couple more newb questions with rew

[hoofenshnoz]

hey guys
so i have my soundcard and meter all set up and and im fooling around with my measurements now. just a couple questions. my first question is trying to understand why you have to loop the output of the soundcard to the L or R input in conjunction with the meter? my second question is setting the target level....when should this be adressed, after the measurement has been taken?

thanks!

L

[Wayne A. Pflughaupt]

You don't have to loop the in/out once the sound card calibration file has been generated.

Set the Target for 75 dB initially, but if your response reading ends up being well above the Target, you can re-align it.

Regards,
Wayne

[brucek]

why you have to loop the output of the soundcard to the L or R input in conjunction with the meter?

It's not looped in conjunction with the meter. It's looped once so that a soundcard calibration file can be created, so that you have a calibrated channel on your soundcard to connect the meter to. The loopback is removed once the calibration routine is complete.

brucek

[hoofenshnoz]

oh ok so once i have say the right channel calibrated on my loop back. I unplug the loopback and plug my meter into that channel?i thought i plugged my meter into the oposite channel i used loopback for.

[brucek]

oh ok so once i have say the right channel calibrated on my loop back. I unplug the loopback and plug my meter into that channel?

Yes.

i thought i plugged my meter into the oposite channel i used loopback for.

No.

I urge you to read and become familiar with the REW HELP files and the REW Cabling and Connections Basics.

Also note the REW information Index and the Download Page.

brucek

[hoofenshnoz]

yah man ive spent hours upon hours reading those help files. obviously i am reading them a bit wrong

[laser188139]

As a relatively new user who also read through all the helpfiles, hoofenshnoz, perhaps I can help.

It appears that REW really supports two techniques, and you can use one or the other, as you like.

You can build a loopback connection once and use it to build a sound card calibration file. From then on, the software does not need the left channel loopback because it uses the right channel to send test signals and to receive audio from the measurement microphone.

You can build a loopback connection and use it all the time on the left channel. This lets the REW software determine the effects of the sound card whenever it takes a measurement, at any frequency, under the assumption that the sound card handles the left and right channels identically. As it can measure instantaneously any reduction in the looped back signal, it does not need to refer to a sound card calibration file. (This is the meaning of the Use Left Channel as Calibration Reference check box in the Settings.)

Some of the pictures in the forum's documentation show how to construct the latter cabling. After you've done it, you can just leave the cables all put together for the next session. The tricky part is to keep track of which is outgoing and which is coming back in.

The cabling is easier, and needs one fewer cable, if you use the loopback only once to calculate a sound card calibration file. You then disconnect it and can use the same connectors for measurement. There is less chance of confusion, as you don't see as many wires when you are plugging into the soundcard at the start of the measurement session. I expect this is one reason why Bruce always recommends the one-time technique, a second being that it avoids the assumption that the left and right channels are identical.

The results are the same, as all it is trying to do is to calculate what signal losses at the bottom end and high end are attributable to the sound card and should be added back into the measurement of the rest of the system. Whether it calculates this on the fly using a loopback connection all the time, or whether the calculation is done once, the results are the same, provided you don't change sound cards.

Hope this helps,
Bill

[hoofenshnoz]

Very helpful Thanks!
So just to clearify, even though you only run the calibration test thru one channel. Would it add the calibration to both L and R channels?

[laser188139]

So, with a loopback all the time on the left channel, while taking measurements through the right channel all the time, one is making the assumption, almost always valid, that the two are identical. I've not even worried about this on the desktop computer I use for my measurements.

If you use the loopback connection one time to build a sound card calibration file, you might as well calibrate using the channel you will use for your measurements. (If you are really worried about this assumption, you could do a calibration on one, then on the other, and compare the two calibration files.)

I shouldn't have mentioned it; forgive the mathematician in me. The issue of left/right channel being identical arises when one follows the idea elsewhere in the forum that one can calibrate through the mic pre-amp to subtract its effects from the measurement. When going through an external device with adjustable controls, the risk that the two channels are not identical far exceeds the benefit of any small accuracy improvement.

Note this is all a discussion of which channel on the soundcard is used for passing information; REW uses one mono channel to send test signals and to receive microphone input. Not to be confused with which speaker channel is being measured. Generally one starts with a Y connector at the receiver/preamp to drive both audio channels identically. Later one can disconnect the left or right from the receiver/preamp or speaker to measure left or right speakers or subwoofer in isolation.

Bill

[laser188139]

To follow-up on my first response, if you look in this other thread, you will see brucek re-iterate his recommendation that the "Use left channel" and loopback during measurement is only for checking system delay.