Home Theater Forum and Systems banner

1 - 20 of 29 Posts

·
Registered
Joined
·
11 Posts
Discussion Starter #1 (Edited)
The basics:

I'm using an HTPC with VT1708B audio, connected through an Onkyo SR608 receiver. Polk R30 centers, R15 surrounds, and CS2 center. I'm in the process of rebuilding my sub, so I have no sub at the moment. I'm using the Audyssey mic that came with the receiver.

So here's the first graph:

40hz.jpg

The first thing you might notice is this big hump centered around 41-42 Hz. None of my speakers are generating any meaningful sound at that frequency. The purple graph is with my mains set to "large" in my receiver. I then changed "subwoofer" to "on" and set the crossover to 60 Hz, giving us the green graph. I did another at 80 Hz.

I thought hey, maybe something in my room is generating some sound at that frequency? It's not very quiet in here. So I muted the receiver and ran the measurement again. Gave me this:

muted.jpg

Doesn't seem to point to ambient noise, but now I get the crazy ramping response, showing an increase of 35 dB from 200 to 20,000 Hz with zero actual change in the sound in the room.

Here's another gotcha. I have a sound meter app on my Nexus S that shows a complete spectral chart. During the full range sweep, the peak hovers around 80-100 Hz until it starts ramping up smoothly. Running a 40 Hz test tone from the PC gives a peak in the 80-100 Hz range too with plenty of SPL. Smartphone mic error? PC problem? It certainly doesn't feel like 40 Hz, it feels more like 80.

What's going on here? Are these microphone artifacts? They can't just be an uncalibrated mic because we don't see the peak from ambient sound, right? I'm sure I'm missing a ton of information here.

Bonus Q: Big peak around 500 Hz and below 300 Hz. How do I approach this?
 

·
Premium Member
Joined
·
15,054 Posts
The basics:

I'm using the Audyssey mic that came with the receiver.
Hi there, this is your first problem. you can not use just any mic with REW so this reading is pretty much useless unless you can find a calibration file that will correct the readings with REW you can only use ones that are designed to work with REW.
 

·
Elite Shackster
Joined
·
7,142 Posts
Hi there, this is your first problem. you can not use just any mic with REW so this reading is pretty much useless unless you can find a calibration file that will correct the readings with REW you can only use ones that are designed to work with REW.
Given Audyssey mics are designed to measure pretty accurately anyway, would it be possible to use it to take a 1m measurement, possible with the sub outside, to extrapolate a more realistic interpretation of what the mic actually measures. From that you could possibly get a better idea of how accurately the mic is.

Of course, the subs natural roll off and the fact we dont know whether or not a C weighting would apply would make it difficult to know if the results can be trusted or not. I dont know anything really of measuring mics etc, but maybe there is a way to use REW to test the accuracy of the Audyssey mic?
 

·
Registered
Joined
·
5,772 Posts
Unless you can get two measurements, one with a calibrated mic and one with the Onkyo, then you really do not know what is happening.
I have the Onkyo mic as well as my reference but do not have an adapter to plug it in.

If I get time and an adapter, I will try to get a comparison.

Cheers,
Bill.
 

·
Premium Member
Joined
·
15,054 Posts
Audyssey mic's are not very accurate by them selves and each receiver also has a built in calibration for the mic. Without it the readings are not going to be very helpful.
the only mic's that REW has calibration files for is the radio shack SPL meters, The Galaxy CM140 SPL meter and the Behringer ECM8000 mic.
 

·
Registered
Joined
·
11 Posts
Discussion Starter #6
Given Audyssey mics are designed to measure pretty accurately anyway, would it be possible to use it to take a 1m measurement, possible with the sub outside, to extrapolate a more realistic interpretation of what the mic actually measures. From that you could possibly get a better idea of how accurately the mic is.

Of course, the subs natural roll off and the fact we dont know whether or not a C weighting would apply would make it difficult to know if the results can be trusted or not. I dont know anything really of measuring mics etc, but maybe there is a way to use REW to test the accuracy of the Audyssey mic?
Well there is no sub at the moment! Only some Polk R30 tower speakers which don't put out lots on the low end. I had a fairly large sonosub but I dismantled it and will be putting it into a box tomorrow. I would have to use a laptop, but I don't have a USB sound device with line in. I could try, though.

What's the cheapest way I could get either a measurement on this particular mic, or get a microphone suitable for calibration purposes?
 

·
Registered
Joined
·
35 Posts
Hi there, this is your first problem. you can not use just any mic with REW so this reading is pretty much useless unless you can find a calibration file that will correct the readings with REW you can only use ones that are designed to work with REW.
I kinda have a problem with this statement.

I use an AKG D880 mic with a generic mic calibration file. Seems to work great for what I'm doing.

Granted I've tried an HK EZSet/EQ mic as well and its horrible. I imagine that ANY AVR calibration mic would be useless. I'm just not convinced that you HAVE to use specific performance mics to be able to use REW for a variety of purposes. Especially if your just a beginner.

Not only that but the OP's Freq response graph is deceiving because his graph goes from 15Hz to 25KHz.

If he scaled the graph from 20Hz to 200Hz that "giant bump" wouldn't seem so severe.


Test,
 

·
Registered
Joined
·
11 Posts
Discussion Starter #8 (Edited)
OK! I went out and bout the Radio Shack sound meter. I took measurements with the Audyssey with C-weighting on and off, then the RS meter with the calibration file enabled with the fronts set to small (crossover 80 Hz) and again with the fronts set to large. Remember, I have no subwoofer.

The RS meter is still showing the big hump at 42 Hz, within 3 dB of the output in the 100-300 Hz range.

many.jpg

So this raises lots of questions:

What's up with the 40Hz hump? Why is output in the 100-300 Hz range so low? Why is there a huge rise from 3000-8000 Hz and such a large dropoff afterwards?

I will admit I am using the headphone and mic port on a laptop, but I did do the sound card calibration.
 

·
Registered
Joined
·
11 Posts
Discussion Starter #9
Oh, and is there a way to maybe generate a calibration file for the Audyssey using the calibrated measurement of the RS meter?
 

·
Registered
Joined
·
35 Posts
OK! I went out and bout the Radio Shack sound meter. I took measurements with the Audyssey with C-weighting on and off, then the RS meter with the calibration file enabled with the fronts set to small (crossover 80 Hz) and again with the fronts set to large. Remember, I have no subwoofer.

The RS meter is still showing the big hump at 42 Hz, within 3 dB of the output in the 100-300 Hz range.

View attachment 32249

So this raises lots of questions:

What's up with the 40Hz hump? Why is output in the 100-300 Hz range so low? Why is there a huge rise from 3000-8000 Hz and such a large dropoff afterwards?

I will admit I am using the headphone and mic port on a laptop, but I did do the sound card calibration.

Try changing the parameters of your graph. Try 30HZ to 200Hz no smoothing.
Since you have no subwoofer it looks like you don't have a hump, but rather a dip at 55Hz.
Your mains seem to be rolling off at 44Hz.

Test,
 

·
Registered
Joined
·
35 Posts
Great!

Now, we'll need some more info. Whats your room layout look like?
Where are your Mains in relation to the nearest side walls, rears walls?
Have you tried moving your Mains side to side a few feet or forward and back?

Try a few different locations and take measurements after each.
Find the one location that smooths out your graph best.

Because you have no sub, I would think you'd want to set your x-over at 60Hz.

Test,
 

·
Registered
Joined
·
11 Posts
Discussion Starter #13
living-room.jpg

The living room area is 14' x 22.5', with a 39-inch opening to a 12x13 kitchen.

The mains are about 4" off the back wall. They're front-ported so it shouldn't matter much. I haven't tried moving them around, but I've moved the mic around the room and it doesn't change the response curve a whole lot.

I'm running the speakers now with the fronts set as Large, but it doesn't matter much because I'm going to be finishing the sub and installing it tomorrow. It's a 17" cube that I hope will work decently in between the two chairs.
 

·
Premium Member
Joined
·
15,054 Posts
I kinda have a problem with this statement.

I use an AKG D880 mic with a generic mic calibration file. Seems to work great for what I'm doing.

Granted I've tried an HK EZSet/EQ mic as well and its horrible. I imagine that ANY AVR calibration mic would be useless. I'm just not convinced that you HAVE to use specific performance mics to be able to use REW for a variety of purposes. Especially if your just a beginner.

Not only that but the OP's Freq response graph is deceiving because his graph goes from 15Hz to 25KHz.

If he scaled the graph from 20Hz to 200Hz that "giant bump" wouldn't seem so severe.


Test,
Believe what you will but the facts are you have to use a mic that has a calabration file asociated with it. You can make a file but you must have the mic's frequency graph from the manufacturer and most do not suply that. Even John who is the maker of REW states this very clearly in the user files for REW.
 

·
Premium Member
Joined
·
15,054 Posts
OK! I went out and bout the Radio Shack sound meter. I took measurements with the Audyssey with C-weighting on and off, then the RS meter with the calibration file enabled with the fronts set to small (crossover 80 Hz) and again with the fronts set to large. Remember, I have no subwoofer.

The RS meter is still showing the big hump at 42 Hz, within 3 dB of the output in the 100-300 Hz range.

So this raises lots of questions:

What's up with the 40Hz hump? Why is output in the 100-300 Hz range so low? Why is there a huge rise from 3000-8000 Hz and such a large dropoff afterwards?

I will admit I am using the headphone and mic port on a laptop, but I did do the sound card calibration.
If the laptop alowes you to change the mic input to a line input then your good to go. also you must load the calabration file for the RS meter, have you done this also?
 

·
Registered
Joined
·
35 Posts
Believe what you will but the facts are you have to use a mic that has a calabration file asociated with it. You can make a file but you must have the mic's frequency graph from the manufacturer and most do not suply that. Even John who is the maker of REW states this very clearly in the user files for REW.
I agree if your using an EQ you want the actual freq response for your room so you can make changes to the proper values.

However, lot of folks want before and after comparisons to changes they make.
If you have a peak, you have a peak. If you have a dip you have a dip. For some like me, it doesn't matter if the Mic is a little off. The actual freq is not what I'm after. Its the changes between location, treatments, etc..
As long as the Mic your using is consistent and predictable you can use REW with ANY decent pro audio mic and get good results.

For example, I have used my HK calibration mic and my AKG D880.
Even though the freq response and SPL's differ between those 2 mics (all things being equal) the SPL curve is very similar.
Both show the same thing. a nasty peak round 80Hz.
I can use that data pretty effectively as is.

But, if i had BFD and needed to address certain eq values, then YES I agree you'd need the mic calibrated with a specific file.

Test,
 

·
Premium Member
Joined
·
15,054 Posts
Ahhh but the mic its self could be reading wrong and give you peaks or dips where their is not.
 

·
Registered
Joined
·
35 Posts
Ahhh but the mic its self could be reading wrong and give you peaks or dips where their is not.
Ok tony.

I don't want to hi-jack the OP's thread too much so I'll leave it at this:

Your right, maybe the dip is at 76Hz instead of 80Hz. I can't imagine a good quality mic showing dips or peaks where there really aren't any.
Maybe an AVR calibration mic would. Or a cheap radio shack unbalanced 1/4 inch.
If all you want to see is a broad brush Peak/Dip, general overall response I don't see any issues with using a decent Mic with a generic calibration file.

OF COURSE it would be better to go out and buy the "approved" and "calibrated" mics but not everyone can do that and not everyone actually needs that kind of accuracy.

I guess it's up to each person and their objective.
REW has helped my setup considerably.
If someone had told me that REW would useless with out the mentioned calibrated mics I never would have tried.
That would have been a bad thing.
I never would have began to learn and my setup would not sound as it does today. Which is pretty good given my limited equipment.

Test,
 

·
Registered
Joined
·
11 Posts
Discussion Starter #19
OK this thread took a weird turn ;)

I built the sub and have a few new graphs. I did some open-air measurements and the response looks identical to the model, which is great. I did one on-axis and one from 90 degrees off out of curiosity. That's of course what you see up top.

Then I did a 10-200 sweep with the laptop output plugged right into the sub. That's the dark red.

Then I hooked it up through the receiver, set it to sub on fronts small and did one 10-20k sweep with the sub turned on and another with it switched off.

sub in room open air.jpg

Then I did another sweeps. I left the "sub turned off, fronts small" and "sub turned on, 80 Hz xover" graphs up. Then I turned the sub off and set the fronts to large and did another. Huge difference from 45 to around 100, but I'm still seeing that huge peak at 42. I guess it doesn't matter because I will be running fronts as small but it's still curious.

sub on vs off no xover.jpg

Any comments about the current working (lime green) trace? Looks like low response in the 150-350 range and a couple of peaks (1500 and 6000).
 

·
Registered
Joined
·
11 Posts
Discussion Starter #20
Also, quick Q: I'm using the RS SPL meter set to A-weighting, the RS calibration file, and "C-weighted" unchecked in REW. Is this correct?
 
1 - 20 of 29 Posts
Top