Now what? Please give comments

[angryht]

The attached files are my room layout and the center speaker REW plot. BFD or no BFD. The location of the sub is between the left speaker and the center speaker. That is what seemed to give me the best response. I put in a house curve of +5 at 30 Hz and 0 at 80 Hz. My crossover is at 100Hz and I am running bass management with an Outlaw Audio ICBM. Please give comments and be brutal.

Thanks - hope I did this right.

[brucek]

center speaker REW plot

Can you explain what you mean by this?

brucek

[angryht]

brucek wrote:

Can you explain what you mean by this?

brucek

angryht wrote:

center speaker REW plot

The graph is the frequency response of the center speaker using Room Eq. Wizard. This includes the subwoofer and the center speaker.

[Otto]

Hi Greg,

Well, we normally measure our sub response by itself -- turn off all other speakers either by cutting power to their amp, or by simply removing the speaker cables from the speakers (or at the amp).

Also, the fact that you were able to measure the center response tells me that you were using some type of audio processing mode -- perhaps DPLII or "multi channel stereo" mode. Either way, disable any of those modes and just measure the sub in a stereo mode with crossovers engaged. REW is a mono signal that's meant to be handled in a simple fashion.

What sub are you using? You're getting a very solid response to 30 Hz, and then it drops like a rock. Is that expected or is the ICBM somehow imposing that?

If you're asking whether or not you should get a BFD, I would answer that I could go either way on that. Your response looks very good in general, and most of us would be happy to have a natural response like that. On the other hand, I love the BFD, and would not be without it or its equivalent.

In general, you're on the right track. It looks like you have REW set up properly, and you're able to take measurements. Keep going!

[angryht]

Thanks, Otto. I have a seperate preamp, an Outlaw Audio ICBM (I just use it for bass management because I have an older preamp/pro: http://www.outlawaudio.com/products/icbm.html, and a 5 channel amp. All of the channels from the preamp go into the ICBM then to the amplifier (and the sub) and out to the speakers. Here is the setup for the signal from the computer (output). I ran the output cable from the computer to the left channel of the inputs of the preamp. Then I ran the left output from the preamp to the channel that I am testing (input for the ICBM) For the graph I show in the first post, it is the center channel which includes the sub.

I have a Def Tech Super Cube II subwoofer. When I did the sub alone (as you suggest by just unplugging the speaker wires) I get a pretty big hump from about 30to 50 Hz (at least I think it's from about 30to 50 Hz - my files are on my home computer). I will go ahead and post that later today. When I connect the speaker it settles down like my graph that I have posted.

Is the drop from 30 Hz unusually steep?? When I tried adjusting the phase the dip from 50 to 60 Hz just got deeper.

[brucek]

Is the drop from 30 Hz unusually steep?? When I tried adjusting the phase the dip from 50 to 60 Hz just got deeper.

The phase will only affect areas that are shared by two speakers. So, the frequencies near a crossover, would enjoy the largest effect.

The quick drop below 30Hz is a bit sharper than a natural rolloff, so it's likely cause by the room.

brucek

[angryht]

brucek wrote:

The quick drop below 30Hz is a bit sharper than a natural rolloff, so it's likely cause by the room.

brucek

Maybe if I move the sub around I could get a better response below 30 Hz??

[Wayne A. Pflughaupt]

Maybe if I move the sub around I could get a better response below 30 Hz??

The location of the sub is between the left speaker and the center speaker.

I certainly would! I tried dead-center-between-two-walls one time, and the response was abysmal. Maybe give it a try in a corner that doesn't have any nearby openings.

Regards,
Wayne

[angryht]

Thanks for the suggestions, Wayne. I will try the corner placement to see if I get a better response from the low end. I also noticed that when I adjust the xover on my ICBM, I don't seem to get much change - I will have to explore that.

I suppose it would be better to use just the signal to the sub only for placement then get it tuned in with the front speakers, right?

I will try to post more graphs when I get time.

Thanks again for your help.

[angryht]

I did some more testing over the weekend. I moved the sub to a number of locations and still did not get much of a response from below 30 Hz. I also looked at the room modes. I got the following using the room mode calc located here: http://www.hometheatershack.com/foru...converter.html

I used dimensions of: 7 FT height, 8'10" width, and 12'8" length. I had to take some averages of the width and height.

Maybe I'm not reading this right, but if the room affects my response below 30 hz, wouldn't that mean there should be a mode around that frequency and that I would be in a null??

There is probably more going on in the room - tangential modes???

I will post more graphs later tonight with just the sub.


I was also curious about the sub itself (Def Tech Super Cube 2). Would the fact that it has one active woofer and 2 passive woofers cause some sort of cancelation? Does anybody else have any experience with the Def Tech subs and REW?

[angryht]

Forgive me for breaking the suggested graph limits but I had to show you guys this. Here is my sub only. I did not think I was getting any response below about 30 Hz. It looks like a huge dip. Is this normal below 20 hz???

Can this be fixed with a BFD?

Incidently, my crossover has two settings - 12 dB/Octave (normal) or 36 dB /Octave (special for THX subs). I have it set to normal or 12dB/octave.

[Wayne A. Pflughaupt]

I dunno – typically a “real” null is pretty narrow. That thing’s more than 2-1/2 octaves wide.

Can this be fixed with a BFD?

I have my doubts, unless you have something like 30 miles of headroom. If not, it’s impossibly deep - 30 dB!!! Aside from that, the lowest filter setting the BFD can be set at is 20 Hz. Sure, you could affect lower with a really wide filter, and then EQ down everything above 30 Hz, but as noted, you would have to have a sub system – both amp(s) and driver(s) - that’s totally loafing right now, using maybe only 5-10% of their capability.

I’m can’t help but think there’s something we’re missing. I can’t say I’ve ever seen anything quite like this. Not in a corner, anyway... Are there any gaping holes in the wall - bar pass-through, doorway, etc. – near the sub?

Regards,
Wayne

[angryht]

Wayne A. Pflughaupt wrote:

I dunno – typically a “real” null is pretty narrow. That thing’s more than 2-1/2 octaves wide.


I have my doubts, unless you have something like 30 miles of headroom. If not, it’s impossibly deep - 30 dB!!! Aside from that, the lowest filter setting the BFD can be set at is 20 Hz. Sure, you could affect lower with a really wide filter, and then EQ down everything above 30 Hz, but as noted, you would have to have a sub system – both amp(s) and driver(s) - that’s totally loafing right now, using maybe only 5-10% of their capability.

I’m can’t help but think there’s something we’re missing. I can’t say I’ve ever seen anything quite like this. Not in a corner, anyway... Are there any gaping holes in the wall - bar pass-through, doorway, etc. – near the sub?

Regards,
Wayne

Well I do have a lot of headroom left. My sub is at about 15 to 20 percent of capacity. I bet I could easily boost it up to get the 15 Hz to 75 dB. But wouldn't that blow out every thing else?

[angryht]

Wayne A. Pflughaupt wrote:

I’m can’t help but think there’s something we’re missing. I can’t say I’ve ever seen anything quite like this. Not in a corner, anyway... Are there any gaping holes in the wall - bar pass-through, doorway, etc. – near the sub?

Regards,
Wayne

No gaping holes. I do have a basement window that is in the room, but it is small (typical basement window). The window is closed, and I have linacoustic lining the walls. The position of the sub is fairly close to the corner - just the right side of the left speaker. When I move the sub around the room I get, in general, the same response.

[angryht]

Wayne,

Here are 50 percent, 75 percent and 90 percent in that order:

[brucek]

The information below 20hz is not actual output from the sub. As the response drops into the noise, the meter calibration file compensates (as it also drops) and applies its opposite compensation.

See in your graph where the cal file drops past 10Hz and carries lower (because C-Weight is incorrectly turned on). This offers increasing compensation, and as the actual signal enters the noise, the meter file creates the strange rising signal.

Your usable response goes a little below 20Hz.

Below I show a couple graphs of a small sub I own.

You can see from the impulse response where the signal enters the noise about ~45dB down.

Then see the associated frequency response plot (with the exaggerated vertical scale) that is created from that impulse response, and you can see where the signal is down about ~45dB, but then begins to rise. This isn't signal - it's noise.

Generally you determine where your sub signal enters the noise and set your bottom horizontal axis to that...... For most people, that horizontal scale is about 15Hz.... The LLT guys use about 10Hz....





brucek

[angryht]

brucek wrote:

The information below 20hz is not actual output from the sub. As the response drops into the noise, the meter calibration file compensates (as it also drops) and applies its opposite compensation.

See in your graph where the cal file drops past 10Hz and carries lower (because C-Weight is incorrectly turned on). This offers increasing compensation, and as the actual signal enters the noise, the meter file creates the strange rising signal.

Your usable response goes a little below 20Hz.

Below I show a couple graphs of a small sub I own.

You can see from the impulse response where the signal enters the noise about ~45dB down.

Then see the associated frequency response plot (with the exaggerated vertical scale) that is created from that impulse response, and you can see where the signal is down about ~45dB, but then begins to rise. This isn't signal - it's noise.

Generally you determine where your sub signal enters the noise and set your bottom horizontal axis to that...... For most people, that horizontal scale is about 15Hz.... The LLT guys use about 10Hz....


brucek

Thanks, Bruce. So based on that I should set my lower limit of frequency back to 15 and try to equalize that. You mentioned that C-weighting is incorrectly turned on - do you mean below 15 Hz? I assume I should have it on for measurements and that it is not relavant below the limit of my small sub.

Who are th LLT guys? I did not understand the reference.

From my graphs of 50, 75 and 90 percent, would a BFD help to tame things with my sub?

Thanks again for the help.

[brucek]

So based on that I should set my lower limit of frequency back to 15 and try to equalize that

You should set the axis on your graphs to horizontal of 15Hz-200Hz and the vertical to 45dB-105dB. The third (blue) plot you show is quite good. The natural low end extension is a bit below 20hz. That's good. You can't equalize below 20Hz with a BFD. Either way, there's no upside to trying to boost below a subwoofers rolloff at the bottom end.

You mentioned that C-weighting is incorrectly turned on

The C-weight check box in REW is only active outside the limits of the meter calibration file. It should be turned off.

Who are th LLT guys? I did not understand the reference

Many members here have built LLT's. See the DIY subwoofer and Soundsplinter sections in the forum..
LLT explained

From my graphs of 50, 75 and 90 percent, would a BFD help to tame things with my sub?

You have a very good response. I wouldn't try and equalize it.

brucek

[angryht]

brucek wrote:

You should set the axis on your graphs to horizontal of 15Hz-200Hz and the vertical to 45dB-105dB. The third (blue) plot you show is quite good. The natural low end extension is a bit below 20hz. That's good. You can't equalize below 20Hz with a BFD. Either way, there's no upside to trying to boost below a subwoofers rolloff at the bottom end.

ok. I will reset the graph limits. That last graph (blue) is just a result of me turning up the gain on the sube to about 90 percent of the total. I did not set the levels before I took the reading. I just did a quick check levels before the test and it said ok. Maybe that's all I need to do?

brucek wrote:

The C-weight check box in REW is only active outside the limits of the meter calibration file. It should be turned off..

I don't understand. I am using the radio shack analog meter, aren't I supposed to have the C-weight box checked? Is it different for just measuring the sub? I think I am missing something. Could you please explain?

brucek wrote:

Many members here have built LLT's. See the DIY subwoofer and Soundsplinter sections in the forum..
LLT explained.

I will check this stuff out. Thanks for the clarification.

brucek wrote:

You have a very good response. I wouldn't try and equalize it.

brucek

That's good news. I just need to get it dialed in with my mains.

[angryht]

One more question. I am running everything through my ICBM for bass management. I did those readings (green red and blue graphs) with my crossover set to 60 Hz. It looks like it is not working correctly. Shouldn't I have a drop off just before 60 Hz? From the graphs, it looks like it's set at about 120 Hz.

[brucek]

I just did a quick check levels before the test and it said ok. Maybe that's all I need to do?

Yep, if it says OK, that's fine.

I don't understand. I am using the radio shack analog meter, aren't I supposed to have the C-weight box checked? Is it different for just measuring the sub? I think I am missing something. Could you please explain?

The C-weight check box only affects the calibration outside the limits of the meter calibration file.

So, if I don't enter a meter calibration file and check the C-weight box, then REW will use all C-weight calibration values for the entire frequency band from 0Hz to 30Khz. Meters unfortunately don't track a C-weight curve exactly, so we need a calibration file instead.

If I enter a meter calibration file that starts at 10Hz and ends at 200Hz, and I have the C-weight box checked, then C-weight calibration values are used below 10hz and above 200Hz. But, it would be meaningless to use a meter outside the meter calibration files outer limits, (since it wouldn't be accurate), so there is no reason to check the C-weight box..

Shouldn't I have a drop off just before 60 Hz?

It should theoretically track the target line set in REW. Be sure to set the desired target frequency and slope in REW that you want to use.

brucek

[angryht]

Yep, if it says OK, that's fine.


The C-weight check box only affects the calibration outside the limits of the meter calibration file.

So, if I don't enter a meter calibration file and check the C-weight box, then REW will use all C-weight calibration values for the entire frequency band from 0Hz to 30Khz. Meters unfortunately don't track a C-weight curve exactly, so we need a calibration file instead.

If I enter a meter calibration file that starts at 10Hz and ends at 200Hz, and I have the C-weight box checked, then C-weight calibration values are used below 10hz and above 200Hz. But, it would be meaningless to use a meter outside the meter calibration files outer limits, (since it wouldn't be accurate), so there is no reason to check the C-weight box..


It should theoretically track the target line set in REW. Be sure to set the desired target frequency and slope in REW that you want to use.

brucek

Now I think I get it. Since I am measuring from ~10 hz to 200 hz and I have loaded the calibration file for my SPL meter (from this site), REW is not using the C-weight because there is nothing being measured outside the limits .cal file. So, if I wanted to use the meter above 200 hz, I should use the c-weight, right? Otherwise it isn't doing anything except under 10 hz which doesn't matter anyway.

[brucek]

REW is not using the C-weight because there is nothing being measured outside the limits .cal file. So, if I wanted to use the meter above 200 hz, I should use the c-weight, right? Otherwise it isn't doing anything except under 10 hz which doesn't matter anyway.

Yeah, if you wanted to measure above 200Hz, you would turn on the C-weight. The C-weight compensation at and around 200hz is about nil anyway in this case, but you've got the idea.....

brucek

[angryht]

Well I did some more testing last night and all I have is questions. The 4 results (sub only) are below.
Graph 1: Set up using the Settings procedure
Graph 2: Increased master volume (pre/pro) about 9 dB from the first reading. Did a quick Check Levels prior to running measurement and got green -12.7dB
Graph 3: Increased master volume 10 dB from #2. Check Levels showed -9.7dB in red.
Graph 4: Increased master volume 10 dB from #3. Check levels showed -8.2dB in red.

It seems that if I simply increase the volume on the sub the response flattens out. Isn't that strange? Or is the sound greater than the meter can read, hence the red Check Level values (maybe top end is being cut off and just being shown flat???). If I am trying to get a flat response, I should just turn up the sub but then I am exceeding the Settings Procedure? Then when I try to turn up the sub alone to blend with the mains it is overwelming to my ears but the response graph out flat around 90 dB?? Am I missing something in the Set up?

[brucek]

Am I missing something in the Set up?

You need to do the setup procedure in REW and then leave the volume levels alone. If you change them significantly as you have, then you should really redo the procedure. There's really no reason to modify the levels once REW is setup. Reasons to re-setup REW would be a change in BFD filters (which alters the overall level from the sub), or adding in the mains with the sub (which again alters the overall level). Turning up the levels can clip the meter or the input of the soundcard...

To go through these steps only takes a second......

1. Ensure the soundcard and meter calibration files are loaded.

2. Run the Check Levels routine with the SPL meter at the listening position and the dial of the meter at 80dB position. This gives the meter lots of headroom from the 75dB target you'll be setting in REW.

3. Run the Calibrate SPL routine in the meter tab to match REW and the SPL meter to ~75dBSPL.

4. Run the Set Target Level routine, which will set the target to ~75dB.

5. Run the Measure routine and set the End Frequency to 200Hz.

6. Measure....................

brucek