Something's Wrong!

Certainly, if you are observing clipping in the red lights on the preamp/mixer, or if REW is warning of excessive levels on the input channels, then you need to tweak the input volume downward at either the preamp or the computer. Whenever you change these, you need to redo the Calibrate SPL step, because you have changed the REW perceived electrical levels while the physical audio levels were unchanged. Because I never know if someone or some other application has changed the levels, I find it best to run the Calibrate SPL step at the start of each session.

Of course, if you are just interested in making the graph have the normal average level of 75dB, you can change the offset to the data to achieve the level you want to see, and can use the Add offset to data button to permanently modify that measurement.

Bill

perritterd said:

Ok Bill. I reduced the input vol. level down to .199 (which reduced the levels to about 20-21), re-ran the SPL check, and set the SPL to 75db. Here is the pic of the graph I generated (BTW, I had both sp. on for the SPL check-is this ok?).

Click to expand...

It doesn't matter per se whether you calibrate the SPL with one or both front speakers. Obviously, if you calibrate the levels to 75dB with both speakers and then you measure only one, the measured level will be lower than 75dB. If you calibrate the level to only one front speaker and then measure both together, the level will be higher than 75dB.

I assume you calibrated with both speakers on and then measured both together. Some of the variation you see in the graph comes from comb filtering between the two speakers, as Wayne mentioned above. You can eliminate this effect by measuring the left and right separately, but that can also uncover variation where one speaker shows dips that the other does not. You can also reduce its effect by using more smoothing.

I also assume the measurement was made with Audyssey on. In which case, much of the rolloff at the high end is a deliberate result from the Audyssey Reference curve.

Bill

perritterd said:

... I do not understand why the rolloff. ...

Click to expand...

Even with all the variation at 1/48 smoothing, the Audyssey On curve looks a little more even than Audyssey Off.

I have some curves that look very similar at the high end to yours. So some of that response may be natural rolloff at the high end from absorption in the room. I know in my room, I have furniture issues that can change the absorption/reflection of high end frequencies from session to session. I'm not sure you have anything to worry about. At my age, my ears don't reach that high.

If you are concerned, you can try individual measurements, 1-2' in front of each speaker with Audyssey Off, to verify how each speaker appears when measured near field. Some speakers have more roll off than others at the high end from the tweeters. If your receiver has preouts, you can measure these, comparing Audyssey Off to Audyssey On, with no mic calibration file, to see if and where Audyssey is maxing out on its boost at the high end. As I recall, its maximum boost is +9dB, so anywhere you see adjustments that high at the upper end, it may be running out of adjustment and not leveling as much as it would like.

Bill