What's going on around my crossover freq?

[pg_rider]

Did some basic measurements with the Hsu bass disc test tones and my Radio Shack meter. I got some very curious results that I wanted your thoughts on.

First, my setup. I've got dual VTF-3's (a Mk III and Mk IV) placed symmetrically up front with my LSA One Monitors. I've run Audyssey XT32 with my Denon X4000, then raised the summed sub level by 3dB. Finally, I left Dynamic EQ on for all measurements.



I ran the tones four times, each with a different crossover setting (80Hz, 100Hz, 60Hz, and 120Hz). At the 80, 100, and 60 Hz settings it appears I have a big problem (dip) in the 80-100Hz range. However, at a 120Hz crossover the response balances out quite nicely. Below 80Hz and above 100Hz the response is very even regardless of the crossover setting.

For the lower crossover settings, does this mean Audyssey is doing a poor job blending the subs and mains around the XO frequency? At first I thought I simply had a bad null around 80-100Hz (which seemed too wide of a band to have a null in), but with the 120Hz XO setting the results are much better. Is that because the sub is flatter through that range than my mains and therefore doing a better job?

Interested to hear your thoughts. Oh, I know using this limited set of test tones and a Radio Shack meter isn't a very high fidelity test, but it at least gives me some data to work with.

[JerryMeeker]

The smoothness of frequency response at the crossover point is influenced by how well the phase of the subs and the mains match up. Unfortunately, Audyssey calibrates the subs and the mains separately, and makes no effort to measure the subs and mains together to optimize the phase at the crossover. This is a common discussion on the Audyssey thread over on AVS Forums. There is a procedure, referred to as the "Sub distance tweak", that involves adjusting the subwoofer delays (distances) while measuring frequency response at the splice, in order to achieve the smoothest response. In many cases, people who have followed the procedure have been able to significantly improve the frequency response.

Addressing this issue, as well as assessing the overall performance of your audio system, really requires a more sophisticated measuring system. One of the more popular systems is REW, which is free software, and can be implemented with a very reasonable investment of around $100 for a USB microphone and cables.

[Tom C]

I have a similar problem. It's centered around 63Hz. As I raised the XO the dip would rise.

I then tried moving mains closer and further from the back wall. As I moved them closer the dip would rise. I got it to an acceptable level. You might try that if you haven't already.

[SME]

I definitely have to agree with jerrymeeker here that the Audyssey MultEQ auto setup often selects a sub distance that yields poor results. I wrote my own thread on this subject along with a suggested course for trying optimize it without sophisticated measuring equipment. (Note: your mileage may vary.)

http://forum.hsuresearch.com/showthread.php?t=95746

For me, optimizing the distance was a literal night and day difference. The 80-100 Hz range is crucial for music. Problems there give the impression of "weak bass" regardless of how good the performance is further below.

[pg_rider]

Thanks for the replies. I definitely understand the value of an REW setup -- just haven't invested the time or money to do it yet.

I didn't realize Audyssey did a poor job around the crossover frequency -- I thought XT32 was the end-all be-all of room correction for all things subwoofer-related. When I get some time I may play around with the sub distances to see if I can get a smoother response. It's weird though that the response is so much more even with a 120Hz crossover compared to the others. I certainly have no problem leaving the crossover that high with the symmetrical placement -- no issues with localizing either of the subs...