PDA

View Full Version : HK 635 and BFD tag-team against the Room from Hell



sensibull
05-07-2005, 08:42 AM
As some of you may remember, my HK 630 recently died on me and HK was kind enough to replace it with the 635, which includes a fairly advanced (so I'm told) EQ ability, particularly with regards to subs. Just before that happened I bought a BFD in the hopes of taming my impossible room. At first I thought the BFD would be redundant, but fortunately that turned out to be anything but the case.

Anyway, I thought it might be helpful for some to see the interactive results between an AVR's EQ and the BFD, particularly when faced with an entry level sub and a very difficult room (http://www.mindserai.com/room%20layout.jpg). As you will see in the .jpg, my room has all the textbook flaws -- its dimensional ratio is almost exactly 2:1, the surrounds are asymmetrically placed and not enough behind the seating, the seating is against a back wall (and a windowed one at that), one main is against a side wall while the other is not, the walls are open in two directions -- the list goes on and on. Unfortunately, for a variety of reasons, none of this can be changed, not even the placement of the sub.

I've been playing around with the HK's EQ, the BFD, and Room EQ Wizard (http://homepage.ntlworld.com/john.mulcahy/roomeq/index.html) over the last week or so, but last night I finally got serious. After a processor reset to make sure everything was set to default, I ran the HK's EQ process in manual mode, with the 340s and 170s set to small, crossovers at 80hz, and the Dayton sub output dial at about 1/4 - 1/3. Then I did a sine sweep from 20-160 in 2hz incremenents with Room EQ. The results are displayed in the Test One graph here (http://www.mindserai.com/FourTests.jpg).
As you can see, the results were pretty bad. The HK did a halfway decent job of flattening the low end, but things got progressively worse closer to the crossover point, with one bad null around 88hz, one outrageous peak at 96hz, and lots of zig-zagging elsewhere. I let Room EQ suggest some filters for me (it will only cut peaks, not boost nulls), tweaked them a bit while watching the graph, entered them, and tested again. The results are shown in Test Two.

No idea why Test Two showed a new peak around 30hz, because I definitely didn't boost anything near there, but otherwise there was a marginal improvement in the bad peak and null. For the next hour or so I continued to tweak the filters and adjust the gain on the sub, but whatever I did I couldn't tame that 96hz peak. In fact, nothing I was doing seemed to touch it at all. Then it dawned on me that the left main's corner placement could be the culprit instead of the sub. Around those tones I was hearing a pretty strong "muddy" resonance from that corner, but I just didn't put 2 and 2 together. So for Test Three, I raised the crossover to 100hz and re-ran the HK's EQ. As you can see, that was the ticket and flattened things out nicely. For Test Four, I simply added one additional filter (making a total of 10) to slightly boost the null at 82 hz. Now, with the exception of one 10db null at 80hz (which I should be able to take care of by widening the filter at 82), I'm at +/- 6db from 20 to 160 Hz -- not exactly prizeworthy, but not too shabby either given the Herculean challenges presented by by room.

Aurally, I'd describe the end result the same way others have described it after EQing their rooms -- cleaner, more accurate, more refined. I would not, however, call it "night and day" different (maybe my ears just suck). I'd been listening to those mid and upper bass peaks for so long, that it's hard to shake the feeling that something is missing, particularly now that the 340s are crossed at 100hz, but the clarity and the seemlessness of the front stage more than make up for the loss of thump.

I'm still learning all this stuff, and sometimes find it difficult to keep track of how each variable will effect the next step. There are many possible ways of approaching this task -- I took just one and arrived at decent results. Am I done tweaking? Definitely not (I will probably create a more bass-heavy curve for movies) but hopefully I can leave things alone for a little while and just enjoy the music...

If you've actually made it this far, thanks for reading. Feedback would be much appreciated -- I promise to keep my Destroyer on a short lease ;-)

shane55
05-07-2005, 11:33 AM
sensibull...
Great post... Congrats on taming the beast! It's amazing how 'used to' sounds and spectra we can become. Right or wrong.

One thing I've learned in testing (new speakers specifically) is to *not* listen to the old ones... the ones that I'm 'used to'... for an extended period, while I only listen to the new ones I'm testing. This seems to 'neutralize' the listening experience somewhat. While at first the new ones might sound 'wrong', it's amazing how quickly one can become 'used to' the new sound and eventually think the old ones sound strange... ;)

Anyway... thanks for the post, and especially for the links.

Cheers.

shane

Nicholas Mosher
05-07-2005, 11:56 AM
By your graphs (and there are no X-axis graduations) it looks like you could do a much better job with more practice. It looks like you still have +/- 6dB swings. With the BFD you should be able to get +/-2dB or better (over the majority of the 20-80Hz range). Pre-measure your response with the BFD in bypass, then use 5/60 bandwidth settings and start at 30Hz placing a filter in each 5Hz increment up to 85Hz. Retest after each filter and aim to make all your levels the same SPL. You could also try placing a gentle descending curve from 30Hz to 85Hz. Some of the filters might be redundant, but this setup will give you a good idea of what it can do, and a platform to work from.

sensibull
05-07-2005, 03:00 PM
You're probably right, Nicholas, and thanks for the suggestions and feedback. Unfortunately, it took my private think tank seven days to plan a strategy for getting the wife and twins out of the house long enough for me to take a honest crack at this. And boy will it cost me tomorrow...

Anyway, like I said, I ain't done climbing, just resting on a ledge comfortable enough to spend the night.

But whiles I'm asking, is there any one else out there who crosses over their 340s at 100hz? Is that just a stupid waste of these speakers?

bikeman
05-07-2005, 04:54 PM
But whiles I'm asking, is there any one else out there who crosses over their 340s at 100hz? Is that just a stupid waste of these speakers?

It is not a stupid waste. You're losing just a bit on the low end but apparently gaining elsewhere. I hope my next receiver has a variable crossover so I can experiment with the 340's. I locked into 80hz with my Kenwood. I could use the crossover on the sub but I prefer not to.
Go with what works. The heck with conventional wisdom.

David

sensibull
05-11-2005, 01:14 PM
FWIW, Nicholas was right. I blocked out some time today and ran like 15 sweeps, tweaking my filters before and after each. My idea was to reset my HK's EQ and start with the BFD -- hoping the HK could fine tune anything I couldn't tackle with the BFD (instead of the other way around, which was how I approached it before).

Anyway, as far as I can tell, the HK is NOT trying to achieve a flat curve because (with the exception of some 20-30 Hz boosts I can't manage otherwise) it just wrecks whatever I can achieve with the BFD. So for now, I'm leaving the HK's EQ off and will sit tight with the following response (just a hair off +/- 2db from 36 to 80 Hz, and only +/- 5db fro 36 to 200 Hz.)

(edit: and this is with the crossovers back where I want them -- all 5 at 80hz -- so the midrange punch is back without sacrificing the clarity and seemlessness achieved before)

20.0, 63.80
21.2, 64.10
22.5, 64.51
23.8, 64.42
25.0, 65.43
26.6, 70.24
28.2, 70.19
29.9, 67.77
31.5, 72.06
33.6, 75.53
35.8, 81.92
37.9, 79.97
40.0, 77.62
42.5, 79.15
45.0, 81.20
47.5, 77.78
50.0, 77.75
53.2, 77.93
56.5, 81.10
59.8, 79.94
63.0, 77.96
67.2, 78.53
71.5, 78.02
75.8, 80.58
80.0, 82.91
85.0, 81.48
90.0, 77.96
95.0, 79.49
100.0, 85.75
106.2, 87.11
112.5, 86.61
118.8, 83.67
125.0, 81.27
133.8, 82.34
142.5, 82.80
151.2, 81.44
160.0, 83.92
170.0, 83.06
180.0, 85.55
190.0, 83.35
200.0, 81.96