Troble running bass test tones. Need help equalizing. (1 Viewer)

[Wayne A. Pflughaupt]

You mentioned earlier that you were doing this with the speakers off. If you had them on this time, that would account for the changes. When the mains and sub are on, that’s when the phase issues come to light. Readings with both running is what you will want to use for equalizing.

Regards,
Wayne A. Pflughaupt

 

[RuslanM]

Wayne,

The speakers were off. Thats why I'm confused. Here're the results:

2073.581.0
2281.588.0
258994.0
289296.0
31.58992.0
369496.5
4091.594.0
458587.0
508283.5
5682.584.0
6372.574.0
7181.583.0
807879.5
8971.573.0
1007577.0
11178.579.5
12564.565.0
142.557.558.0
1605655.5


You can try the link, if it doesn't work copy and paste the URL.

Thanks,
Ruslan.

 

[RuslanM]

Sorry, here's the link:

http://www.geocities.com/ruslan24/graph3.JPG



Ruslan.

 

[Wayne A. Pflughaupt]

Ruslan,

Sorry to leave you hanging; it’s been one of those weeks.

The peak centered on 30Hz has changed, so we’ll have to change our approach, too. Since the center of it is now a dip, not a peak, it will take at least three filters to address it, not the single one I recommended before.

I’d still start with the 1/3-octave filter I mentioned earlier, only move the center frequency a little, to 31.5Hz. Cut about 6-7dB, not the 15dB I recommended before (we’re trying to make it “bottom out” at the 85dB line).

This will leave two other peaks to deal with, one in the neighborhood of 40Hz and one around 25Hz. You may want to take new readings from 20-50Hz before you do the next two filters to determine exactly what those center frequencies are. Those two filters will need to have bandwidths in the neighborhood of 1/6- or 1/8-octave. The idea is to end up with a reasonably flat line between 20-56Hz, so play with the bandwidth and cut values of those two filters until you get something close to that.

Next up, the hump at 71Hz. Actually, I wouldn’t view this one as a peak, but as two valleys. If we adjust as two valleys we can be simultaneously working towards a house curve.

The 63Hz dip is pretty narrow and can be eliminated with a 1/12-octave filter boosted 7-8 dB. This should put the 56-71Hz area fairly flat and hovering at the 83dB area on the chart.

The next filter should be set at about 90Hz, 1/6-octave bandwidth, and boosted 6-7dB. This should put the 80-100Hz area at about 80dB.

So, we should end up with something of a sloping curve from 111Hz to 45Hz, and flat below that point. Personally I like a house curve that “tops” out at around 30-32Hz, and flat below that point. You may be able to achieve this by tweaking the ~40Hz filter we used earlier, cutting some more, changing the bandwidth, etc.

If you don’t like the results of that, you might try brucek’s trick of overlaying a broad filter to get the house curve, at the link gave in my earlier post.

Regards,
Wayne A. Pflughaupt

 

[RuslanM]

Thanks Wayne,

Your response is highly detailed and informative. I'll play with BFD this weekend and let you know how it goes.

Thanks again,
Ruslan.

 

[RuslanM]

Hi Wayne,

I finally got the time to experiment with BFD. The following two graphs are showing sub only measurements and measurements with the mains on.

This is the best I could do so far. One thing I want to point out is that the dip at 89 did not want to go away. No matter how wide the filter was or how high I was boosting it didn't change. Then I connected the mains and played the tones again. Now the dip got even lower. So, I reversed the phase on the sub and measured again. The graph with mains on got a little bit better in that region, but goes wacky above 125. What would be your next suggestions? What can I do more to improve the graph?

By the way, I must say that I've tried doing it differently than you suggested the first time and couldn't get even close to what I have now following your suggestions on bandwidth and gain/cut values. You can't imagine how much I appreciate your help Wayne. Couldn't do without you.

Anyway, here're the values:

The filters:

1. 31.5 1/3 -6

2. 26.3 1/6 -5

3. 38.3 1/6 -5

4. 63 1/12 +7

5. 90 1/6 +10

Number 5 is probably pointless. I didn't try to measure without it yet.


Mains OFF:

207178.5
227783.5
258085.0
288084.0
31.580.883.8
368284.5
4082.885.3
4580.582.5
5080.882.3
568384.5
6380.582.0
718283.5
807879.5
897071.5
1007476.0
1117778.0
1256767.5
142.55858.5
1605756.5


http://www.geocities.com/ruslan24/graph-sub-only.JPG



Mains ON:

2068.576.0
227783.5
2579.884.8
287983.0
31.582.585.5
368385.5
4083.586.0
458789.0
5080.582.0
5681.583.0
638081.5
7180.582.0
8078.880.3
897677.5
1007981.0
1118586.0
12580.581.0
142.58484.5
1608988.5


http://www.geocities.com/ruslan24/graph-mains-on.JPG

Thanks,
Ruslan

 

[Wayne A. Pflughaupt]

Hi Ruslan,

It’s looking pretty good on paper, and I’ll bet it sounds better, too.

You didn’t say specifically but I’m presuming that the “mains on” graph is with the sub out of phase?

Both graphs look pretty good, actually. The sharp rise above 100Hz with the mains on may be nothing more than a need to re-set the sub’s level now that it has been equalized. Remember, with equalization you’ve lost 10dB of output. That’s very substantial. I’d have to see some data with the sub in phase to before making any “concrete” observations beyond that.

The reason the 89Hz area is not responding is because you have a null there. As you can see from the low point that was at 63Hz, not all response depressions are nulls. A true null, as you’ve seen, will resist equalizing. It would be best to ignore the null rather than try to equalize it. Any filter set to boost expends some amplifier headroom; no reason to waste it if it doesn’t get you anything – i.e. an audible improvement.

Regards,
Wayne A. Pflughaupt

 

[RuslanM]

Hi Wayne,

Yes, the sub sounds much better now. Honestly, I was a bit skeptical on the differences before I started equalizing, but I wanted to give it a shot and I don't regret it.

The graph with the mains on shows the sub with phase set to 180 (reversed). Before I reversed the phase I had a huge hole at 89Hz. After I reversed the phase the hole almost disappeared. But I still have that 5th filter on. I wonder if it has any affect.

I will take new measurements this weekend with mains on and the phase at 0 and 180. I will also try to turn off the boost at 90Hz, see if it has the affect. I've recalibrated the sub with Avia (raised the volume slightly on the sub) but I didn't remeasure after that. I'm not sure how accurate the current measurement is (I don't remember if I remeasured the hole graph or just 89Hz range after I reversed the phase).

Anyway, thanks again and I'll post the new graphs this weekend. Talk to you then.

Thanks,
Ruslan.