Dual Opposed AV15H Subwoofer Build - Page 9

To me the 96db is a heck of a smooth curve without any equalization. And from looking at the 6" away graphs to the LP ones, you must have a ton of room gain.

What bugs me is why the 39Hz room mode does not increase with higher levels.. A room mode just sums the waves at that LP. If the amplitude is larger so will the final sum. Right?
Looks like an EQ might not be helping you for that peak either.

I am not familiar with the Galaxy mic. I have an RS analog one, and when I want to measure at a higher level I need to recalibrate the mic in REW. Did you have to do something similar?

Also, 95db at 10Hz with more room to go sounds good. Does the PB13U do that?

So - big question here...

How does it sound at this point...?

Yup, you probably have it at the 50-100dB 'Medium' range....you'll need to recal with the 80-130dB range selected on the meter....I use the same one.

JSS

Max is right.

Just EQ that peak at 40Hz down and you've got a very nice start at a great response.

I'm interested to know if, when you get things tweaked, you notice a different sonic sig than the Ultra with the 15H subs. I know it's hard top describe such things, but if you could give it a shot when the listening is done, I'll appreciate the input.

Bosso

I think it's possible that a given room mode can only support so much energy, and maxes out beyond that. I may have read that somewhere, or I may have just made it up; latter has been known to happen...

I see. If that's the case, and not some calibration or measurement issue, then I do not see how this can be fixed with an EQ. The amount of equalization would have to be different at every volume level (I am sure I am making this up as I go :-))

"I'm interested to know if, when you get things tweaked, you notice a different sonic sig than the Ultra with the 15H subs. I know it's hard top describe such things, but if you could give it a shot when the listening is done, I'll appreciate the input."

me too. :-)

"Here is the response 6" from the cone for each sub ..."

those close mic and room response graphs looks really good.

i was expecting the sub to start rolling off around 65hz at 12db/oct and that appears to be about what it does. this is another confirmation that models such as winisd are accurate at predicting 2pi response.

the in room response looks great too. you are really picking up a lot of gain. by 20hz, you are picking up about 20db of gain, that's pretty big. after beating down the 40hz mode, you will be into tweaks.

again, very nice.

Yes, if that were the case, and one was a perfectionist, eq would have to vary with level, which is not impossible to do. But really, some of us are happy with good enough, which splitting the difference (say, half the full-blown correction?) might be.

How do I do it? Very tediously.

First, the PB Ultra is EQ'd in the plate amp, so you're comparing an EQ'd anechoic response to an un-EQ'd response, which you can see in the attached graph.

The roll off caused by the 150Hz 4th order LP looks quite normal. The 'drop off' is actually just a fairly normal in-room mountain-valley response seen between 100-300Hz.

YES, the peak is skewing the calibration. That peak is +15dB and will certainly get the meters attention when setting the SW level. So, if the movie clip you're watching has less content at 40Hz and more content below that, it will sound attenuated... because it is.

The pods emerge scene, for example, has content from 4Hz to 90Hz with the strongest content at 25-30Hz, where you subs would be 6-12dB lower than the Ultra, depending on the tune, when both are calibrated the same.

This is a testament to proper calibration.

The graph is labeled, so it should be simple enough to see what is being compared. The Ultra in sealed mode shows a 6dB boosted response, which is what I would recommend for your 15Hs for flat to single digits response.



Bosso

Nice job so far...I hope you figure out the clipping issues because your AV15s with that amp should go louder then your PB13.

When you used WINISD did you add the 2nd order LP filter @ 150Hz to see the simmed slope?

A 2nd order LP @ 150Hz will be down -6dB at 150Hz so it actually slopes before. This is so LP and HP filters slopes sum to create a smooth flat transition at 150Hz.

also the distance of the woofer to the wall could create a null around that frequency (1/4 wavelength is just under 2 feet).

BTW,

The dual, dual-opposed AV15H + RMX 5050 sub system will absolutely dismember a PB13 Ultra, lest anyone is confused.

Penn,

A 2nd order LP @ 150Hz will attenuate the response by -3dB at 150Hz. Most receivers apply a 4th order LP on the sub which will put the response down -6dB at 150Hz. The AV drivers have a natural dip around 200Hz so the response will be naturally headed down a bit above 100Hz in addition to the filters influence.

LTD,

The room is adding approximately +6-9dB of gain by 20Hz. Here's the previously posted graph with PBC's close mic and my own close mic of a Raven, which, as I stated earlier in the thread, is a bit smaller than PBC's box, laid over.

PBC's posted graph indicates a high Q alignment, which I assure you is not possible in the chosen enclosure size. If it were, the Raven would show an even more pronounced high Q peak.

Given that you now have the anechoic response and the in-room transfer function, can you suggest the DCX settings and filters required for a flat in-room response?



Bosso

I even edited above thinking it was -6dB and I was still wrong. I thought 2nd order filter was -6dB? I guess I didnt look at the slopes very well.

While I agree the PB13 looks to be a really nice sub (never heard one), there's no way that little woofer can truly do what you describe. If you look carefully at the distortion charts, the high-level 10Hz distortion is literally off the charts. On the "105dB" trace, the so-called 10Hz output could well be more distortion product than actual 10Hz. Can't cheat the physics...

Gotta agree with Bosso. The two systems are not nearly in the same league, except for maybe parts cost of one vs. purchase price of other.

Then why do you think you're anywhere near max output?

In my experience, REQW is worthless if you're trying to get accurate graphs of maximum output. You can't calibrate it to a volume over 99.9dB and there isn't enough headroom left in the system to keep pushing it. If you just want to see the shape of the graph that's no problem (you just intentionally mis-calibrate it), but if you want the graph to have accurate SPL numbers you're out of luck.

You can turn your receiver / pre-pro up past 0dB, but there's limited room there. In my system I can only go to +10dB. You can turn down the sensitivity of the microphone / line level input on your sound card to gain headroom, but if you have it too low REQW starts complaining.

FWIW, they should let you calibrate up to 120dB.

Basically you can use it to dial in your EQ at lower levels until you get the response you're after and from there you've got to go by ear. Crank the volume with content and listen for distortion or bad noises.

Looks fairly close to me. Real world differences in the subs, the xover point and slope and the Mic + cal file, and the micing scheme could account for the differences.







The bottom 2 are fine and should represent your real response. The top 2 show signs of severe compression or in this case clipping of the SC input. Disregard those.

In order to do high volume REW measurements you need to turn down the input sensitivity of your soundcard and recalibrate to a higher level. I use the maximum of 100db when I'll be getting very loud. This will cause lower volume measurements to register low and be more prone to noise but they are still useable for the most part though. Once you know what your response is, it's easy to spot random weirdness in the measurements that is not real.


Yep. Bosso covered it already. Your peak is dominating the calibration reading so that the rest of the range is much lower in volume. Chop the peak down and recalibrate.






Ok. Just eyeballing your graphs I'd guess at trying a 39hz cut of about 8db with a Q of 4 or 5 and see what that does. Adjust the Q and cut depth until it looks about right. Try to get the 35-50hz range about in line with the 80hz area. As for a low end boost try a 26 or 27hz, 12db shelf filter with about 5 or 6db of boost to start. Don't go too crazy with the boost level. You may need to add another fairly narrow band of EQ cut around the shelf hinge frequency to even that area back out afterwards. Next set of graphs extend the FR down to 8-10hz.

I thought your amp was clipping, nevermind.....