Initial REW Setup

**falcon40** · **06-10-09, 01:26 AM**

I have read the help files for REW setup and am confused about which method to use for sound card compensation. I currently have the loopback setup in place (right channel meter input--left channel looped, etc.). The following quote from the help file suggests that everything is accounted for automatically by the program due to the loopback setup.

The other input and output channels do not need to be used. The response of the soundcard itself can be compensated for by taking a reference measurement with the output connected directly to the input and configuring REW to subtract that measured response from subsequent room measurements. However, it is also possible to use a loopback connection from the soundcard's left output and left input to have REW automatically compensate for the soundcard's response when it makes a measurement. If you wish to do this you may need an additional RCA phono plug coupler to make the loopback connection. An advantage of using the left channel loopback is that it provides REW with a timing reference and the delay through the system being measured can be calculated, which can be useful for getting speaker delay settings correct in multi-channel systems. A disadvantage is that there is a small reduction in the signal-to-noise ratio of the calculated impulse response, although this is so small as to be of no consequence for acoustic measurements. Whether REW uses the left channel as a reference is controlled by a check box in the Soundcard Settings.

My question is, for my setup do I need to do any compensation measuring? If not, should I just skip the measurement steps in the soundcard setup?

I apologize for not being able to get past the first few steps but I do not want to proceed under a false sense of security.

Thank you in advance for any help.

falcon40

**brucek** · **06-10-09, 08:01 AM**

Quote:

or my setup do I need to do any compensation measuring?

There are two methods to using REW. The first is the use of a one time created soundcard calibration file and the other is the use of the left channel of your soundcard being used dynamically as the calibration.

I highly, highly, highly recommend using the soundcard calibration file method.

It's quite simple. Connect a single loopback cable from line-out to line-in on the right channel and run the soundcard calibration routine and save the file as instructed by REW. Remove that cable and now use that channel for your SPL meter and output to the receiver.

brucek

**falcon40** · **06-10-09, 08:46 AM**

Thank you brucek for the quick response. I will try to proceed as you suggest. When I tried using the dynamic method I must be doing something wrong because my measurements always end up with the exact same SPL on REW's measurement--that can not be right.

One more (I hope) question. Am I just using one channel to measure SPL or using both. Measuring just one channel seems unrealistic. Does the system need to measure the total output from both speakers to get the sum impact on SPL?

Thanks again for the advice.

falcon40

**brucek** · **06-10-09, 09:05 AM**

Quote:

Am I just using one channel to measure SPL or using both

That's up to you. You feed the single mono line-out channel of the soundcard to your CD or AUX input of the receiver and use a Y-splitter. This allows you to see the response of the sub alone or each speaker alone, or both speakers, or sub plus both speakers. The object of the exercise is not to determine level, it's to measure the frequency response in a room.

Whenever you change what speaker you want to measure, you reset the levels in REW to 75dBSPL. This allows you to plot the responses with a trace around 75dB. There is also a trim adjust in REW to help overlay them if needed.

brucek

**falcon40** · **06-10-09, 11:51 PM**

OK--I realized one of my problems was I needed to use the 44.1 kHz sample rate instead of the 48 kHz. I did as you suggested and used the soundcard file calibration method. SInce I used this method, is it even necessary to have the left channel looped back or is it O.K. to have the left and right channel output go to my preamplifier? It seems like my current wiring was for the dynamic calibration method.

It is nice to start seeing realistic graphs on REW. What a fantastic piece of software. I will certainly be making a donation soon for this current product in hopes of continued development/refinement of this product or others that may be in the works.

falcon40

**brucek** · **06-11-09, 07:28 AM**

Quote:

is it even necessary to have the left channel looped back or is it O.K. to have the left and right channel output go to my preamplifier?

You're not suppose to have the left and right channel looped when you use the soundcard calibration file method. The loop is used with the dynamic calibration method along with the checkbox enabled called ("Use Left Channel as calibration Reference"). Be sure to uncheck it when using the soundcard calibration method.

It's best to use the calibrated channels output connected at the receiver with a Y-splitter to feed both left and right.

brucek

**falcon40** · **06-16-09, 10:15 PM**

brucek,
Thank you for your patience and explanations. After much trial and error, I think I have finally figured out the correct steps to evaluate my room acoustics. I would like to post some graphs because I suspect I still may not be doing something correct. In what forum should I attempt to post some graphs and what pixel width is acceptable to the forum population? I assume I would use the "Save Graph as JPEG" function and then post to the forum. In addition, if and when I do post a graph, what graph plots should be included vs left out, i.e., date plot, Corrected, Target, Mic/Meter, etc.? Please advise.
falcon40

**brucek** · **06-16-09, 11:08 PM**

Quote:

In what forum should I attempt to post some graphs

Post in this thread.....

Quote:

what graph plots should be included

Just post a measurement trace with a target.

For subwoofers, always use the standard Vertical graph axis of (45dB - 105dB) and the Horizontal graph axis of (15Hz - 200Hz) using the Graph Limits button in the top right corner of REW.
Use a measurement start frequency of 0Hz and an end frequency of 200Hz.

Always set a measurement target of 75dBSPL.

Graphs are best posted with an 800 wide resolution using the Save Graph as Jpg function.

brucek

**falcon40** · **06-17-09, 12:41 AM**

Thanks for the quick response. Here is a plot of my subwoofer. It looks way too artificial so I am wondering what I may have done wrong. Please examine and provide feedback. I calibrated the soundcard and am using the Radio Shack calibration file as well. Thank you.

Name: 123.jpg
Views: 196
Size: 51.8 KB

**brucek** · **06-17-09, 07:13 AM**

Quote:

Here is a plot of my subwoofer.

No, that's a plot of a loopback cable with a soundcard calibration file loaded and a meter calibration file loaded.

If you don't have a loopback cable installed, then it is a result of monitor mode being turned on in the cards application software. Turn it off and re-measure.

BTW. Re-read my post above on the proper graph axis to use.........

brucek

**falcon40** · **06-17-09, 09:34 PM**

O.K. Thank you for the feedback. I will tinker with my sound card settings some more. I am using GNU/Linux so it is not as straight forward as when in Windows on this particular issue. I did check the scale and thought it was already set-up to render the proper scale. I now realize I have to click "Apply Settings." Sorry for being difficult. Your responses are helping me learn, however. I appreciate it. I will report back if I am ever able to turn the monitor mode off on my sound card.

falcon40

**brucek** · **06-18-09, 07:10 AM**

The shape of the red grapg trace is the clue that reveals the monitor mode is on.

Notice that the trace is the exact inverse of your meter calibration file. That is the result you would get with a loopback cable or anytime the output is connected to the input channel that you're using to measure.

If you cleared the meter cal file, the trace would be a perfect flat line (the indication the the soundcard calibration file is loaded and good).

The result that monitor mode is turned on is that the Record Mixer has line-in enabled. Only the Playback Mixer should have line-in enabled. Sorry, but I can't help you with finding that in Linux.

brucek

**falcon40** · **06-19-09, 04:16 AM**

I think I solved the sound card "monitor" issue. I routed the microphone (RS 33-2055) through Aux 2 instead and turned the capture slider off. At least it looks like I am progressing to something closer to reality but it still looks odd. Here it is. Any feedback is appreciated. A simple "Go" or "No Go" will suffice. I will assume a "No Go" means I still have something wrong with the capture/monitor issue.

falcon40

Name: sub plot test.jpg
Views: 152
Size: 37.1 KB

Name: sub plot test.jpg
Views: 152
Size: 37.1 KB

**kjgarrison** · **06-21-09, 11:27 AM**

No go.

**brucek** · **06-21-09, 11:57 AM**

Quote:

but it still looks odd

That's a graph of the C-Weight box being checked while taking a measure of a loopback cable.

You don't have a loopback cable connected do you?

brucek

**falcon40** · **06-21-09, 05:47 PM**

kjgarrison and brucek,

Thank you for the response. I do not have a feedback loop setup. I did use a feedback loop to establish a calibration file, however. The problem lies in my sound card configuration--I believe. The help files allude to concerns with Audigy2 sound cards which I have. The problem is I use GNU/Linux and alsa sound server and the configuration process is apparently not as straight-forward or complete as the MS Windows process. I have put in a request on a linux forum to try and eliminate the sound monitor issue.

I will not waste yours or anyone else's time until I achieve some sort of "peak and valley" reading from REW.

Thanks again.

falcon40

Advertisement