EQ frequencies posted...please offer suggestions (1 Viewer)

[Gary Thomas]

Freq.Seat1 Seat2 Seat3

17hz77.576.578.5

20hz86.584.586.5

22hz85.583.585

25hz84.58282.5

28hz84.58380.5

32hz858581.5

36hz8384.580.5

40hz8767.585.5

45hz827684.5

50hz8274.574.5

56hz8267.575.5

63hz74.568.580

72hz6967.575.5

80hz68.576.581.5

89hz55.572.580.5

Everything looks great below 40hz... It's amazing how different the response can be from the 3 locations. Seat 1 is my recliner, seat 2 is a loveseat in the center of the room and seat 3 is a couch along a side wall.

What adjustments should I start with on my BFD?

 

[ColinM]

Watch the movie from seat 3!

Sorry....

 

[Wayne A. Pflughaupt]

Gary,

Which is the primary viewing position? That’s the one you want to EQ for.

In my experience, if the main position is in the room (i.e., out some feet from the wall) you can EQ from that position, and it will deliver acceptable sound at all other positions except those situated against a wall.

Regards,

Wayne A. Pflughaupt

 

[Gary Thomas]

The main position would be seat 2, which is a loveseat directly facing the tv. So, should I cut the frequencies to match what's happening at 40hz?

 

[Ray R]

Before starting with the BFD, I would suggest moving furniture. It doesn't have to be a complete shuffle. A few inches might make the difference. Play the 40 Hz test tone while sitting in position 2 and then move the soundlevel meter front and back a few inches. If your lucky it will only take that much and you will have the room to move the seat. You could also try moving your main speakers a little as well, but this could effect all listening positions. Moving your sub might also have an effect.

Something else that effects the cancellation frequencies is the subwoofer phase control. My sub has a continually variable phase control as opposed to a 0/180 switch. I know that minor adjustments to the phase control will effect dips, but I was unsuccessful at making this work to my advantage.

Once you done what you can with moving things, then I would take Wayne's suggestion and just EQ for on position. You will drive yourself crazy otherwise.

One last thing, when you start EQing with the BFD notice if the cancellation frequencies change slightly. I was able to get good results, but I did notice a slight shift in the frequencies where I had cancellation. Just wondering if others have seen this happen.

 

[Wayne A. Pflughaupt]

Gary,

Unfortunately, your response is very ragged. The two major problems are the huge 17dB “cliff” above 36Hz and below 40H, and the deep notch centered at 40Hz.

Separately, EQing below 36Hz and above 40Hz is no problem, but blending the two “plateaus” will be problematic – seriously so. Bottom line, Gary, it will be virtually impossible to achieve smooth response with the sub in this location.

The notch suggests some cancellation. If the seat is halfway between the sub and a wall, you will need to move it. A couple of feet will do.

Do you have the sub in a corner? The ragged response suggests that it is not. Corner placement will give the smoothest unequalized performance, with any response peaks easy to EQ.

If the sub is already is in a corner, try a different one. Typically you want the corner with the longest uninterrupted wall length (i.e., no open doorways, etc.) in both directions.

If you end up moving the seat and/or the sub, post some new readings for us.

Regards,

Wayne A. Pflughaupt

 

[Gary Thomas]

Thanks Wayne:

I tried the sub along a side wall, the only other acceptable position, and the response was worse! I'd love to stick it in the corner where it belongs, but it's not aesthetically possible w/ our tv & furniture. (speak to my wife!)

Anyway, the only real problem area, even uncalibrated, is at 40hz from the loveseat seating position. I was able to get an excellent curve from my recliner, but I had a pretty good curve to start with.

Given my placement restriction, I'm not sure if the eq really helps all that much...a few db's here & there.

I had all of the other speakers turned off. Unfortunately, I upped the volume a few db's for my final reading, but you can see the relationship between the frequencies on the before & after readings:

Recliner Before After

16hz 69.5 82.5

20hz 80 86.5

22hz 78.5 89

25hz 78 88.5

28hz 77.5 90.5

32hz 80.5 91

36hz77.5 87

40hz81.5 85.5

45hz76 83.5

50hz74.5 83

56hz74.5 81.5

63hz68.5 74

72hz60 63.5

80hz56.558.5

89hz6257.5

100hz58.556

Loveseat BeforeAfter

16hz68.581

20hz78.582

22hz76.586

25hz7584.5

28hz76.587

32hz8188

36hz78.585.5

40hz60.562

45hz6974

50hz67.572.5

56hz55.561

63hz60.561.5

72hz58.557.5

80hz61.558.5

89hz62.558.5

100hz5956

Couch BeforeAfter

16hz71.583.5

20hz8084.5

22hz78.588

25hz7685

28hz74.585

32hz7684

36hz72.579.5

40hz79.582

45hz7983.5

50hz69.574.5

56hz69.572.5

63hz7071.5

72hz65.564.5

80hz66.563

89hz55.553.5

100hz6057

 

[Wayne A. Pflughaupt]

Gary,

...it's not aesthetically possible w/ our tv & furniture. (speak to my wife!)

If all else fails, try bribery. Diamonds will usually get the desired “compromise.”
I’ll work from the loveseat, since you said that was the main seat.
I don’t know what you changed with the “Before” and “After’’ readings, but the “After” is much better (read “easier to equalize”), despite an extra 5dB of response deviation (24 vs. 31dB). In addition, response below 20Hz drops only 1dB instead of 10dB.
I spent some time applying various EQ adjustments, and here’s what I’ve come up with. It will require four filters. The first two do most of the correction; the second two are “clean-up,” or fine-tuning, and they will overlap the first two.
I’ll refer to the bandwidth settings in octaves or fractions of octaves; I’ll leave it to you to translate that to the BFD’s peculiar bandwidth fractions.
Set the first filter for 1-octave bandwidth, centered at 28Hz and cut 12dB. This will tame the hump at the lower octave. (The sub’s response and output covers a 2-octave range.)
Set the second filter for 1-octave bandwidth, centered at 72Hz and boost 8dB. This will bring up the overall level of the upper octave.
These adjustments should result in the following readings (modified from your “After” readings):
16Hz81dB
20Hz78dB
22Hz80dB
25Hz73.5dB
28Hz74dB
32Hz76dB
36Hz78.5dB
40Hz61dB
45Hz74dB
50Hz74.5dB
56Hz66dB
63Hz69.5dB
72Hz65.5dB
80Hz64.5dB
89Hz62.5dB
100Hz57dB
This reduces the overall response deviation from 31Hz to 24Hz. That may not seem like much, but if you plot both readings on graph paper you can see that overall response is much smoother. For instance, the peaks at 32 and 45Hz were 13dB apart; now they are only 4dB apart.
Now, I imagine some will cry foul at boosting 72dB by a significant 8dB over a large bandwidth, with concerns of overdriving the speaker or amp. The prospect of overdriving the amp has merit. Suffice it to say, Gary, your placement demands some radical EQing. If your amp has “just enough” power with no EQ, you may well run into trouble. Indeed, you may have to upgrade your amp to pull this off. Factually, you can get away with much more clipping in the sub compared to the mains, as low frequency distortion is largely inaudible. But if your clip light is on continually, you have a headroom problem.
Regarding the possibility of overdriving the speaker, that won’t be a problem. You have reduced the lower octave, i.e., the range that demands the most excursion from the driver, a as much as 12dB. This gives you plenty of “headroom” to place greater demands on the upper octave, which requires significantly less excursion.
At this point I would put your sub “to the test,” before tweaking further. Try out your most bass-heavy DVDs to see if these EQ changes will fly with the amp and sub. If you try a CD with music, pick one that has bass extending from high to low notes. You should be able to hear the difference.
If you think your rig is up to it, Gary, you might try these next two filters, to further smooth response.
The first should be no problem. Set a 1/3-octave filter at 28Hz and boost 4dB. This will reduce the hole between 22Hz and 36Hz by 2dB and significantly flatten response over the entire region. It is critical here not to let the filter ”spill over” beyond a 1/3-octave “perimeter.” The outer limits of the filter are 22Hz and 36Hz. Response for those frequencies should not rise at all; if they do, tighten up the filter to say, 1/4 octave.
The new readings in this range should now be:
22Hz80dB (no change)
25Hz76dB
28Hz78Db
32Hz78dB
36Hz78dB (no change)
The second is a bit risky, because we’re boosting the upper octave even more. Set a wide 1- 1/3-octave filter at 100Hz and boost 6dB. This will “shore up” the sagging fringes of the upper octave.
The numbers for this range should now read:
63Hz70.5dB
72Hz68.5dB
80Hz67.5dB
100Hz63dB
Now, the problem here is that we have boosted the top end of the upper octave significantly, as much as 9dB (comparing your initial 80Hz reading [58.5dB] to the last one [67.5dB]). Hopefully your amp can handle this. If not, reduce the last filter enough to keep the amp from clipping too bad.
You will notice that I have not tried to address the deep null at 40Hz. Nulls are usually tough to deal with, although I have seen some people claim to do it. It is very narrow, only 1/6-octave wide. Fortunately, since it is a null it will be much less audible than if it were a peak. However, the adjustments we’ve made should reduce it from a 25db hole to a 13-15dB hole.
If you’re feeling brave, you might try and see what you can do with it. Try setting a 1/6 octave filter at 40Hz and boosting say, 8dB, and then take new readings. If you see a 6-8dB improvement at 40Hz, great. However, if you see only something like 3-4dB improvement, it would be best to give it up. It wouldn’t be worth the loss of headroom.
Keep in mind again the filter limits, which will be 36 and 45Hz. If the readings at those frequencies increase, you will need to tighten up the filter, say to 1/8 octave.
Good luck, and let us know how it goes.
Regards,
Wayne A. Pflughaupt

 

[Gary Thomas]

Wayne: I sincerely appreciate all the effort you put in to helping me! I can't wait to try out the suggestions.

I've thought of another possible spot for the sub. If you check out my HT Pics, I'm thinking of a spot next to the recliner, below a side speaker. What do you think?

As far as amp capacity, I should be ok w/ a Samson 1000. It has 500w. The sub I'm using is a SVS CS-Ultra.

I'll be home Friday & plan to try out the new sub location. If I don't get a better response, I'm planning to try out your eq suggestions.

Thanks again!

 

[Wayne A. Pflughaupt]

Before you do that, Gary, you might want to take a look at my post on this thread (the “epic” near the bottom).
http://www.hometheaterforum.com/htfo...threadid=41399
Regards,
Wayne A. Pflughaupt

 

[Gary Thomas]

So...using a side wall will give me an additional boundry as compared to corner placement.

Given this, with the current location of the sub, the opposite wall is probably 30 feet away, whereas a side wall placement would have the opposite wall only 14 or 15 feet away. This would produce cancellation at 22hz. So...it's probably not a good idea to move the sub to a side wall!

If I leave the sub in it's current spot on the front wall, how close to the wall should I place the sub? I can go from 0 - 2 feet.

 

[Wayne A. Pflughaupt]

If I leave the sub in it's current spot on the front wall, how close to the wall should I place the sub? I can go from 0 - 2 feet.

Cancellation won’t be an issue at two feet, but you might get a little better output with it against the wall.

Regards,

Wayne A. Pflughaupt

 

[Gary Thomas]

OK Wayne...one last time!
Here are the un-equalized measurements from the center (main) listening position:
17.5hz: 89.5
20hz: 97.5
22hz: 95.5
25hz: 96
28hz: 96.5
32hz: 98
36hz: 97.5
40hz: 82.5
45hz: 87
50hz: 86.5
56hz: 79.5
63hz: 75.5
72hz: 71.5
80hz: 78.5
89hz: 80.5
100hz: 76
I took the measurements with only the sub active, all other speakers have been disconnected. Normally, I set my mains & all speakers to "small" and use a 80hz crossover.
What's your eq prescription?
Thanks!!!

 

[Wayne A. Pflughaupt]

Gary,

Thanks for the readings. I’ll study them and post recommendations tomorrow.

Regards,

Wayne A. Pflughaupt

 

[Wayne A. Pflughaupt]

Gary,

Sorry, but these readings are the worst yet.

Comparing all four plots (including the first one from the love seat location, and the second “before” and “after” readings), I think the first readings you give us are the best (despite what I said before). They only had a 17dB response deviation, compared to 26-32(!) dB for the others.

Common to all readings is the null at 40Hz. As I’ve mentioned, it may or may not be possible to address this with equalization. More hopeful is the valley above that point, shown in varying degrees between 45 to 100Hz. It’s too wide to be a null, so it probably can be equalized.

Therefore, Gary, I’m basing the following recommendations on the first readings you posted. Hopefully that is a location you can put your sub permanently.

First I suggest taking new baseline readings, and don’t change the volume setting until equalization is completed. It is not necessary to duplicate the previous readings that were posted. I’ll use the old readings, but indicate the amount of change you should see (i.e., “+4dB”) at each frequency. Thus you will be able to know what changes to expect with your current readings.

Also, I don’t know how it escaped me before, but some of your previous readings were way too low. When you get SPL readings down in the 50s, you are competing with ambient room noise, which can obscure and prevent accurate readings. For best results, I suggest making sure the lowest readings are above 70dB. 75dB would be even better, if that’s possible.

First set a filter for 2/3-octave, centered on 32Hz and cut 6dB. Technically, the outer fringes of the filter should be 25Hz and 40Hz. If the BFD’s filters behave like most, the filter will actually extend a little past the technical fringes with a cut this deep. If so, then you should see a 1dB reduction at 25 and 40Hz, which is what we want. If you see no changes at 25 and 40Hz, increase the bandwidth incrementally until you see a 1dB change at 25 and 40Hz.

After applying this filter you should realize changes in this region similar to this:

22Hz 83.5 (no change)

25Hz 81 (-1dB)

28Hz 79 (-4dB)

32Hz 79 (-6dB)

36Hz 84.5 (-4dB)

40Hz 67.5 (-1dB)

45Hz 76 (no change)

This leaves us with a slight peak at 20Hz. You may or may not want to reduce this. I suggest setting a filter for it, and then decide with listening tests if you want to use it or not (often some boost at the lowest frequencies is desirable).

Set a 1/3-octave filter for 20Hz and cut 2dB. The outer limits of the filter are 17 and 22Hz. Since this is a small cut, you should see a 2dB reduction at 20Hz, and no change at 17 or 22Hz. If so, incrementally narrow the filter until there is no change at 17 or 22Hz.

After EQ, the readings in this region should look like this:

17Hz 76.5 (no change)

20Hz 82.5 (-2dB)

22Hz 83.5 (no change)

Our next problem is bringing up response above 40Hz.

Set a 1-octave filter at 56Hz and boost 8dB The outer limits filter should technically be 40Hz and 80Hz. Again, a filter this deep should “spill over” beyond the limits, which is okay. However, we don’t want more than 2dB increase at 40 and 80Hz. Again, incrementally tighten the bandwidth if necessary. Not knowing exactly the filter “shape” some of the readings might vary a dB or two, as indicated.

After applying this filter, you should see the following changes in this region:

36Hz 84.5 (no change)

40Hz 68.5 –69.5 (+1 or 2dB )

45Hz 78-79(+2 or 3dB)

50Hz 78.5-79dB (+4 or 5dB)

56Hz 75.5 (+8dB)

63Hz72.5-73.5 (+4 or 5dB)

72Hz69.5-70.5 (+2 or 3dB)

80Hz77.5-78.5 (+1 or 2dB )

89Hz72.5 (no change)

I experimented with boosts of 6, 8, and 10dB; on paper it looks like an 8dB boost will give the best results. You might try experimenting with a 10dB boost and see what you think of the results.

There is still a narrow spike at 80Hz we can deal with. Set a 1/3-octave filter at 80hz and cut 4dB.

The outer limits of the filter should be 71 and 89Hz. However, this is not a deep filter, so the outer limits should not change much. In fact we don’t want them to change at all. If they do, tighten the bandwidth a little.

Readings in this range should now read:

72hz69.5-70.5 (no change)

80hz72.5 (-4dB)

89hz72.5 (no change)

This leaves the 40Hz null. You might try addressing it the way I recommended in the post above. Remember, if an 8dB boost gets only gets a 2-3 improvement, it would be best to leave this problem alone.

The new full-range readings (including the 20Hz filter, but not addressing the 40Hz null) should be:

17Hz 76.5dB

20Hz 82.5dB

22Hz 83.5dB

22Hz 83.5dB

25Hz 81dB

28Hz 79dB

32Hz 79dB

36Hz 84.5dB

40Hz 68.5 –69.5dB

45Hz 78-79dB

50Hz 78.5-79dBdB

56Hz 75.5dB

63Hz72.5-73.5dB

72Hz69.5-70.5dB

80hz72.5dB

89hz72.5dB

The response deviation has been reduced from 17 to 14dB. Ignoring the 40dB null, the response deviation has been reduced from 17dB to 11dB.

Hope this helps. Keep us posted on the progress.

Regards,

Wayne A. Pflughaupt