REW is GREAT!

[Tong Chia]

REW is a great program. I used it to analyse the performance of the SVS AS-EQ1
subwoofer controller based on Audyssey MultiEQ.

Many thanks to JohnM for a tool that just works and to brucek's helpful posts
that helped me avoid the potholes.

Measurement setup
EMU 0404 USB with calibrated Earthworks M30. Shockmounted on a Rycote INV-7
Apple Macbook running WinXP SP3

Sub
Velodyne HGS18MkII crossed over to mains at 90Hz

Measurement.
8 positions around the listening area, in the exact points the Audyssey measurements
used.

Green plot is before. Purple is after.
The results look decent but could be better. The peak at 130Hz is from the mains.
There is a slow 6db/octave rolloff from 50Hz down is a bit of a mystery. My current theory is
one of the positions is adding unusual bias. Deep Red graph in the individual plots.

This is a thankyou note and to describe what I am doing. I have the room details if anyone is interested.

Before (green) and after (purple) EQ plot



Individual responses of the 8 positions after EQ. The red laggard is the position 3 ft behind the listener.


Individual responses of the 8 positions before EQ.

[Wayne A. Pflughaupt]

Wow, looks like a reverse house curve. That can’t sound good! “Before” response looked pretty good, actually; just the dip south of 50 Hz needed some boost. A single parametric filter could have taken care of it.

Regards,
Wayne

[Tong Chia]

Wayne A. Pflughaupt wrote:

Wow, looks like a reverse house curve. That can’t sound good! “Before” response looked pretty good, actually; just the dip south of 50 Hz needed some boost. A single parametric filter could have taken care of it.

Regards,
Wayne

I used the SMS-1 previously and it had no trouble handling the dip at 50Hz. The dip is
actually caused by the mains cancelling out the sub. I did a few quick sweeps of just the
sub alone and the dip is no where as deep. I will take a closer look over the weekend.

The main object of the exercise is to characterize the AS-EQ1 using REW. The calibration
certificate from the Audyssey software provides a before and after trace as to what it
thinks is the response (attached). It looked optimistic to me. I use tape markers and
a plumb line to make sure all 3 mics (Onkyo 885 preamp, AS-EQ1 & REW) were in the
same spot.

Audyssey's cal can suffer the "garbage in-garbage out" syndrome if some of the
positions are vastly different. I am hoping REW can help avoid the worst of the trouble spots.

I will rearrange the listening area if need be.

Overall I am pleased with the AS-EQ1 and REW. It is still in the early stage and I have a lot
more work to do to establish a consistent baseline measurement with REW as well effort
selecting the AS-EQ1 measurement positions.

[thewire]

Wayne A. Pflughaupt wrote:

Wow, looks like a reverse house curve. That can’t sound good! “Before” response looked pretty good, actually; just the dip south of 50 Hz needed some boost. A single parametric filter could have taken care of it.

Regards,
Wayne

This is the subwoofer only? Looks like something is wrong here.

[Tong Chia]

REW trace is sub + main.

AS-EQ1 trace is sub only.

I will rerun the REW sub only traces again. I clobbered it while trying to save the trace.

[clubfoot]

Are your mains to small and you shouldn't have the cancellation/dip issue.

[OvalNut]

Are you sure the mic calibration file is loaded in REW? The slow shallow roll off toward the lowest frequencies looks perhaps like a measurement without a mic/SPL meter calibration file in use.


Tim

[Tong Chia]

clubfoot wrote:

Are your mains to small and you shouldn't have the cancellation/dip issue.

As a workaround, I set the crossover to 130Hz, while I start looking at the source of the problem

[Tong Chia]

OvalNut wrote:

Are you sure the mic calibration file is loaded in REW? The slow shallow roll off toward the lowest frequencies looks perhaps like a measurement without a mic/SPL meter calibration file in use.


Tim

That was the first things I looked at. A cal file is physically loaded (REW popup). Earthworks says that
their mics are flat below 1kHz and the thing is speced to 5Hz. Their cal file only
has data down to about 700Hz.

I am thinking about getting the mic independently calibrated.

[clubfoot]

Tong Chia wrote:

As a workaround, I set the crossover to 130Hz, while I start looking at the source of the problem

Set your xover to 80 and your mains to small and take a measurement without filters, it may surprise you.