SB Audigy 2 Platinum and REW

**jagman** · **10-20-06, 01:25 AM**

I must have luck. Java is up to date so I re-set the meaurement level. When set at -12dB and the integrated's volume was at 20 the SPL meter registered 72dB, so I dropped the measurement level to -10cB and that resulted in the SPL meter flickering between 74 and 75dB. Good enough. Since I can't finish setting the input volume, I left it as is since everything made sense. So, I calibrated the SPL reading without a hitch. Then I tried to calibrate the sound card. Since the stereo cables plugged into the LINE_IN and LINE_OUT jacks have left and right RCA's, I connect the LINE_IN left to the LINE_OUT left and the LINE_IN right to the LINE_OUT right to create the loop. I then clicked "Measure Coundcard Response...", kept the level at -6.0dB, clicked "Continue" and nothing happened. No measurement came about. The REW software looks like I never pressed "Measure Coundcard Response..." in the first place. Is this normal? Maybe I should just download REW again.

**jagman** · **10-20-06, 02:14 AM**

I decided to play around and started pressing buttons. Low and behold I hit "automatic measurement" and sure enough, it measured the soundcard response. If that was in fact the right thing to do, it would probably be a good idea to have that mentioned in the Help File. So, I got my Soundcard measurement, except "Actual RMS at 1kHz" was not measured and neither were the 2nd or 3rd Harmonics at 1kHz. OK. I continued on anyways, clicked "Left", unselected "Compensate for C Weighting", hit "Automatic Measurement" and sure enough I got a nice flat line. Beautiful, except it's flat at 81.5dB and not 75dB. Is it supposed to be this way or am I just doomed? Is the program set up properly to start taking measurements of my system? Hopefully I'll be able to do that this weekend.

I've had enough for one night. Time for bed. Sorry for the hand holding but please tell me if I'm doing this properly or not. I'm very greatful for the help, yet at the same time it's hard to figure things out when the advice I get doesn't happen when I'm actually playing with the program. Sorry for the rant. I just want to get this program running so I can get great sound out of the sub.

**brucek** · **10-20-06, 08:20 AM**

Quote:

I connect the LINE_IN left to the LINE_OUT left and the LINE_IN right to the LINE_OUT right to create the loop.

No, you only make a single loopback connection. If you are going to use the right channel for measurements, then you remove the connections to the SPL meter and receiver and short the LINE_OUT right to LINE_IN right. The other left channel always remains unused... Use the right channel only. The loopback is only used to do the soundcard test and to create a soundcard calibration file, then the loopback is removed and your regular cabling put back in place.

Quote:

When set at -12dB and the integrated's volume was at 20 the SPL meter registered 72dB, so I dropped the measurement level to -10cB and that resulted in the SPL meter flickering between 74 and 75dB. Good enough

It would have been better to leave the level at -12dB and simply turn up the volume of your receiver. The actual volume setting of the receiver makes no difference. It can be -20dB or -5dB or whatever. You're simply attempting to create the correct input and output conditions of REW so it can accurately measure your response and recommend BFD filters.

Quote:

Since I can't finish setting the input volume, I left it as is since everything made sense.

As long as you feel the input VU meter looked like it was responding with some signal and wasn't in the red. If you think the input volume is a bit low you can simply click the little thumbwheel and turn it up a bit. You want the input VU meter to be in the yellow when it's perfect. Since you can't use the automatic input volume setup, simply play some pink noise with REW after you've 'set the measurement level' and see where the input VU meter is responding. If it's low, turn it up.

Quote:

The REW software looks like I never pressed "Measure Coundcard Response..." in the first place. Is this normal? Maybe I should just download REW again.

It should just work when you press the button.

Quote:

I continued on anyways, clicked "Left", unselected "Compensate for C Weighting", hit "Automatic Measurement" and sure enough I got a nice flat line. Beautiful, except it's flat at 81.5dB and not 75dB. Is it supposed to be this way or am I just doomed? Is the program set up properly to start taking measurements of my system?

Use the right channel and stick with it. All loopback and measurements will be done with one channel.
'Compensate for C weighting' should be checked and you should also have the calibration file (.cal) for your specific meter loaded from here.
Make sure the sub tab is selected also. You also should select the target level before doing an automatic measurement.

brucek

**JohnM** · **10-20-06, 11:43 AM**

After hitting "Continue" on the soundcard measurement level dialog a progress panel should pop up and the process should run through automatically. If that isn't happening some sort of error might have occurred, if so there will be a log of what happened in the debug files. To locate those, open the Help -> About... box to see where your home directory is (typically c:\Documents and Settings\yourusername), the files are stored there. Look for files called roomeq_wizard0.txt, roomeq_wizard1.txt etc. Open each file in any text editor (e.g. Notepad) and see if there are any error messages, if there are then email me the file. A new file is created each time REW is started, the last 9 copies are kept in a rotating list.

**jagman** · **10-20-06, 12:38 PM**

Bruce and John... thank you for the advice. OK. When I took the measurement of the souncard, the help menu said to unclick C Weighting to see if the measurement was flat, but to otherwise have it clicked. That is my intension. I also checked the left channel as that is what the help menu said to do. I will recalibrate the level to -12dB and then attempt to recalibrate the soundcard using the right RCA only (as well as the right channel). I believe I'm getting an error with the system when setting the input volume as nothing pops up after playing the 75dB pink noise. I also think I'm getting an error as nothing pops up after clicking continue in the soundcard calibration. John... I will check for errors tonight and e-mail anything out of the ordinary. It sounds like I just have a few bugs to work out. One other thing, I already have the calibration file for my RS SPL meter... but how do I apply it to the program?

**jagman** · **10-20-06, 12:47 PM**

Nevermind... I assume I click "load meter/calibration file..." under the meter section. If there is anything else to making REW use it please tell me.

**jagman** · **10-20-06, 10:41 PM**

Well, I did everything as described, but when I went to calibrate the soundcard the same thing happened... no popup. But, in addition, the plot looks like a 9.5 earthquakes reading on a seizemograph (sp?). It also says the impulse peak is not where it should be.

**brucek** · **10-20-06, 10:48 PM**

Quote:

but when I went to calibrate the soundcard the same thing happened... no popup

I'm certainly at a loss to know what's wrong? Don't know of anyone having this trouble.

If you send the debug file to John, I'm sure he will be able to come up with more answers...

brucek

**jagman** · **10-21-06, 01:48 AM**

The files are long. I don't know what that means, but there is a ton of stuff in them. As for what's going on, I figured out a couple things. The graph went crazy because REW reverted some things on the soundcard Mixer. After I took care of that, I reset the values and come up with this for an unalterred plot...

Any opinions on where to start?

Obviously I have that nasty null at 74 Hz. I'll just have to let it go and when I have the capability cross over to the mains below it. As for what's going on elsewhere, there are definitely places to cut, but should I boost the lower frequencies to bring 'em up? I've played around with it quite a bit and came up with different plots, but I'd rather ge your advice before enacting them. What would you consider the maximum boost advisable?

Here's another idea. If I set the Target to 65dB (instead of 75dB), for the most part, all I'd have to do is cut the output at selected frequencies. I could get a very good looking plot, but how would that affect the overall volume coming from the sub after calibration? Would it be too quiet? I don't think it would be because without the BFD the sub is very loud. When I did raw measurements with the SPL meter (with the pink noise at 75dB) using the updated correction factors, the response was 25 dB hot in places! I would imagine if I calibrate with a target of 65dB I would get a great curve all the way to 10 Hz. And, since I'll be cutting only 15 Hz or so, that means I'd still have a 10dB house curve... does that make sense?

**brucek** · **10-21-06, 06:47 AM**

First, why are you measuring to 24Khz? You're equalizing a sub woofer with a limit range.

Set the Graph axis limits for every graph with a vertical scale from 45dB to 105dB and a horizontal range from 15Hz to 200Hz.
Reset the graph after every measurement to that range setting.
There is an icon called Set Graph Axis Extends at the bottom right hand side of the graph to store and reuse each measurement to make it convenient.

Set the sweep measurement from 20Hz to 200Hz.(it will actually sweep wider than this to 400Hz, but use this setting)

Quote:

If I set the Target to 65dB (instead of 75dB),

No, use 75dB.

Quote:

but should I boost the lower frequencies to bring 'em up?

No, you simply cut the higher frequencies and compensate later for any levels with your sub amplifier.

Quote:

how would that affect the overall volume coming from the sub after calibration? Would it be too quiet?

Turn up the sub.

Quote:

I would imagine if I calibrate with a target of 65dB I would get a great curve all the way to 10 Hz. And, since I'll be cutting only 15 Hz or so, that means I'd still have a 10dB house curve... does that make sense?

No, the purpose of REW and a BFD is to remove peaks caused by resonances. It's intent isn't to make a sub extend to 10Hz when it isn't capable.

The idea is to locate a sub to get the best response and then remove room resonant peaks with a few cutting filters.

brucek

**JohnM** · **10-21-06, 11:44 AM**

Hi Krister,

Got your email, thanks. The Exceptions (aka errors) that are logged look they might relate to the look & feel you have selected, do you run a theme other than one of the Windows defaults on your system? There is a bug within Java that can cause errors on systems not using the default look&feel themes, which looks like the culprit here. If you are using a non-standard theme please post what it is, I can then feed that back to Sun to ensure the problem has been resolved for the Java V6 release.

**jagman** · **10-21-06, 02:36 PM**

John... the only thing out of the ordinary on the way I have windows set up is I have the screen black. Besides that, it's just the norm. After work, I double check the exact setting, and then reset the diplay to the original Windows theme and see if that makes a difference.

Bruce... I'll be sure to standardize the graph and readout response. I didn't get the plot out until late last night so at that point I was just happy to have a plot to play with

. As for adjusting the sub volume, I have an IB sub powered by a prosound amp. The cut on the sub amp is at zero (the nob on the front is turned fully clockwise). The only other way to increase the volume relative to the mains is to increase the gain of the SUB OUT on the integrated but that was set to optimize the level input to the BFD so as to not make it clip. Obviously I don't want to change that. That's why I questioned whether I should decrease the gain (or increase the cut) on the Samson itself before calibration. Having headroom to adjust the volume of the sub vs the mains after calibration would be nice .

Having said all that, there is the fact that the sub played significantly louder than the mains over the bulk of it's frequency range before calibration when the gain on the Samson was at the same level. I tested the system response previously with test tones playing from 10Hz to 100Hz in 1Hz steps, and then spaced out from there up to about 200 Hz. When I tested the combined output of the sub and mains, the pink noise was set to 75 dB. The in room response from the main seated position started at 77dB at 10 Hz and steadily and for the most part evenly rose to 97 dB at 37 Hz (there were a few minor blips). It then dropped to 83 dB at 46 Hz. After that it rose to 100 dB at 56Hz (and stayed there until 63Hz). From there it droped to 95 dB at 76 Hz and then rose to 99 dB at 81 Hz (and stayed there until 100 Hz). Then it slowly dropped to about 75 dB at 140 Hz. As stated before, all of this was done with the amp gain turned up 100% and the level output from the integrated amp set to a point where the clip light on the Samson S700 has only flickered once.

What I take from that is with the gain of the sub amp where it is right now, I should be able to generate a plot without boosting anything that is flat to 10 Hz (except for where the dip is at 74 Hz).

**brucek** · **10-22-06, 07:36 AM**

Quote:

What I take from that is with the gain of the sub amp where it is right now, I should be able to generate a plot without boosting anything that is flat to 10 Hz

I understand the point you're making, and in the case of an IB (which would normally extend very low without problems), you may be correct.

Here's my point. Consider a sub that has a stated response from 25Hz to 150Hz. A graphed response check reveals (as expected) that it's down 10dB at 15Hz. A creative person might say, "I could reduce all the frequencies above 15Hz down to the level of the 15Hz signal and then turn up the main gain and my response would go from 15Hz to 150Hz".......
And indeed it would, except there would still be minimal output at the low end since that sub simply cannot produce any SPL level from 15hz to 25Hz. It just isn't capable.

From the levels you post (and evidenced from your graph), your contention is that you have a very large peak from 25Hz to 140Hz. I guess this could be so, but I can't say either way. I wonder what sort of SPL level you can muster up at 10Hz?

Anyway, there certainly are devices like this that can be placed between the BFD and pro amp that raise the line level. That way you'll have some control over the sub amp gain dial......

brucek

**jagman** · **10-22-06, 10:52 PM**

Well, I played around a bit, went with my hunch, and the results are very satisfactory. I tried different filter configurations before coming up with one that I like. What I did was compare the REW raw plot to my measurements pre BFD and pre REW. I ultimately cut a lot in most places (since the sub was playing so hot) and boosted mildly in a few others. I know boosting is at my own peril, but I decided to try it. I kept the volume down and increased it as I got more comfortable with the amps ability to handle the signal.

The upshot is I remeasured the response, and it is much better. It starts at 75dB at 10 Hz and steadily increases to 87dB at about 30 Hz. It then slowly decreases to about 81dB at 120 Hz and then drops to about 75dB at 140Hz. Voices sound much more realistic... the lower voices were way to heavy with the sub as hot as it was before. It does seem much quieter, though. Kind of weird. Music is much better. I'm still getting used to the quality of a an IB sub that is properly calibrated. The only problem is there is an unavoidable 4dB dip between 40 Hz and 60 Hz. There isn't much I can do about that and it probably isn't that discernable anyways. Another thing is with the sub response dialed in the amp doesn't come close to clipping. Overall I'm very happy

.

**jagman** · **10-22-06, 10:53 PM**

Once again, I just want to thank Bruce for his repeated help. Sorry for needing so much hand holding.

**jagman** · **10-26-06, 06:33 PM**

After a few more adjustments (it never ends

), I have a plot that really makes me happy.

Just ignore the dB's on the left. Actual measurements are about 20 dB higher, but the overall shape of the REW plot is almost identical to the corrected RS SPL meter's measured response. There are only couple difference. The first and obvious difference is when the sub and mains are both on the response flattens out at 80Hz and stays at 75dB (which correlates to ~56dB on the REW plot) from there on upward. The other difference is the exception to what I just stated... there was a peak 120Hz when the the sub and mains played simultaneously (which is why there is a dip in the REW plot at 120 Hz). The combined result after EQing is a beautiful house curve with a peak at about 30 Hz that is 11 dB hotter than the pink noise.

I absolutely love the sound now that it is finalized. BTW, the amp is nowhere near clipping, so even though I boosted a few areas there appears to be no side effects. Even still I'll be careful when watching WOTW for the first time. Batman Begins absolutely kicked $#^!

Next project will be a good PJ for the HT!

**jagman** · **10-29-06, 04:46 PM**

Bruce... I couldn't figure out what the was going on. To say the least it was really "irritating". Anyways... I finally figured out what the problem was... the "in/out" LED wasn't lit

. Now the is output is amazing!!!

The following is with +13 boost at 20 Hz with the bandwidth at 76:

The following is with +8 boost at 20 Hz with the bandwidth at 76:

So... the beauty is I can cut back on the bass boost if needed after I figure out what the system can handle. Obviously I'll be very carefull in the mean time.

**brucek** · **10-29-06, 05:28 PM**

Quote:

I finally figured out what the problem was... the "in/out" LED wasn't lit

Yeah, been there, done that....... I sure find the filters work best when turned on...hehehe

bruce

**jagman** · **10-30-06, 09:53 PM**

It's amazing... you feel so stupid yet so happy to get it working. Anyways... the boost I put in was way too much. The subs started to bottom (I think) in WOTW. I kept repeating and slowly increasing the volume in the scene where the pavement cracks and the first machine rises. At the highest volume I like to listen to movies the subs weare easily moving 2.5" if not 3" peak to peak (3" is peak to peak Xmax). They were moving slowly, but a moo baby moo of a long way. Creepy and scary to say the least. I heard a couple of small pop like noises when they moved that much, so I quickly turned it down. Luckily no damage was done. At any rate, I alterred the boost in the lower frequencies so now it tapers to no boost at 10 Hz. Before it was boosted about 8dB! Stupid is all I can say. After dropping off the boost they moved about 1.5-2" without a hitch. That seems to be about right. Since the scenes are such an extreme, I think they'll be safe. They also seemed to move about 1.5-2" during the helicoptor scenes in BHD. It's a rush to stand below the IB, look up into it, and see those puppies moving so much.

Here's the new one...

**jagman** · **02-18-08, 02:17 PM**

OK guys,

I'm back at EQing my sub. I've changed a few things around in the room and want to re-optimize my response. Since I had so many problems and the dB level was not accurate, I want to go over the steps I've taken so far to be sure I've done things correctly.

I downloaded the updated version of REW.
I clicked settings and proceeded to set up my system.
Under Soundcard, I have the input and ouput devices as my SB Audigy 2. Output is "Speaker" and input is "LINE_IN (Line-In 2.Mic 2)"

Wave Volume is 1.000 and unmuted
Output Volume is 0.505 and unmuted
Input Volume is 1.000
Sweep when I turn of REW is 20

I'm having trouble calibrating my soundcard. When I go to do it, I set the Sound Mixer as follows:

Master Control...
Volume 51%
Bass 50%
Trble 50%

Source...
Wave 100% and unmuted
MIDI Synth 0% and muted
CD Audio 0% and muted
Line-In 100% and unmuted
Line-in 2/Mic 2 0% and muted
Microphone 0% and muted

Record...
Analog Mix (Line//CD/Aux/TAD/PC) 97% and Record Without Monitoring is NOT checked.

With the mixer set like that, I manually change the Sweep Level to -16.0 dB and the Output and Input meters read as follows:

Output: -13.0 at the top and -16.0 at the bottom
Input: -12.9 at the top and -15.9 at the bottom

I think that's set properly as it's in the green (not yellow) which I take it as meaning it's safe.

So, I hit next in the help menu at the bottom of the settings window since the input volume has been set. The help menu says the soundcard can be measured, so I hit next. As predicted, the settings window minimizes and the new measurement shows. But, unlike predicted, the Settings window doesn't pop back up. When I manually bring it back up, I can hit finish.

That's great... the plot looks pretty standard, but how do I save my calibration file? The "Make Cal..." button is greyed out.

Thanks advance for your help.

**brucek** · **02-18-08, 02:32 PM**

Quote:

That's great... the plot looks pretty standard, but how do I save my calibration file? The "Make Cal..." button is greyed out.

Press FINISH and then the Make CAL is enabled....

**jagman** · **02-18-08, 05:31 PM**

I did that but it remained greyed out. I repeated twice but it never became enabled.

**JohnM** · **02-18-08, 05:54 PM**

Two things seem a bit odd, you say you have:

Source
Line-In 100% and unmuted
Line-in 2/Mic 2 0% and muted

for Source, why is Line In unmuted? If you need the input unmuted in Source to work then shouldn't it be Line In 2/Mic 2 that is unmuted as that is the one you say is selected as the input?

You also say "Record Without Monitoring is NOT checked". If the box does say Record without monitoring then you want it to be checked.

**jagman** · **02-18-08, 06:39 PM**

Thanks guys.

I'll double check things tonight and reread the help menu. I'm sure it's something simple.

**jagman** · **02-18-08, 11:41 PM**

JohnM -

I double checked my conections and settings. There are two inputs listed as LINE_IN (Line-In 2/Mic 2) in Sound card tab of the Settings window. Why I don't know, but they are listed that way. When I select the first one and use LINE_IN 1 on the back of the sound card, I get a reading (so it must be the right one since I don't get any readings when any of the other tabs are selected). I also realized that I'm supposed to have "Record without Monitoring" button selected, so I did that.

Still, even when things are set up properly, I can't calibrate my soundcard. When I click the Measure button in the Calibration section of the Soundcard tab of the Settings window, I click next in the first window because my loop is set up properly. I then click next again because the Sweep Level, Wave Volume and Output Volume are set as instructed. In the following window, with Sweep set to -16.0, the Output put goes from -13.0 to -16.0 and the Input goes from -13.1 to -16.1, I so I hit next again. It says the soundcad can now be measured, so I should hit Next. It says the window should shrink and a measurement progress panel will appear. Well, when I hit next, teh settings tab minimizes but all I'm left with is the main REW window without any measurement. I don't know what I'm doing wrong. It's very frustrating.

Advertisement