Well, in this instance, you'd be moving the horn forward to shift the nulls to be more evenly centered about the listening axis.
Also, I was describing a left-right shift of the horn relative to the driver. This could help 'aim' the nulls a little more as well, horizontally rather than vertically. The extreme of this would be a side-mounted waveguide.. but by making an angle, you shift the interference lobes accordingly.
Also, I was describing a left-right shift of the horn relative to the driver. This could help 'aim' the nulls a little more as well, horizontally rather than vertically. The extreme of this would be a side-mounted waveguide.. but by making an angle, you shift the interference lobes accordingly.
dnewma04 said:
Outside the nulls, the tweeter is out of phase with the woofer. So it's not the ideal listening position.
I know you're trying to drive home the point that as long as you're 10 or 12 feet back, you're alright. That's true.
Be sure to understand that if you're less than 10 or 12 feet back from one or both speakers, you're not alright. You're probably sitting in a null.
That means at every show, the listeners were treated to nulls in their face. The rooms aren't big enough for listeners to be far enough away. No problem, we all know show conditions suck. But then again, so do smaller living rooms and many bedrooms and offices. Recordsing studios are out - too close. There are lots of applications where you may need to be closer than 10 feet.
If you want to use speakers that have a narrow vertical null angle, I suggest angled stands or risers. It's just good practice. In fact, I like to use angled risers even if the nulls are spaced farther apart. Never hurts to put the forward axis - where there is closer to zero phase offset between woofer and tweeter - right on target where the listeners are.
noah katz said:
Nope, you are wrong once again with respect to my participation. He said he doesn't have the TIME to post them, not that he doesn't have the data. In fact, he has the data on his systems, and says they are all about the same, but:
gedlee said:
That's in the Nathan thread, and a bunch of us then proceeded to demonstrate that many manufacturers posted that data for their systems, most prominently JBL.
I don't know where Wayne found the Summa verticals, but Markus measured his Nathans, and it's clear why Earl has "declined" to post them for his products:
http://www.diyaudio.com/forums/showthread.php?postid=1623275#post1623275
At another point, Earl asserted that ALL products have terrible-looking verticals and we laid THAT rationalization to rest, as well.
Earl tells us that the nulls are at ~20° across all of his offerings, and that they affect the response at narrower angles, as well, very much as Wayne is documenting here, under apparent protest....
"I will post the vertical at some point, but owing to the fact that its not as good AND its not as important as the horizontal, I'm not in a hurray to do that. The point here is that no one else ever shows this data at all!! "
I guess I misremembered - I apologize for wrongfully accusing you.
I guess I misremembered - I apologize for wrongfully accusing you.
Wayne Parham said:
Aw, c'mon Wayne, there's no need to exaggerate. It's unlikely that you'd be sitting in a null at any reasonable distance. Sure you might be able to stand up into a null or lie down into a null but that's not what you said.
And, I'd think that most people considering 15" floor-standers would be able to get 10'-12' away from them anyway. I mean, they aren't really designed to be bedroom speakers. 😉
noah katz said:
It's not a problem really, Noah, and I understand why members wanted me to dial this down in Markus's Nathan thread; it was somewhat inappropriate there. In this thread, however, these matters are decidedly on-topic....
Wayne Parham said:
*sigh* driver spacing means nothing with out reference to frequency overlap or crossover frequency. Why do you keep fixating on driver spacing as if it is an issue in a vacuum? I could have two woofers spaced 20 feet apart and still have a big lobe if they were playing 5hz signals. So there are two-inherently related-aspects and thus solutions to the lobing problem. Push the drivers together or cross low. Earl crosses low, you push the drivers together. Whether one is better than the other is completely implementation specific, you just can't make a general statement about it.
noah katz said:
Because there's a big ol' hole in the spectral content of the first reflections from the floor and ceiling.
As we discussed before, that requires corrective damping and/or diffusion to mitigate.
augerpro said:
The sims show it: you can't cross low enough to move the nulls outside the dispersion pattern using a round 70° - 90° axisymmetric waveguide.
Further, it's obviously not as if the nulls occur at only the angle where they are deepest from full interference cancellation; their influence extends well beyond there, and may be seen in the measurements as affecting the response at narrower and wider angles, as well....
"Because there's a big ol' hole in the spectral content of the first reflections from the floor and ceiling.
As we discussed before, that requires corrective damping and/or diffusion to mitigate."
Fine, but this is a 2nd-order effect far different than the original assertion that the notch could be in the listening window.
It could even be a 3rd-order effect - with toe-in, the reflections will likely miss the listening position.
Depending on how narrow they are in the horizontal axis; still waiting for Wayne to address that.
As we discussed before, that requires corrective damping and/or diffusion to mitigate."
Fine, but this is a 2nd-order effect far different than the original assertion that the notch could be in the listening window.
It could even be a 3rd-order effect - with toe-in, the reflections will likely miss the listening position.
Depending on how narrow they are in the horizontal axis; still waiting for Wayne to address that.
noah katz said:
Considering that there may also be a hole at a different frequency in the first lateral reflection, as well, the effect on the perceived spectral content is indeterminate. Clearly, it's far better if these anomalies are not there to begin with, kinda like the HOM thing, except that these we can actually measure and see.... 😉
"Considering that there may also be a hole at a different frequency in the first lateral reflection, as well, the effect on the perceived spectral content is indeterminate."
You rest my case.
You rest my case.
ZilchLab said:
You can't? So physics breaks down in your world? The sims show no such thing.
Now I remember why I stayed off the forums all winter. This is just ridiculous.
Here is an idea for these little rectangular horns that everyone crosses too low. Cross it right where the horn has completely lost pattern control vertically. You can optimize on axis fine, while as you move offaxis vertically the null will be filled in by the response flare. It won't be textbook but then neither are non-coincident sound sources. Might be a bit messy and unpredictable but I bet the sound power would average out to be smoother.
The flare is sort of makes up for the driver's natural directivity and non-coincident layout.
Not sure if this is genius or really stupid. After I take an Ambien I can't really tell...
The flare is sort of makes up for the driver's natural directivity and non-coincident layout.
Not sure if this is genius or really stupid. After I take an Ambien I can't really tell...
Member
Joined 2003
Is the current discussion of CTC overrated?
From my (admittedly limited) understanding, low frequency "cutoff" determines the minimum mouth dimension, whether the horn/WG is axisymmetric or asymmetric. Assuming the same relative off-axis vertical pattern control, the CTC distance will be the same for both (but the asymmetric will be ~2x the physical width of the axisymmetric).
If the above is true, a different discussion might be the relative value of the two different radiation patterns. Personally, I like the idea of an asymmetric pattern, but don't have the space to house a WG twice as wide. Whether an individual chooses axi/asymmetrical can be more about design trade-offs than "superiority" of either approach.
From my (admittedly limited) understanding, low frequency "cutoff" determines the minimum mouth dimension, whether the horn/WG is axisymmetric or asymmetric. Assuming the same relative off-axis vertical pattern control, the CTC distance will be the same for both (but the asymmetric will be ~2x the physical width of the axisymmetric).
If the above is true, a different discussion might be the relative value of the two different radiation patterns. Personally, I like the idea of an asymmetric pattern, but don't have the space to house a WG twice as wide. Whether an individual chooses axi/asymmetrical can be more about design trade-offs than "superiority" of either approach.
catapult said:
If you're facing the woofer as opposed to facing the tweeter and you're less than 10 feet away, you're sitting in a null. It doesn't get any more plain than that.
I told you one way you can help this situation - Just angle the speakers down a few degrees. At least that will split the difference, push the troublesome lower null down a bit.
Paul W said:
Hi Paul
I tend to agree with you, which is why I am not jumping to any conclusions about how an elliptcal waveguide will work out in a real situation. In all of my room designs I control the floor and ceiling bounce and I'd have to say that I am still inclined to prefer this approach which makes the narrower vertical pattern above crossover moot.
In the elliptical waveguide that I am making, I took an approach opposite to what you suggest, for exactly the reasons that you suggest. I could not see doubling the width of the waveguide in order to maintain control in the vertical, so I maintained the width and halved the height. Now this is certainly less than ideal and precisely one of the trade-offs that you alluded to. This shortened vertical dimension will cause a fluctuation in the vertical polars at precisely the crossover - unless the crossover point is raised, which in itself is not desirable. Now, and I'm sure there will be lots of guesses out there, it's not at all obvious if this will improve or degrade the vertical polars at the crossover.
As I have said over and over, it is not clear that an asymmetric device will yield an improvement or not. Everybody wants to make it so simple - its not.
augerpro said:
Look at post #992. Perfect. Two problems:
1) Can't cross a 12" round waveguide at 800 Hz, and
2) 12" woofer dispersion is how wide at that frequency?
Wayne Parham said:
Wayne, you're doing what JonMarsh calls obfuscation through elucidation. 😉
The woofer of the Summa is sitting on the floor so how could your ear be at that level when you're sitting in a chair? Even the tweeter (center of the horn) is well below ear level so a lobe that's tilted upward would probably be a good thing to get the lobe centered on your ear. Now I've never seen, let alone heard or measured, a Summa and I'm not claiming Earl made all the same choices I would have in the crossover. And he will tell you I'm not hesitant to speak up when I disagree with him. But for a guy who claims to be an objective scientist, you're spewing nonsense. It's beneath you.
- Status
- Not open for further replies.