[Krumlov]

Hi Guys,

Am thinking of using REW to see how my room "looks" acoustically. I currently have an RXV-2400 and I have used Yamaha's auto setup feature to set the levels of the speakers including the sub.

I understand the theory of what REW does and the necessary hardware required to do it. I am curious as to why the notes say to limit the RS measurements to 3kHz? What is the downside of using it up to 20kHz? Can you also quantify the downside? I say that because there are levels of dedication in this "hobby" and I would not say I was a hardcore audiophile! I am just interested in what the response of my room is and how well the Yamaha has set the levels. Will the RS SPL do a better/worse/same job as the Yamaha's internal setup?

My thinking at the moment, depending on the response to this post, is to start with the Behringer UCA202 and my RS SPL and if I want to move up from there purchase the Galaxy CM-140. This would be the extent of the system that I would looking at.

Thanks!

 

[JohnM]

The main problem is a fairly pronounced hump in the frequency response centred at roughly 5kHz, on top of which is the C weighting response itself which is down about 11dB by 20k. The meters also show a lot more unit to unit variation at high frequencies. The CM-140 is a lot more consistent and, if calibrated, can be used for reliable full range measurement.

 

[Krumlov]

The main problem is a fairly pronounced hump in the frequency response centred at roughly 5kHz, on top of which is the C weighting response itself which is down about 11dB by 20k.

Thanks John. However I am still trying to understand how bad that really is. Yes it is by no means perfect but if I used it would it be a better/worse/same result as the Yamaha's internal setup?

 

[laser188139]

I don't have any experience with the Yamaha microphone and built-in system. But in this graph I compared my early measurement with the RS meter, and then my first measurement with the calibrated EMM-6. In blue, you see the result with the RS meter, which gives the impression that Audyssey in my Denon left a lot of unevenness in the response. The red curve is the identical setup, no changes whatsoever except perhaps volume setting, but now measured with a calibrated EMM-6 microphone. The curve follows the tendency of the published Audyssey target curve.

I've not seen a statistical study of a large range of RS meters, but it is hard to believe their manufacturing tolerances are tighter than the curves of the ECM8000 and EMM-6 curves that Herb generated. I would not want to make any adjustments at the higher frequencies based on the RS meter.

Bill

 

[Krumlov]

But in this graph I compared my early measurement with the RS meter, and then my first measurement with the calibrated EMM-6. In blue, you see the result with the RS meter, which gives the impression that Audyssey in my Denon left a lot of unevenness in the response. The red curve is the identical setup, no changes whatsoever except perhaps volume setting, but now measured with a calibrated EMM-6 microphone. The curve follows the tendency of the published Audyssey target curve.

Thanks Bill. The way you have written this suggests that Red Curve is somehow better than the blue curve. To my uneducated eye they are very similar. Certainly up to 2k they follow the same pattern just slightly separated? If I had a calibration file for above 1K then I assume that curve could be corrected? So my next questions are:
How many dB different could the different SPL meters be? (I note that the graphs you linked are up to +/- 5dB at the extremes) (Thus how far out *could* the meter be from the cal file)
Can the RS SPL Meters be used if a custom cal file was obtained?
Why has it been possible to create a generic cal file for the Galaxy CM-140?

Thanks for indulging me!

 

[gsmollin]

The real problem with the RS meter is that the microphone is too large to measure frequencies above 5 kHz with any accuracy. There is no calibration curve you can apply, because the microphone becomes very directional, and the off-axis response is unpredictable. Given you are in a "surround" field, and are also subject to specular reflections at high frequencies, you can't control the direction of the sound and can't predict the result. You need a measurement mic with no larger than a 1/4 inch mic capsule, and preferrably 1/8 inch if you are serious about measuring high frequencies.

 

[Krumlov]

You need a measurement mic with no larger than a 1/4 inch mic capsule, and preferrably 1/8 inch if you are serious about measuring high frequencies.

Thanks. So I assume the Galaxy CM-140 fits this requirement then?

 

[gsmollin]

Thanks. So I assume the Galaxy CM-140 fits this requirement then?

They only claim 8 kHz in their specs. I use a Behringer ECM8000 microphone, plus a cal curve. Parts Express has an inexpensive mic too. Then you can spend some real money and get the Earthworks M30. All these mics need a preamp with phantom power. I use the Behringer 2-channel Euromixer, although they are showing the Behringer Xenyx 802 now. For not much more than the CM-140 you can get response to 20 kHz. Of course you still need a SPL to set levels.

 

[JohnM]

The CM-140 measured responses look reasonable, allowing for the roll-off of the C-weighting.

 

[gsmollin]

The CM-140 measured responses look reasonable, allowing for the roll-off of the C-weighting.

I'm glad you qualified that statement, because I don't think -15 dB at 20 kHz is reasonable performance, if I am interested to measure 20 kHz. There is also a 5 dB variance between random and 0 degree incidence, which indicates the omnidirectionality of response has been lost at 20 kHz. I'd say this meter is good to 12 kHz, which is good for an SPL, but the OP is asking about full range performance, and the last audio octave requires equipment optimized for that. In any event, the Parts Express and Behringer equipment is pretty inexpensive, I think. I know I've been satisfied with mine, and for casual use it is a very good value.

 

[Krumlov]

I use a Behringer ECM8000 microphone, plus a cal curve.

Do you mean the cal curve from here or did you get one for your specific mic?

... the Behringer Xenyx 802 now. For not much more than the CM-140 you can get response to 20 kHz. Of course you still need a SPL to set levels.

So you are suggesting the ECM8000 (AUD$79) and Xenyx 802 (AUD$89) makes AUD$168 which does compare favourably with a calibrated CM-140 from Cross Spectrum delivered at AUD$172? I am assuming that the cal file from this site would be adequate? If not then that would be an additional cost.

Combine all that with a Behringer UCA202 (AUD$49) and an Ebay SPL meter (AUD$40) for a grand total of AUD$257. Does that set up suit my initial query of a setup for full range measurements?

 

[gsmollin]

I used the generic cal curve from here. If I thought I needed custom cal, I would also need a better mic because even the Behringer and PE mics are becoming directional at 20 kHz. If you really need accuracy to better than 3 dB at 20 kHz, you need the M30. I often use two mics and average them. REW also has an averaging facility for various positions.

The Behringer mixers work well, and I take it you have no aux-in on your computer so you are adding a Behringer USB aux input. That looks intriguing, and I may get one myself. Yes, a SPL meter is still important for setting levels, although not essential. You can always set levels by ear. When the pink noise is drowning out the competing outside noises, then it is loud enough to measure.

OBTW, a word about outside noise. You need to turn off all competing noise sources. Fluorescent lights, lamp dimmers, HVAC, fish tanks, computers, TV and radio, refrigerators and freezers, washers and dryers, everything you can find. Then you may need to wait for the middle of the night. I remember a professional reviewer who complained that the Yamaha setup wasn't working well, and he had left his projector running.

P.S. Don't forget the microphone cables. You need balanced lines with XLR connectors.

 

[Krumlov]

The Behringer mixers work well, and I take it you have no aux-in on your computer so you are adding a Behringer USB aux input.
[snip]
P.S. Don't forget the microphone cables. You need balanced lines with XLR connectors.

The REW cabling setup page mentions "an external USB audio interface with integral microphone preamp with phantom power" but I haven't been able to find a sound card which has such a capability. Anyone have any ideas so that I could simplfy things by removing the Xenyx 802?

And yep wouldn't want to be short of mic cable to put it all together! :doh:

 

[JohnM]

The REW cabling setup page mentions "an external USB audio interface with integral microphone preamp with phantom power" but I haven't been able to find a sound card which has such a capability. Anyone have any ideas so that I could simplfy things by removing the Xenyx 802?

M-Audio MobilePre USB would look to fit the bill there, but the Xenyx 502 (which nowadays has phantom power) and a cheap and cheerful stereo USB soundcard would likely work out cheaper.