. It doesn't sound perfect when you are laying on the floor, or if you are 7 feet tall and standing--so what?
By the way, the 1/2 wavelength rule works in all three dimensions. Often, you can pack your tweeters more tightly by making a zig-zag line instead of a straight line. The zig-zag doesn't even have to be left and right: you can put half of the tweeters a tiny bit closer to the listener so that the driver mounting hardware is overlapping (but be careful of other effects that may be introduced by this, like diffraction created by edges).
As long as the individual drivers are within 1/2 wavelength of our theoretical line and no gaps of greater than 1/2 wavelength exist, we still have a perfect line source up to that frequency.
So that's what it looks like with 2 drivers. It's the same with two microphones as well. In fact, when you see someone on at a podium with two microphones, a smart engineer is likely only using one of them to get rid of these interference effects.
When the placement of two drivers is small compared to the wavelength, then they are working together and cannot cancel each other out (when fed the same signal). When they are further apart, the difference in distance to the ear allows the two signals to arrive 1/2 wavelength or more apart in time.
The constructive and destructive interference results in points where the signal is twice as loud and points where it cancels completely to silence (for a given frequency). These points are clustered into "lobes" of higher and lower sound pressure levels.
So that's the real reason why the answer is "no" to your question. If some areas have a signal at 2x loudness and others are at 0x loudness, how can you boost or cut a signal to compensate? At the 0x spots, an infinite boost won't help, meanwhile the 2x spots just get louder.
Now, one method of EQ to eliminate the comb filter effects IS available. A complete cutoff (shelf filter) above the critical frequency. This will kill the comb filter effects.
At that point, you need to either live without those frequencies, or supply them in another form. Both methods are employed, as is just leaving the comb filter effects in the soundfield by ignoring them.
Many concert systems roll off early, even as low as 8kHz to get rid of the annoying comb filter effects. One member here added a single supertweeter to add the extra highest frequencies back in.
see http://www.diyaudio.com/forums/show...e=25&highlight="apex+jr+tweeter"&pagenumber=3
The is a pretty decent solution for home listening. The technical problem with it is that the mix of the two types (line source and point source) results in the balance of sound received being different at different distances. (If you get up close the single driver gets louder and as you get further away the line source stays louder and the point source gets quieter faster).
For a given listening distance, you can just adjust the level of the super tweeter for pretty acceptable results.
If you have a somewhat small vertical listening zone, and a fixed number of tweeters, you can try packing the tweeters really tightly vertically at ear level. This will limit the distance of the nearfield zone, so you'd want to be careful of this.
The ideal line source array would be a continuous full range driver that goes from the floor to the ceiling (and touching the highly reflective ceiling being more important than extending all of the way to the less reflective floor).
It would also be infinitely narrow (horizontally).
Understanding how departures from the ideal will affect the sound will allow you to make something listenable.
Any departure that involves a placement error of less than 1/2 wavelength of the highest frequency being driven can be considered an non-issue.
Limiting the listening area to a zone of accurate sound can allow you to make many compromises as well. It doesn't sound perfect when you are laying on the floor, or if you are 7 feet tall and standing--so what?
By the way, the 1/2 wavelength rule works in all three dimensions. Often, you can pack your tweeters more tightly by making a zig-zag line instead of a straight line. The zig-zag doesn't even have to be left and right: you can put half of the tweeters a tiny bit closer to the listener so that the driver mounting hardware is overlapping (but be careful of other effects that may be introduced by this, like diffraction created by edges).
As long as the individual drivers are within 1/2 wavelength of our theoretical line and no gaps of greater than 1/2 wavelength exist, we still have a perfect line source up to that frequency.
if the tweeter spacing is 1 inch center to center the available frequency before combing is, lets say x for this example. so this array is good until we reach combing. And the other array has 3.5 inch center to center, its combing frequency is y. Can not this be done in the area of our normal listening environment. Making each array work in its own frequency band before we experience combing. It gets worse as we get to the bass region i assume so we do not use an array for low frequency. i have listened to some very good music on line arrays, they do have some very good aspects. I am not a physicist so i do not know the math for these systems. Just trying to learn as i go. Thanks Tad
I'm not exactly sure what you are asking here.tryonziess said:
In an array, there will be a frequency above which combing starts to occur.
For the tweeters at 1 inch c-t-c spacing, this theoretically begins at about 6500 Hz. (Assuming that the tweeters are radiating in an omnidirectional manner.)
For the other array at 3.5 inches c-t-c, the combing begins at about 1940 Hz. (If they are each radiating equally in all directions--at least in the half space facing outward from the speaker.)
In practice, this tends to be a bit overkill. Spacing individual tweeters at less than a centimeter apart rarely happens to make the 1/2 wavelength rule at 20kHz. And the combing between drivers spaced more than 1/2 wavelength and less than 1 full wavelength tends to shoot nearly straight up and straight down. At 1 wavelength spacing, lobes shoot 45 degrees up and 45 degrees down from the drivers.
This means that if you have 1 wavelength spacing at 20k and your drivers go from floor to ceiling in an 8 foot tall room, a destructive lobe could hit your ears if you get really close or really low to the ground or really high towards the ceiling (and you are only 4 feet away). Even then, it will only affect 20k and higher.
Unless you have a reflective ceiling. Then that lobe goes up and back down on you. With a highly reflective floor and ceiling, you can get several bounces for anything larger than 1/2 wavelength.
Absolutely. Use a crossover between the two arrays.
I'm not sure what you mean by "it gets worse" there.
The spacing gets much easier with lower frequencies. For example, if you have woofers that won't be playing anything higher than 200Hz, these can be placed 33 inches apart or closer, center to center.
Those are the theoretical limits. You are always safe packing the drivers closer (other than the crazy expense). You can get really acceptable results with drivers spaced further apart as well. With the exception of line arrays using continuous ribbon drivers, almost no one packs their tweeters within the 1/2 wavelength rule. Especially not floor to ceiling. That would mean 284 tweeters -- per channel.
If you are that crazy/ambitious, the Apex Jr. tweeters come packed 333 to the box, at $166.50 per box, plus shipping.
AFAIK, it's not dead, however, as always when a high-end / priced product is introducted, there is an initial flurry of interest, sparking debates, which then generally ease off over time. The IDS is clearly a low-volume, bespoke product, built to order, probably with a high profit margin (you either go high volume, low margin, or low volume, high margin), so in terms of making money, you don't need to sell too many per year to make a living. Which is fair enough.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.