Home Theater Forum and Systems banner

1 - 16 of 16 Posts

·
Registered
Joined
·
6 Posts
Discussion Starter · #1 ·
First time to use REW. Setup is EMU0404 USB, DSP1124P to power amp driving a pair of HSU TN1220.
I have attached the measurement before and after EQ. Measurement was done directly from EMU0404. One channel (R) out is fed to both channels of DSP1124P, then to a stereo power AMP driving 1 sub on each channel.
My focus is to flatten out 20-80Hz. 80Hz THX crossover should be done by AV receiver, so I'm not asking the DSP1124P to do the 80 Hz roll off. The peaks I can understand, but the dip at around 65Hz need more than 14dB of gain to compensate, which doesn't seem quite right to me. the two subs are in the two opposite corners of the room along the shorter wall. I limit the gain to 12dB in the "after" curve.
Any suggestion on what could possibly cause this dip? Should the 2 channels have different EQ?

Thanks.
 

Attachments

·
Registered
Joined
·
2,604 Posts
Boosting a dip can often be fruitless, as they are often caused by cancelling reflections, whether modal or not. Boosting boosts both the main signal and the reflection, netting no achievement other than eating headroom. Not written in stone, but something to watch out for.
So, try to isolate the problem: do you get a dip from either sub by itself or only when the 2 are combined? If it's only with the two, then there's a phase cancellation between them. Otherwise it's acoustics. Where you go from there may depend on the room layout and whether there's a dip in both or only 1.
 

·
Super Moderator
Joined
·
9,279 Posts

Welcome to the Forum, Joe!

For starters, you might want to extend your graph down to 15 Hz. It looks like your subs are putting out below 20 Hz. Also, we prefer graphs with a vertical axis of 45-105 dB.


My focus is to flatten out 20-80Hz. 80Hz THX crossover should be done by AV receiver, so I'm not asking the DSP1124P to do the 80 Hz roll off. The peaks I can understand, but the dip at around 65Hz need more than 14dB of gain to compensate, which doesn't seem quite right to me.
I’d recommend engaging the crossover and re-equalize. With the upper bass frequencies rolled out, you’ll find you won’t need nearly that much boost in the 65 Hz region.


Any suggestion on what could possibly cause this dip? Should the 2 channels have different EQ?
As Greg noted, the dip can be caused by any number of things, especially if your subs are not co-located or if separated, are not in locations with symmetrical boundaries. However, it’s typically best to equalize the sub’s total output, not individual. With the latter, you’ll probably find that painstaking separate equalizing will be trashed once they are both turned on together.

Regards,
Wayne
 

·
Super Moderator
Joined
·
9,279 Posts

Boosting a dip can often be fruitless, as they are often caused by cancelling reflections, whether modal or not. Boosting boosts both the main signal and the reflection, netting no achievement other than eating headroom. Not written in stone, but something to watch out for.
This is the popular wisdom, but in reality I can probably count on one hand the number of cases I’ve seen over the years where someone truly couldn’t eliminate a depression with EQ. A while back I dug up this article which helps explain why. This quote sums things up nicely:

These nulls are related not only to the distance from the rear wall, but also from the other walls, the floor, and the ceiling. So in order to create a deep null a precise balance is needed, and that balance is easily disrupted by the many other reflections bouncing around the room.

If I understand that properly, sound is reflecting from every boundary in the room, not just the one that might be responsible for the signal to be 180˚ out of phase. The other reflections exhibit varying degrees of phase, more or less than 180˚. This gives an equalizer plenty of other, non-canceled signal to work with. Plus, the EQ introduces some phase changes itself. Most likely this is why you’ll seldom see a case where a depression won’t respond to EQ.

From what I’ve seen over the years, the only depressions that didn’t respond to equalization were deep and narrow, like the ones in the graphs below.




Here's that last graph, after equalization. As you can see, the nulls didn't budge.


Graph with nulls after EQ.jpg



Regards,
Wayne
 

·
Registered
Joined
·
2,604 Posts
Well, on the one hand, I don't buy the theory of other surfaces, as every reflection should change by the same amount, including phase shifts. The only part of that I buy is with more gain you cna get an extra reflection that wasn't getting to the mic before, but this effect should be minimal.

That being said, I will ammend my statement to exclude broad dips. I believe narrow ones are usually due to modes and other reflections, which should be expected not to respond too well to boost, but broad dips may be other phenomena, which would not be subject to the same limitations... I wonder what those would be...
 

·
Registered
Joined
·
66 Posts
Hi Joe,

I'll leave the technical stuff to the other guys, but wanted to say hi to another TN1220 user. :wave: I also have a (single) TN1220HO. Your equalized results look pretty sweet to me. I wasn't able to get mine that flat. It looks to me like only a 5 - 6 db difference at your 68 hz dip ?

Are you going to use a house curve ?

Regards,

Rog
 

·
Registered
Joined
·
6 Posts
Discussion Starter · #7 · (Edited)
Hi Wayne,

Thanks for the comment. I have attached the revised graph with the suggested range. There is indeed useful output down to 10Hz. Compare to the graphs that you post, I'm glad that I don't have a notch cancellation as you have indicated. The dip respond to the boost very well. I am just concerned about such a large boost could cause headroom issue. As a rule of thumb, how much boost could be apply? Another way I was thinking is to reduce the level of the target curve and adjust the filter that way. Then increase the AV receiver subwoofer output to compensate the level. The objective is to maintain headroom at both the A/D and D/A stage in the DSP1124.
I'm a bit confused by the suggestion that I don't need that much boost with crossover applied. It is true that the output at 65Hz will be reduced when crossover is applied. However, if the response is reduced by, say 6dB before applying crossover, it would still be down 6dB from the targeted response after crossover is applied, though at a lower level.
On top of that, I haven't figure out how to do measurement thru the AV receiver since I have set up the main speaker to be "large". So if I fed a 2-channel signal into it, all low frequency will just go to the main speaker instead of the sub. What I did was I do the measure and apply the EQ, then recalibrate all channel levels again at the AV receiver.

regards,
Joe
 

Attachments

·
Registered
Joined
·
6 Posts
Discussion Starter · #8 ·
Greg,

Thanks for the suggestion. The 65Hz dip of concern respond to boost fairly linearly. So I don't think it is a simple cancellation or null point at the measurement location. I will try to move the sub to redo measurement, but due to WAF, the sub location will be very much limited.

regards,
Joe
 

·
Registered
Joined
·
6 Posts
Discussion Starter · #9 ·
Hi Rog,

I actually have the TN1220 sitting around for a while before "activating" it recently; very much thanks to this forum. Otherwise I would have no way nor idea how to adjust the sub other than using the AV receiver "calibrate" function, which only adjust level.

I am still not entirely sure about "house curve" and what is the appropriate house curve. With the current post EQ response, I already think that there is more than sufficient bass. Though do note that I have a "large" main speaker. So only LFE and center/surround bass end up in the sub.

One annoying problem with the TN1220 is that it vibrates on the floor and give out a buzzing noise. I thought a 3-point contact should form a stable "plane" but the whole thing buzz when bass level is moderately high. It will stop buzzing when I apply some down force on the sub. It looks like the sub is not heavy enough. Do you experience the same problem? I am thinking about adding some weight to the bottom. Since you can't buy leadshot any more, I am wondering if there are better alternatives.

regards,
Joe
 

·
Super Moderator
Joined
·
9,279 Posts

The dip respond to the boost very well. I am just concerned about such a large boost could cause headroom issue. As a rule of thumb, how much boost could be apply?
The only rule of thumb is, how much headroom to you have? If you have a 50-watt 8” sub in a 8000 cu. ft. room, then you probably can’t do any kind of equalizing (cut-only included). If you have sixteen 18” subs powered by 20,000 watts of power in that same room, you can probably boost to your heart’s content.


Then increase the AV receiver subwoofer output to compensate the level. The objective is to maintain headroom at both the A/D and D/A stage in the DSP1124.
That’s not really a concern. As long as you have a decent level on the BFD meters, you’re good. Better to leave some headroom than to max out the BFD’s input levels.


I'm a bit confused by the suggestion that I don't need that much boost with crossover applied.
Usually people are equalizing to the Target Curve, and @ 60 Hz the crossover is starting the roll out the bass (see the graphs I posted). So less boost would be required to get response up to the Target.

Regards,
Wayne
 

·
Registered
Joined
·
2,604 Posts
Greg,

Thanks for the suggestion. The 65Hz dip of concern respond to boost fairly linearly. So I don't think it is a simple cancellation or null point at the measurement location. I will try to move the sub to redo measurement, but due to WAF, the sub location will be very much limited.

regards,
Joe
Well, I'd say the proof of Wayne's comments is in your plots... I'm always a fan of trying it empirically as long as we know the dangers we're looking out for!
That being said, I'd also say that if you were going to run into headroom problems with a boost, applying cuts and boosting the main level could run you into the same headroom problem. Similarly, if you lower the target curve, chances are you'll end up increasing the volume on the receiver after the fact, netting the same...

I'm not saying you're going to have a headroom issue...
 

·
Registered
Joined
·
66 Posts
I am still not entirely sure about "house curve" and what is the appropriate house curve. With the current post EQ response, I already think that there is more than sufficient bass. Though do note that I have a "large" main speaker. So only LFE and center/surround bass end up in the sub.
Wayne has several good write ups about house curves and how to determine starting and ending points. What I took away from these articles was that "flat" response may not be best for listening because of the way we hear low frequencies. We actually need higher levels for some low frequencies. I do think there's a certain amount of subjectiveness to it, since I've read others describe different curves for different types of listening (movies versus music).

I'm not sure how you'd do that, though with the mains in large mode. I guess the bottom line is that if you like how it sounds that's good :bigsmile:

One annoying problem with the TN1220 is that it vibrates on the floor and give out a buzzing noise. I thought a 3-point contact should form a stable "plane" but the whole thing buzz when bass level is moderately high. It will stop buzzing when I apply some down force on the sub. It looks like the sub is not heavy enough. Do you experience the same problem? I am thinking about adding some weight to the bottom. Since you can't buy leadshot any more, I am wondering if there are better alternatives.
That's interesting. My sub is sitting on carpeting, so I don't notice that problem. Are yours on carpet or hard surface ?

Rog
 

·
Registered
Joined
·
66 Posts
Joe,

Looks like the sub dude link may be the tickets to fixing this. The sub is probably vibrating on the hardwood floor on the spikes, which doesn't happen with carpeting. The sub dude thingy might fix that.

Another option would be to lay the sub sideways, although that may not be possible for your installation. (Less expensive than $50 x 2 though ) :sn:

Rog
 
1 - 16 of 16 Posts
Top