HOLMImpulse: Measurements in practice

I am struggling to get consistant measurements, the first file shows the same speaker (Saba) measured 3 times with exactly the same setup, done about 10 seconds apart.

The second one shows three different mids measured. the measurement of the Saba in the second file looks like it may be right but who knows

The bottom line is I'm getting wildly different measurements & I have no idea why.

I am running ECM8000, through a small Behringer mixer through the computers soundcard.

Any ideas?

thanks for the advice re screen shots
 

Attachments

  • Saba x 3.JPG
    Saba x 3.JPG
    357.7 KB · Views: 442
  • Test Saba, Seas Kef.jpg
    Test Saba, Seas Kef.jpg
    397.2 KB · Views: 422
DQ828 said:
Any ideas?

All six of those traces seem to be contaminated by a secondary input signal .
-The source of contamination could be caused by a few things ;

(i) Your (input) is running in mono mode creating a feedback loop .
(ii) Your card isn't fully duplex ( though most are these days ).
(iii) You've got "input monitoring" turned on ( within your sound control panel ) creating a feedback loop .
(iv) You are somehow adding in extraneous sounds onto the line level input .
- Most likely originating from the computers builtin mic if it's not muted .

- One solution; Buy an inexpensive Behringer UCA-202 USB card ( it uses the computers own sound drivers ) .
- Just make sure to disengage the "monitoring" switch on the side of the UCA-202 ( again,,, to avoid creating feedback loops ) .

<> EarlK
 
Last edited:
Well I think I have found the problem, I recieved the new ECM8000 mic & M-Audio Pre today. I decided just to try the mic to see if the otherone was a problem. I appear to be getting measurements that make sense. Now all I have to do is make sense of them. I would greatly appreaciate any input you are willing to give.
I will introduce the M-Audio pre on the weekend when I have more time. It should also give me consistancy when I move from the PC to laptop.

I have posted some of the screen shots in the hope you may be willing to give me some input. I have also shown the setup, this setup will only be used when I'm playing around as it obviously wont give me proper readings with the speaker sitting on the floor.

I have some questions I hope you are willing to answer

  1. I appear to get almost no reading below 100 Hz & all drivers rise sharply from there. Is that because of the way I have got the speakers sitting on the floor?
  2. Is it possible to get good measurements below 100Hz?
  3. It looks to me like I'd throw the Bass driver in the bin, & cross the mid Tweeter over at 6 kHz then cross the mid at 800Hz with another Bass driver, assuming these where proper measurements.
  4. The impulse time window appears to be telling me that the drivers are not lined up correctly, with the bass & tweeter about 25Ms apart, which would be audible?
  5. Phase, well who knows?
  6. If you can see anything you'd like to comment ok please go right ahead.
David:D
 

Attachments

  • Saba T,M & B, Impulse Time Window.JPG
    Saba T,M & B, Impulse Time Window.JPG
    464.4 KB · Views: 380
  • Saba T,M & B 10 smooth.JPG
    Saba T,M & B 10 smooth.JPG
    421.5 KB · Views: 371
  • Test Setup.jpg
    Test Setup.jpg
    362.2 KB · Views: 372
  • Saba Mid, Tweeter, Bass.JPG
    Saba Mid, Tweeter, Bass.JPG
    436.5 KB · Views: 180
I would measure the speaker very close (<1cm) to see what is happening. A speaker without baffle will fall 6 db/octave (acoustic shortcut), and the parallel floor and ceiling will cause major peaks and dips in the modal region (< 200Hz). So it could be that you have a major dip around that frequency, and that dip combined with the 6db/oc falling response could result in a steep acoustical highpass filter.

Measuring the speaker very close almost completely eliminates both phenomenon’s. If you still have the same effect, then there something wrong with your measurement setup (mic pre-amp?)
 
I would measure the speaker very close (<1cm) to see what is happening. A speaker without baffle will fall 6 db/octave (acoustic shortcut), and the parallel floor and ceiling will cause major peaks and dips in the modal region (< 200Hz). So it could be that you have a major dip around that frequency, and that dip combined with the 6db/oc falling response could result in a steep acoustical highpass filter.

Measuring the speaker very close almost completely eliminates both phenomenon’s. If you still have the same effect, then there something wrong with your measurement setup (mic pre-amp?)

Thanks, once I get the program running again, I'll give it a go.
 
I would measure the speaker very close (<1cm) to see what is happening. A speaker without baffle will fall 6 db/octave (acoustic shortcut), and the parallel floor and ceiling will cause major peaks and dips in the modal region (< 200Hz). So it could be that you have a major dip around that frequency, and that dip combined with the 6db/oc falling response could result in a steep acoustical highpass filter.

Measuring the speaker very close almost completely eliminates both phenomenon’s. If you still have the same effect, then there something wrong with your measurement setup (mic pre-amp?)

What amount of dB do we need to adjust the very close woofer measurement to splice it back with the gated full range measurement ?