DSP1124p EQ Boost?

[MichiganMan]

I played with REW and used it's peak detect feature to automatically calculate the filters for a DSP1124p.

I noticed when applied the filters created some dips/valleys.

Would it be Ok to add filters with boost/gain to smooth out the dips as long as the boost does not exceed the original signal db?

[brucek]

Would it be Ok to add filters with boost/gain to smooth out the dips as long as the boost does not exceed the original signal db?

Generally boosting is not advised, as dips are caused by phase cancellations, where the sound reaching the measurement point is a combination of an original direct soundwave and a reflected sound that is 180 degrees out of phase at the dip (one half wavelength). When you add a gain filter at that dip frequency, not only does the direct sound increase by the number of dB of the filter, but unfortunately the 180 degree out of phase signal also applies an equal and opposite signal to counteract. The result is that your dip is still there and you have wasted the gain you've thrown at it.

Sometimes you can add a small amount of gain and remove a dip, but if you find the dip isn't responding, don't bother.

brucek

[Wayne A. Pflughaupt]

MichiganMan wrote:

I played with REW and used it's peak detect feature to automatically calculate the filters for a DSP1124p.

I noticed when applied the filters created some dips/valleys.

Would it be Ok to add filters with boost/gain to smooth out the dips as long as the boost does not exceed the original signal db?

Valleys from phase cancellations are one thing; valleys created by EQ filters are another. Basically, response shouldn't be worse after using REW's recommended filters. If so, something's wrong. I suggest posting some before and after graphs for us to look at.

Regards,
Wayne

[MichiganMan]

Gentlemen,

I played with the EQ filters offline and this is what I've come up with.

I placed two filter with boost at 20 and 74hz. The 20hz filter prevents early roll-off, and the 74 was to correct a dip.

In neither case am I boosting the corrected signal above the original sweep. I'm using one filter to correct or smooth out the effect of another filter.

I haven't actually hooked up my DSP1124p so I don't know the real world results yet. Is there anything wrong with the way I'm applying the filters in this case?

[brucek]

Is there anything wrong with the way I'm applying the filters in this case?

Filters 1 and 2 are unnecessary - one will suffice.

But that's not the bigger issue. If this is a measure of sub + mains, then you can't apply filters to the sub alone and expect the results to affect the mains signal that doesn't run through the BFD. You must first equalize the sub, paying close attention to the signal below the crossover, as the signal above the crossover will become increasingly overwhelmed by the mains signal as the frequency increases.

Once the sub is equalized, then add the mains and remeasure. Then use the subs phase control to smooth the crossover transition.

brucek

[Wayne A. Pflughaupt]

MichiganMan wrote:

Is there anything wrong with the way I'm applying the filters in this case?

Actually, yes. You’re starting with your level way above the Target curve, so you’re using the EQ as a de facto gain control. Is this the level you got after calibrating for 75 dB? Either reduce the sub level, or move the Target up to ~83-84 dB before you start equalizing. You’ll see then that you won’t need more than a few filters. You might want to take a look at the first post of my minimal EQ article – you’ll find a link to it in my signature.

Regards,
Wayne

[MichiganMan]

Wayne A. Pflughaupt wrote:

Is this the level you got after calibrating for 75 dB?

Wayne,

Yes but I'm using the "main speakers" to calibrate the levels. I used a RS meter to equalize the levels for my 5.1 speakers including the sub. I think the processor broadcasts pink noise.

I then calibrated REW levels using the "main speakers" and RS meter setting to 75db C weighted.

What I notice is that I get a high level at the sub frequencies and occasionally a clipping notice. If I calibrate using the "subwoofer" I get low main levels on a full sweep and the warning/notice to increase processor levels.

[Wayne A. Pflughaupt]

For taking subwoofer measurements, you should use the subwoofer itself to calibrate the levels. The calibration test signal REW generates has primarily lower frequency content that will be mostly below the crossover frequency of the main speakers. IOW, its level will be fairly low in the main speakers, so you'll have to turn the receiver's volume up to get a 75 dB reading. Naturally, when you go to measure the sub, its level will be much higher, which is why you're getting the "clipping" indicator.

Regards,
Wayne

[bcharlow]

Brucek,

If you can't practically (or shouldn't) smooth out a serious dip using EQ, then what can you do for that dip?

I've got my BFD doing a terrific job with the peaks and the x-over set (24 db/oct published figure) at minimum of 40Hz. Yet I've got a serious dip at about 110 Hz for a few Hz around that point. I realize that level is probably room interaction with both mains and sub and is probably a null at listening positino.

Is there any hope for curing that one dip????

[Wayne A. Pflughaupt]

Well, I assume 110 Hz is above the crossover point, and the BFD is only affecting frequencies below the crossover point. So the answer is "probably not." Your only option is a variable phase control on the sub or time alignment from the receiver. However, if I read your post right and your crossover frequency is 40 Hz, even that won't help. Typically phase/time alignemt issues have the greatest detrimental effect in the octave above or below the crossover point.

Regards,
Wayne

[brucek]

then what can you do for that dip?

Move the sub, or live with the dip.... Often dips at a shared frequency are easy to remove with phasing.

brucek

[bcharlow]

Brucek & Wayne,

I have played with placement as far as WAF allows (my 20' ICs are more flexible than my spouse on this issue!) Unfortunately, I seem to be stuck with that dip. In fact, I think it's there in the mains too, so that's definitely a room issue.

FYI, I'm the odd man out here, as I'm not HT, but stereo for music only. So I don't need, can't afford a receiver with all the timing DSP that a modern HT receiver has. I'm using a NAD C740 receiver, strictly a 2 channel deal. My only DSP is in the BFD unit, and it will probably stay that way for a long time. My speakers are Von Schweikert VR1 monitors, which go deeper and better that way than their specs: down to 40 Hz. The sub is a cheap KLH 120W 10" model designed for entry level HT, which goes from about 28 - 120 Hz usably. I turn that down to x-over at 40 Hz (minimum allowed), and that's its best setting by test; with it's 24db/oct slope it should be delivering no more than 25db at the 110 point (by crude estimate). That's certainly below audible levels one way or the other. So the issue must be with my mains interacting with the room and those mains are not movable under the circumstances.

I cannot tell you what a sharp dip at 110Hz actually sounds like. Since I know that's in the range of many instruments and even the female voice, it could contribute to less than accurate sounding vocals for example. But that's just a surmise, not a measured fact. I'm guessing that I'm missing something I'd like if I had it....

My sub has only a 0 and 180 degree phase control. It's entirely possible that phase contributes, but if my above analysis is correct, the sub isn't the issue here really. In fact, I'm using the BFD to control, among other peaks, two large peaks from the sub at 100 and 120 Hz, so effectively I've really tuned the sub out of this equation when I think of it.

[Wayne A. Pflughaupt]

bcharlow wrote:

The sub is a cheap KLH 120W 10" model designed for entry level HT, which goes from about 28 - 120 Hz usably. I turn that down to x-over at 40 Hz (minimum allowed),

I have played with placement as far as WAF allows (my 20' ICs are more flexible than my spouse on this issue!) Unfortunately, I seem to be stuck with that dip. In fact, I think it's there in the mains too, so that's definitely a room issue.

The issue is with the main speakers, not the sub, so there’s nothing you can do about it with the BFD. Of course, adjusting the mains’ placement can help (not sure from what you said if your WAF placement issues were only for the sub or beyond that); that’s your only option short of equalizing the main speakers themselves.

I cannot tell you what a sharp dip at 110Hz actually sounds like. Since I know that's in the range of many instruments and even the female voice, it could contribute to less than accurate sounding vocals for example. But that's just a surmise, not a measured fact. I'm guessing that I'm missing something I'd like if I had it....

If it’s an ultra narrow dip, then it’s not worth worrying about. Those are typically barely audible, if at all. With program material there is plenty of other information from harmonics, overtones etc. that will essentially mask it.

In fact, I'm using the BFD to control, among other peaks, two large peaks from the sub at 100 and 120 Hz, so effectively I've really tuned the sub out of this equation when I think of it.

Hmm. I have to wonder how well the crossover on this sub works. If it’s really 24 dB/octave, and was set for 40 Hz, a spike in response at 100 or 120 Hz would have to be something like 30-40 dB in order to be an audible or measurable issue. Regardless, 100 Hz and 120 Hz are close enough to 110 Hz. You might try letting one or the other of those peaks come back through enough to reduce the 110 Hz dip.

Regards,
Wayne