My first REW Measurement - DIY Speakers - Did I get speaker polarity wrong?

[nikb]

today, I made my first REW measurement. My goal is to check my kit-speaker build and then maybe improve my listening position and the room in general.

Ten years ago I built a pair of Troels Gravesen Seas 3 Way Classic mkII. When I finished the build I was so happy to get sound out of them that at first I did not notice that something was sounding off. Later I did not quite know what to do about it. At the time I was using a diy amp that had some errors, all of which did not help. I ended up listening to different speakers.

Now I have set up the speakers in my living room and measured them with REW. My first question is: Did I get the polarity of the chassis right or could anything be connected the wrong way? Is it possible to see that in the SPL graph?
Also: my living room is really bare with lots of echo - I think I need some carpet and some decoration on the walls.

 

[kemmler3D]

today, I made my first REW measurement. My goal is to check my kit-speaker build and then maybe improve my listening position and the room in general.

Ten years ago I built a pair of Troels Gravesen Seas 3 Way Classic mkII. When I finished the build I was so happy to get sound out of them that at first I did not notice that something was sounding off. Later I did not quite know what to do about it. At the time I was using a diy amp that had some errors, all of which did not help. I ended up listening to different speakers.

Now I have set up the speakers in my living room and measured them with REW. My first question is: Did I get the polarity of the chassis right or could anything be connected the wrong way? Is it possible to see that in the SPL graph?
Also: my living room is really bare with lots of echo - I think I need some carpet and some decoration on the walls.

View attachment 486383

AFAIK We can't judge polarity unless you also take a measurement with both speakers playing at once, in that scenario you will see extra dips where the two speakers cancel each other out.

However, the response here looks like the woofers aren't connected or maybe the crossover has a serious problem. 100hz should not be 20dB below 10khz, no matter what is wrong with your room.

Clearly something has gone wrong with these speakers or amp, inverted polarity is possible but it's not just a polarity problem.

Alternately, What mic did you use for the measurement here? Maybe it's not calibrated?

 

[solderdude]

Measure each speaker outside and look at the phase plot and compare the speakers.
When you have incorrectly wired one of the drivers you can also look at the frequency response (outdoors) and see if near the crossover frequencies the response is wrong.

 

[digitalfrost]

The measurements seem very strange to me. You have 20-30dB drop from 10khz to 50hz? That should sound anemic. Did you window the measurement? Can you share the .mdat file?

 

[Digital_Thor]

First off, check whether your soundcard invert the signal anywhere - some do, some don't.
Next, measure each driver individually and see whether one has a pulse that start by going up or down. If everything is in phase, the impulse will go up.
It'll take a little time, but read this... Or at least skim through...
https://www.roomeqwizard.com/help/help_en-GB/html/graph_impulse.html

 

[Digital_Thor]

Kinda looks like you somehow moved the slider somewhere and unknowingly asked the software to ignore some of the impulse, hereby getting a limited piece of the whole frequency response

 

[nikb]

Thanks for all the feedback - so clearly something seems to be off with the measurements. I have attached the mdat files to this post.

I live in an apartment building near a busy street, so I will try to fix my measurements before I try to measure the speakers outside. Also I have a considerable noise floor even when all windows and doors are closed in the room. And I listen at pretty low volume.

I am using a measurement microphone and mic preamp with calibration file, not a usb microphone like the umic. I have tried to calibrate the REW spl meter using a random app on my phone - tomorrow I will have a proper spl meter for calibration - todays db readings might be off.

But: There should be no considerable drop between 10khz and 50hz, at least according to the measurements of the speaker builder here: http://www.troelsgravesen.dk/SEAS-3-Way-Classic-mkII.htm. First I will do a measurement with both speakers at once. And I will recheck my measurement setup - yesterday I did a loopback calibration of my audio interface which looked strange to me (did not save the mdat for that).

 

[nikb]

Kinda looks like you somehow moved the slider somewhere and unknowingly asked the software to ignore some of the impulse, hereby getting a limited piece of the whole frequency response

What slider could that have been? I am doing some silly things while getting a grip on REW.

 

[nikb]

AFAIK We can't judge polarity unless you also take a measurement with both speakers playing at once, in that scenario you will see extra dips where the two speakers cancel each other out.

However, the response here looks like the woofers aren't connected or maybe the crossover has a serious problem. 100hz should not be 20dB below 10khz, no matter what is wrong with your room.

Clearly something has gone wrong with these speakers or amp, inverted polarity is possible but it's not just a polarity problem.

Alternately, What mic did you use for the measurement here? Maybe it's not calibrated?

I am using a different amp now, that I trust (purifi class D amp).

 

[Keith_W]

It appears that the polarity between L-R speakers is correct.

Method 1: compare the step response or impulse response of L/R speaker. Both are pointing in the same direction.

Method 2: sum the left and right speaker channels and look for comb filtering. There is none.

There are a lot of problems with your measurement, starting with that absolutely bizarre frequency response. A 20dB upwards treble slope like that with no bass would make your loudspeakers sound like a tin can. Do they sound like that? If they don't, then something has gone terribly wrong with your measurement. Please read this eBook.

 

[kemmler3D]

A 20dB upwards treble slope like that with no bass would make your loudspeakers sound like a tin can. Do they sound like that?

To be fair, OP did mention them sounding off, to the point that they didn't use them. I wouldn't listen to speakers, either, if they sounded like that graph looks...

 

[nikb]

Again, thanks for the help and the replies. I am still in the process of reading up on REW and how to use/interpret the graphs.

I have listened to the speakers again, and I came to the conclusion that there is no/not enough bass coming from the left speaker. I am suspecting that this is part of the bad response. I have moved them several times in the last years and something must have come loose.
I will try to correct that tonight and post updated measurements.

I also suspect that this will only be a part of the problem and that there are more issues lurking in the crossovers. First I will try to have both speakers sounding the same.

Also: yes, they sound anaemic and I kind of knew that, but I sometimes have a hard time trusting my ears (my hearing is actually fine). That is where the measurements will help.

 

[Keith_W]

It's not just your left speaker, it's both. And don't be in a hurry to take new measurements, you have big problems with both your speakers. If it was ONE speaker, I would suspect that a cable has gone awry, or a woofer has gone to heaven, or something like that. But it's both speakers - this suggests a fundamental design flaw. Maybe you didn't read the crossover diagrams properly, chose wrong component values, etc. IMO that seems most likely to me.

I would suggest you dismantle the speaker and inspect the crossover, make sure the components are as specified in the design. If in doubt ... you have an interface. You probably have enough DAC and amp channels to make a single active speaker for testing. The procedure to do this is rather involved, and requires you to download and learn lots of different software. It may be a bit too much for you given that you've only just picked up REW.

 

[nikb]

It's not just your left speaker, it's both. And don't be in a hurry to take new measurements, you have big problems with both your speakers. If it was ONE speaker, I would suspect that a cable has gone awry, or a woofer has gone to heaven, or something like that. But it's both speakers - this suggests a fundamental design flaw. Maybe you didn't read the crossover diagrams properly, chose wrong component values, etc. IMO that seems most likely to me.

I would suggest you dismantle the speaker and inspect the crossover, make sure the components are as specified in the design. If in doubt ... you have an interface. You probably have enough DAC and amp channels to make a single active speaker for testing. The procedure to do this is rather involved, and requires you to download and learn lots of different software. It may be a bit too much for you given that you've only just picked up REW.

I agree that I should thoroughly check the crossovers - I already have an idea about what could be wrong. Before that, I want to make sure that my REW setup is Ok and that my measurements can be trusted. I want to recalibrate levels and my audio interface in REW.

First Question: when measuring with REW, do you connect the audio interface directly to the amp or can I use the preamp line-in? If I am using the line-in, what level do I set the preamp to? Full blast or half way? My amp is powerful enough to blow the speakers - I have to be careful with the output level.

 

[digitalfrost]

Before that, I want to make sure that my REW setup is Ok and that my measurements can be trusted. I want to recalibrate levels and my audio interface in REW.

First Question: when measuring with REW, do you connect the audio interface directly to the amp or can I use the preamp line-in? If I am using the line-in, what level do I set the preamp to? Full blast or half way? My amp is powerful enough to blow the speakers - I have to be careful with the output level.

The stuff you are seeing is so severe, the absolute level or your soundcard calibration does not really matter. You could even ignore mic calibration even though that has impact on the measurement. But do not worry about these small things, absolute level is irrelevant and we can see your soundcard is fairly linear from the measurements you uploaded.

When measuring for applying EQ, I would try to get as close as possible to the real world setup (i.e. including preamp). If you worry about levels have the preamp in there. You do not need to measure super loud either.

edit: If you have other known good speakers, go and measure them so you will be confident your setup is creating good measurements.

 

[nikb]

Hehe, thanks for the heads-up! I will continue with the speakers and worry about REW later

 

[Juhazi]

Looks like your woofers are not connected... http://www.troelsgravesen.dk/SEAS-3-Way-Classic-mkII.htm

 

[nikb]

Opened the left speaker. The Woofer was connected all right!
So now I'll double check the crossover.

 

[nikb]

Here is a measurement of a pair of known good 2-ways.
Same measement setup as before (though levels are different). Still roughly 20db of difference between 100hz and 10khz.

 

[Pareto Pragmatic]

Here is a measurement of a pair of known good 2-ways.

Putting the speaker issue aside, is your room highly reflective? Lots of glass, bare walls and ceiling, hard floor? Because the only way I can get something like that in my room is to maximize my reflections. And I do have a lot of them.

If you can take three measurements it can help tell if this is room reflections. Point the speakers just to the outside of the mike, then straight into the room, then half way between. If this is a reflection issue, that upper frequency tilt should change.

Change, not disappear... unless you are very lucky.