Home Theater Forum and Systems banner

1 - 13 of 13 Posts

·
Registered
Joined
·
35 Posts
Discussion Starter · #1 ·
Hi. I am interested to know what is the best way to go by optimizing my subwoofer. I had used REW, and the recent V5.0 had this nice room modes calculation. How should i use that information to help my bass response?

What I had been doing so far is to bascially match response to the target graph, and then click "optimize" and input the filter values into BFD to get a nice FR graph. By doing that, I notice that my FR looks quite good, but then my waterfall graph does not change much. I believe I may have missed something?

My room mode (from REW):

24.9Hz - 79.6( Peak dB) - 303 (T60)
33.4Hz - 69.3 (Peak dB) - 674 (T60)
36.3Hz - 69.4 (Peak dB) - 722 (T60)

My understanding is that with that low frequency, any reasonable size bass trap will not be useful. So, what's the best way to handle those?

I have read the help file in REW and in the EQ filter section, there is a MODAL field (basically a PEQ setting) that I can select the target T60 values. However, what values should I choose for the Gain?
When I change the Gain, I can see that the mode T60 values and Filter T60 values decreased. I am not sure what those values mean and how I should optimize. Anyone can help what I should do and what those settings mean?

Thanks a lot.
 

·
Super Moderator
Joined
·
9,280 Posts

T-60 isn’t viable for low frequencies in small rooms. By most accounts it’s not viable for the upper frequencies in small rooms either.

That said:
When I change the Gain, I can see that the mode T60 values and Filter T60 values decreased.
Since you’re trying to use RT60 I assume you have some idea of what it is and how it “works” (pardon the poor terminology). It should be obvious that any time-based measurement such as RT60 or waterfalls is naturally going to change relative to gain. Increase the signal level from a reference point and it takes longer for it to decay down to ambient noise, and just the opposite is true for decreasing the signal level.


What I had been doing so far is to bascially match response to the target graph, and then click "optimize" and input the filter values into BFD to get a nice FR graph. By doing that, I notice that my FR looks quite good, but then my waterfall graph does not change much. I believe I may have missed something?

My room mode (from REW):

24.9Hz - 79.6( Peak dB) - 303 (T60)
33.4Hz - 69.3 (Peak dB) - 674 (T60)
36.3Hz - 69.4 (Peak dB) - 722 (T60)
Can’t tell you anything specific to your situation without seeing your graphs, but not every peak in response is a room mode, and not every room exhibits a mode at the measurement location. In either case, you might not see much difference in a waterfall before and after EQ.



My understanding is that with that low frequency, any reasonable size bass trap will not be useful. So, what's the best way to handle those?
There’s nothing you can really do about low frequency decay times, unless you are able to put a lot of bass traps in your room. Parametric EQ can help with a low freq room mode, but that’s about it. And only to the extent that it will bring its decay times back in line with what’s the norm for the room.

Regards,
Wayne
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #3 ·
Thanks Wayne!
My understanding of RT60 is it measured the time it takes for a signal to decay 60dB... I thought it is due to room acoustics mostly, and I do not understand why a higher gain signal will decay slower than a lower gain signal. Can you help explain a bit?

I have attached my frequency response and waterfall here. I used your hard-knee taget graph from 30 to 80Hz (using 8dB instead of 6dB though). And you are correct that my before/after EQ waterfall almost looks identical.

With that said, can I go and say that as long as I optimize per the FR target I want, that's pretty much all I can do (given it is impractical to put a lot of bass traps in my room)?

I heard people saying that you need a fast decay time to get the bass to act "quick". How should I go about doing this?

Thanks a lot for your help.
 

Attachments

·
Registered
Joined
·
35 Posts
Discussion Starter · #4 ·
BTW, please comment on how this looks and what improvements or things I need to look at. I am now using DSP1124 and set 5 filters to achieve this:

26.3Hz BW 6/60, Gain -4
30.2Hz, BW 15/60, Gain 7
32.35Hz, BW 5/60, Gain -6
50Hz, BW 15/60, Gain -4
80Hz, BW 9/60, Gain -7

My crossover is set at 80Hz.

This is with 2 SVS PC12+. I am also considering selling those and get a JTR Orbit shifter. I believe my <20Hz should improve a lot, any comments?
 

·
Super Moderator
Joined
·
9,280 Posts

My understanding of RT60 is it measured the time it takes for a signal to decay 60dB... I thought it is due to room acoustics mostly, and I do not understand why a higher gain signal will decay slower than a lower gain signal. Can you help explain a bit?
I didn’t say a higher-level signal would decay slower. I said it would take longer to decay down to the room’s ambient noise level. That may sound like “splitting hairs,” but it isn’t. Really, it isn’t. :)

As you probably know, RT60 is about calculating a signal's rate of decay; i.e. the time it takes for a signal to drop 60 dB from its original level. For something resembling a real-world example, let’s say you generate a 1 kHz test signal at a specified SPL – let’s say 85 dB - in a room that has a concrete floor and cinderblock walls. As you can imagine, the sound is going to bounce all over the place and take a while to fade away.

Now let’s throw down some wall-to-wall carpet in that room, and hang heavy draperies on all the walls. After applying these absorptive treatments, our 85-dB, 1-kHz signal is going to disappear really quick, isn’t it? What has happened? The treatments absorb the signal and prevent it from bouncing around the room. So we have increased the signal’s rate of decay. In other words, it fades to “nothing” much faster than it did originally.


Rate of Decay reduced.JPG
Signal Rate of Decay


Now let’s look at decay time relative to the signal level. The thing to keep in mind is that in the real world, our rooms have a certain level of ambient noise. The noise floor in a soundproofed recording studio might be as low as 25 dB; in our living rooms it’s typically more like 35-40 dB.

So, let’s revisit our theoretical 85 dB, 1 kHz signal. In a quiet living room, the signal only falls 45-50 dB before it’s buried in the room’s noise floor. What happens if we increase the signal level to 105 dB? Now it has to fall 65-70 dB before it fades into the noise floor.

Well naturally, a signal that only has to fade 45 dB before it hits the noise floor is going to “get there” sooner than one that has to drop 65 dB, even if the RT60 is the same for both. So the time it takes the signal to fade is not the same as its rate of decay (RT60).


Gain and Decay Time reduced.JPG
Signal Gain and Decay Time


Regards,
Wayne
 

·
Super Moderator
Joined
·
9,280 Posts

With that said, can I go and say that as long as I optimize per the FR target I want, that's pretty much all I can do (given it is impractical to put a lot of bass traps in my room)?
That’s pretty much correct. Fortunately, smoothing response (reducing the “distance” between the peaks and depressions) is sufficient to improve sound quality.


I heard people saying that you need a fast decay time to get the bass to act "quick". How should I go about doing this?
That pretty much requires bass traps. Real Trap’s Ethan Winer claims his bass sounds tight and fabulous, but it took something like 35 bass traps in his living room. :eek:

Regards,
Wayne
 

·
Premium Member
Joined
·
6,711 Posts
Hi. I am interested to know what is the best way to go by optimizing my subwoofer. I had used REW, and the recent V5.0 had this nice room modes calculation. How should i use that information to help my bass response?

What I had been doing so far is to bascially match response to the target graph, and then click "optimize" and input the filter values into BFD to get a nice FR graph. By doing that, I notice that my FR looks quite good, but then my waterfall graph does not change much. I believe I may have missed something?

My room mode (from REW):

24.9Hz - 79.6( Peak dB) - 303 (T60)
33.4Hz - 69.3 (Peak dB) - 674 (T60)
36.3Hz - 69.4 (Peak dB) - 722 (T60)

My understanding is that with that low frequency, any reasonable size bass trap will not be useful. So, what's the best way to handle those?

I have read the help file in REW and in the EQ filter section, there is a MODAL field (basically a PEQ setting) that I can select the target T60 values. However, what values should I choose for the Gain?
When I change the Gain, I can see that the mode T60 values and Filter T60 values decreased. I am not sure what those values mean and how I should optimize. Anyone can help what I should do and what those settings mean?
The T60 decay times REW determines from its modal analysis are the decay times of the specific modes identified, not a room RT60 figure (which is the overall decay time within a 1/3 or 1 octave band). They are valid in any size of room. Using the "Modal" filter type configures a parametric EQ filter so that filters zeroes lie on the poles of the mode (the poles are the resonant parts of the mode's response) - more precisely, it places the zeroes as close as the resolution of the equaliser's settings allow. As the gain is adjusted it varies the filter bandwidth setting to keep the zeroes as close as possible to the mode's poles, to provide the best counter to the mode's effect. The "ModeT60" figure tells you where the filter's zeroes have ended up with the current equaliser setting (ie. the filter would be ideally matched to a mode with that 60dB decay time), the "Filter T60" time is the decay of the filter itself - it shows the location of the filter's own poles. The greater the filter cut, the lower the decay time of its own poles - on the other hand, a filter with boost has poles with decay times that get longer the greater the boost applied.
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #8 ·
I didn’t say a higher-level signal would decay slower. I said it would take longer to decay down to the room’s ambient noise level. That may sound like “splitting hairs,” but it isn’t. Really, it isn’t. :)

As you know probably, RT60 is about calculating the “rate of decay.” For something resembling a real-world example, let’s say you generate a 1 kHz test signal at a specified SPL level – let’s say 85 dB - in a room that has a concrete floor and cinderblock walls. As you can imagine, the sound is going to bounce all over the place and take a while to fade away.

Now let’s throw down some wall-to-wall carpet in that room, and hang heavy draperies on all the walls. After applying these absorptive treatments, our 85-dB, 1-kHz signal is going to disappear really quick, isn’t it? What has happened? The treatments absorb the signal and prevent it from bouncing around the room. So we have increased the signal’s rate of decay. In other words, it fades to “nothing” much faster than it did originally.


View attachment 23686
Signal Rate of Decay


Now let’s look at signal levels. As you know, RT60 measures the time it takes the signal to decay 60 dB. The thing to keep in mind is that in the real world, our rooms have ambient noise floors. The noise floor in a soundproofed recording studio might be as low as 25 dB; in our living rooms it’s typically more like 35-40 dB.

So, let’s revisit our theoretical 85 dB, 1 kHz signal. In a quiet living room, the signal only falls 45-50 dB before it’s buried in the room’s noise floor. What happens if we increase the signal level to 105 dB? Now it has to fall 65-70 dB before it fades into the noise floor.

Well naturally, a signal that only has to fade 45 dB before it hits the noise floor is going to “get there” sooner than one that has to drop 65 dB, even if the RT60 is the same for both. So the time it takes the signal to fade is not the same as its rate of decay (RT60).


View attachment 23687
Signal Gain and Decay Time


Regards,
Wayne
Thanks Wayne! That makes perfect sense now!
So, do you see anything in the FR response and EQ setting that I should take a closer look? Does the response looks smooth and good to you?
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #9 ·
The T60 decay times REW determines from its modal analysis are the decay times of the specific modes identified, not a room RT60 figure (which is the overall decay time within a 1/3 or 1 octave band). They are valid in any size of room. Using the "Modal" filter type configures a parametric EQ filter so that filters zeroes lie on the poles of the mode (the poles are the resonant parts of the mode's response) - more precisely, it places the zeroes as close as the resolution of the equaliser's settings allow. As the gain is adjusted it varies the filter bandwidth setting to keep the zeroes as close as possible to the mode's poles, to provide the best counter to the mode's effect. The "ModeT60" figure tells you where the filter's zeroes have ended up with the current equaliser setting (ie. the filter would be ideally matched to a mode with that 60dB decay time), the "Filter T60" time is the decay of the filter itself - it shows the location of the filter's own poles. The greater the filter cut, the lower the decay time of its own poles - on the other hand, a filter with boost has poles with decay times that get longer the greater the boost applied.
Thanks John! So, the Filter T60 values are the "additional" decay time that are caused by the filter itself.. is that correct?
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #11 ·
Response looks good, but the house curve looks more like 10 dB, not 8. Of course, it’s up to you to determine if that’s what your room needs.

Regards,
Wayne
Yeah.. I notice that too. Actually, the target curve had even more differences from 30-80..much like 12. However, since my main take over around 60-70 and adds to it, the difference now becomes 10.
I think the 24/dB octave setting I have in my subwoofer setting "adds" to the target curve.
SHould I set the setting to "none" in speaker setting?

Actually, I have a question on what to put as "Target settings". Should I use "subwoofer" or "Full Range"?
I assume since my "base graph" is taken with speakers+sub, I should pick Full Range, is that right?
However, since my EQ only goes to sub, and at around 70-90, the mains started to have a pretty big effect, EQ around there does not help much.

Also, I use the SPL meter to set my mains to 75dB with the receiver test tones before. However, as you can see from the graph, the level is somewhat lower (measured to about 70dB only..) in the REW curve. I do not know where the 5dB goes... any hint? Should I trust my receiver test-tones + SPL meter, or the REW curve? 5 more dB differences do sound a lot to me.

BTW, where can I get some free sine-wave test-tones so maybe I can check that way?
 

·
Super Moderator
Joined
·
9,280 Posts

Actually, I have a question on what to put as "Target settings". Should I use "subwoofer" or "Full Range"?
You typically only use the Full Range setting when taking full range measurements.


Should I trust my receiver test-tones + SPL meter, or the REW curve? 5 more dB differences do sound a lot to me.
It happens; you can always re-adjust the Target before equalizing. It’s under the “Target” icon to the left of the screen.


BTW, where can I get some free sine-wave test-tones so maybe I can check that way?
There’s a link to some sine waves on Pg. 2 of our BFD Guide. :T

Regards,
Wayne
 

·
Premium Member
Joined
·
6,711 Posts
Thanks John! So, the Filter T60 values are the "additional" decay time that are caused by the filter itself.. is that correct?
Not really "additional" decay time, the decay of the mode is removed (if the filter exactly matches its T60 time) and replaced by the decay time of the filter.
 
1 - 13 of 13 Posts
Top