Unique question about comb filtering of speakers...

[bbecker1983]

Unique question about comb filtering of speakers...

How does one measure its effect?? Other than a shift in the distance between say tweeters to the listener?

I wonder how Axiom and M&K get away with this, let alone a Line Array?? Or do these designs not apply because the tweeters/drivers are on the same baffle and aligned so closely that the sound produced arrives equally?

I can understand that improper cancelation cause of out of phase sound could cause this too but taken that into account couldn't one produce a speaker designed properly w/o the issue of CF?

Bobby

[fbov]

Comb filtering is ubiquitous. Stand between two speakers and listen for sounds that seem to originate between them. You are hearing comb filtering; identical signals from two sources received by sensors roughly 8" apart (i.e. your ears). It's associated with a sense of spaciousness, a major result of strong first-reflections, another source of comb filtering.

If comb filtering were an issue, we'd listen to mono in anechoic chambers. Generations of listeners agree that stereo is superior and moderately lively rooms sound best. This is more a topic for those who like to argue the unimportant, or believe strongly in unsubstantiated (or refuted) rules of thumb.

Have fun,
Frank

[cjd]

There are many reasons it may work. Or not. And there's no guarantee that just because someone out there sells a product that does, it works.

1. People don't actually use their ears and their brains at the same time, so they don't hear the problem, or don't care. Horizontal MTM's are easy to hear if you can. I can make dialog almost impossible to understand using my MTM's placed horizontally and just slowly tweaking toe-in. Pretty simple experiment with very obvious results. Catching it with music is a lot harder for most people.

2. The lobing (in M&K's case) is intentional, not to mention carefully crafted, to limit vertical dispersion.

3. The issues (line array) are moot at distance. Line arrays are interesting beasts.

I have no idea the specific issues Axiom has introduced as I've not measured that setup.

Dual mono != stereo, and you almost never get exactly the same signal out of both speakers exactly out of phase once it hits your ear, but if you DO play dual mono and make sure the signal reaches your ear 180 out of phase, you get a very weird (as in interesting weird) thing going. And as you rotate closer to in-phase it does fun things. That is how we place things spatially.

As with most things, comb filtering, when discussed with respect to speaker design, is largely speaking to single-unit related issues where you end up with large dips or nulls in response as a result of driver spacing vs shared frequency load.

[bbecker1983]

I've always questioned the belief of the phenonema but never understood how it was perceived.. It almost like the DSP effects (psycho-accoustics) where careful timing and phase changes create a odd effect of spaciousness.

Lobbing I've heard and recognize myself but never been succumbed to comb filtering (that I know of) I have listened to the MK speakers and never heard anything I could say bad about them other than they are overpriced!

I still haven't found anything really bad per say about the Axiom or the MK designs..

Image not available

[cjd]

No.

Comb filtering in a design, such as an MTM placed horizontally and then listened to (or measured) off-axis, has very clear results. From my MTM pages:

Off-axis measurements with the speaker horizontally placed, as it would be when used as a center channel. If you remain witn 15-20° off-axis at the edges I think that there will be no trouble. For most folks sitting on a couch 2-3M back, this is reasonable. However, as you can see the off-axis measurements at 30° and above are rather dismal. On-Axis (yellow), 15° off-axis (red), 30° off-axis (green), 45° off-axis (pink), and 60° off-axis (blue).



It's not a psychoacoustic issue, and it's not something to believe without basis. It is very clearly and easily measurable, and the results are audible.

Comb filtering, as we usually worry about in a speaker design (such as 2 tweeters) is NOT the same as DSP effects. It is cancellation we're worried about.

[Rick Craig]

Easy to measure but difficult to define how much impact it has on our perception of performance. Right now I'm working on a CBT array project and will see how things progress.

[cjd]

Easy to measure but difficult to define how much impact it has on our perception of performance. Right now I'm working on a CBT array project and will see how things progress.

Totally agree. I was just arguing against the idea that it was something you only heard because you thought you should.

It's like wiring crossovers wrong - most people don't notice. Many will hear it and realize something is a little different, but pinning it down to a reverse wired driver is hard.

With the MTM's it really is possible to make it nigh impossible to understand what is being said in a movie. But with music, the response dips are far less obvious.

And arrays absolutely are a very different beast.

As always, when you understand the pile of issues to worry about, you can adequately pick and choose what compromises are acceptable.

[Rudolf]

If the comb is the result of two different sound sources interfering at the head, it is easily recognisable by our pair of ears and the brain and does not offend.
If the comb is coming from ONE point, the brain will recognise it as a bad thing.

That's what Tool's book says.
The horizontal center should be something in between.

[twest820]

Agree with both cjd and Rudolf. Any thoughts on what the required separation for sources to be separate is? I would assume the usual 2ms or so.

OT: Spaceballs! :banana:

Image not available