Input on driver selection

[dlr]

Window options

I'm a bit late to the party. Did you figure out how to set the window in ARTA yet? Left click before the impulse and right click before the first reflection.

That should do the trick. It looks like only one of the two markers was placed. What windows are available? A half Blackman-Harris is my preferred window. There's a good Wikipedia page showing the different windows and the way the results fall into the MLS bins. The choice depends on what one is seeking and I can't explain the various uses. The Blackman-Harris looks to work best for basic SPL response. The Hann doesn't seem to have the selectivity (if that's the description to use) to produce the closest response to actual given the spacing and height of the bins around the center frequency, but it looks like those two are closest to each other. Different needs dictate different windows.

However, it MUST be a half-window.

Dave

MLS Windowing

Here's another link. It looks like the Hann window may not to go zero at the stop marker. There's a one or two line description below each that explains the basic use. Those that indicate tone measurement are probably the best for SPL frequency response.

More window information

[Face]

Ok, you can say I've changed a few things around.

This is with a ScanSpeak Air Circ up top, Seas Excel W12CY in the middle, and a Seas Excel W26FX on the bottom.



I do have a few questions about the dip centered @ 350hz and the dip at 60hz. They aren't in my plots from a few months ago, and I haven't changed the crossover since. I did move the speakers around a little since(which may explain the dip at 60hz), I lengthened the woofer's port, and I spiked the cabinet. But, here's where things get weird. This was taken with speaker workshop from the same position.



Can anyone explain the differences?

Otherwise, I'm very pleased with the sound and never had so much enjoyment listening before. The speakers disappear in the room, they're very transparent, and full top to bottom. The only thing I may change is add a small amount of padding to the tweeter since I like things on the neutral/warm side. Otherwise, I plan on building another pair of nearly identical cabinets, but doing a neater job this time so veneering won't be a nightmare.

[Face]

Ok, here's a graph that tells a little more.



Blue=Port
Green=Woofer
Red=Mid
Purple=Tweeter

Those were all taken nearfield, mounted on my baffle/enclosure.

Off white, taken 3' way, midrange level.

The measurements of my other speaker look similar, except it's flatter below 200hz, must be due to the room.

[Face]

Ok, not everyone at once.

I believe I finally have something that looks good. But, my only concern is that the break ups may be too high.

[JonMarsh]

What is your target transfer function overall? It seems your midrange doesn't follow an upper end roll off very long- the possible difficulty is that linear distortion artifacts may be only 15 dB down, and you're never want to tolerate that for non-linear, would you? AT least the woofer peak is about -40 dB.

How does the phase tracking look, and what about off axis behavior, at least laterally?
On business travel this week, not much time to read or post.

[Face]

It was LR2, but the mid rolled off too early and created a BBC dip. I then changed a few parts around to correct the dip and now I have the funny rolloff. I ordered a couple more inductors to play with, hopefully they'll be here before Friday so I can fix the roll off. I'll measure off axis again tonight as well as phase. I have a feeling I may have to use a 3rd order LP on the mid.

Next time I'm following your advice from the start.

[Face]

Ok, aiming for steeper slopes.


Better?

There is a little hump @ 1700 at 30 and 45* off axis.



Min phase.


Excess phase.

I'm waiting on some parts to tune the mid's LP. As for the mid's HP, the next pair of cabinets I build(in the process now), will have a smaller mid chamber.

TIA,
Mike

[Face]

Started putting together my new cabinets tonight. I work slow, so I hope to have them done by the end of the year. The baffle widths at the woofer and mid are the same as my ugly 2 piece setup, so the x-over shouldn't change, much. I'm going with 18mm baltic birch with the front panel doubled up. My biggest concern is fitting the crossover inside, I should have enough room between the bottom of the cabinet and the woofer section.

Speaking of, I have some new measurements on my other computer. Laying one of my GIK panels on the floor between the speaker and mic removed the floor bounce dip and smoothed out the rest of the response curve too.