ARTA

Do two loopback measurements


With the wrong polarity, I figured it out this way. In the control panel of the sound card, I turned on the inverse of the microphone input. But I want to get to the truth. I do not quite understand your advice. But I checked such a thing. When I invert both inputs from the microphone and from the speaker in the sound card's control panel, I get a negative impulse again. If I invert only the microphone input, I get a positive impulse. These facts help the investigation of who is to blame?
 
I do not quite understand your advice.

With a dual channel measurement, the software will deconvolve the DUT (=device under test) signal from the REF (=reference / loopback) channel. If your soundcard inverts the polarity, it will probably do it on both channels. The deconvolution result should therefore have normal polarity.

With a single channel measurement (DUT channel only), there is no deconvolution. If your soundcard inverts the polarity, you will see it in the result.
 
I found out who reverses the signal polarity. I connected another microphone to my sound card. It does not invert polarity. Then another one. And this one inverts. These two and mine are capacitor. Two inverts polarity, the third does not invert. Therefore, the sound card is not to blame. The microphone remains. And the question remains. If it changes the polarity of the signal not by 180 degrees, but by 120 or 150, will the phase-frequency characteristic of the speaker obtained by such a microphone be accurate?
 
I realized that I can use such a microphone and trust its data. If he biases the phase data, then he does it for all speakers the same way. My question was about data analysis in Vituix. Namely, the phase-frequency characteristic. It matters not only +180 and -180.
If it is not possible to find out how much the microphone shifts these phases, then for now, we must forget this question.
In ARTA, the impulse measurement time step is 0.021 ms. This is 7.2 mm. Those. if you move the microphone 7 mm, ARTA will say add a delay time of 0.021 ms. This is very small. But the phase-frequency characteristic can shift in the HF range by 45 degrees. Is there any criterion to consider that during the measurement I set the delay time correctly? For example, according to the figure of the phase response during the preliminary measurement of the frequency response - FR2?
 
Disabled Account
Joined 2019
Hi,

A noob question with Arta and the soundcard, please ?

I want to measure my loudspeakersin room at 1 m axial and at listening position: each louspeaker then both at the same time, what the best please :

- I plug a wire between the left output and the right or left compact disc input of my single ended preamp, then I inject a swept tone from Arta ? I use a single mic with the phantom 48v power of the soundcard.

Or

- I putt in my library an external .wave swept tone I play from my DAC which is certainly quieter than my Pre Tracker E-MU soundcard -though powered by an Ifi low noise dart- . From there easy to measure the L or R channel I want or both at the same time.

In both case how to setup the volume pot of the preamp please ? Have I to measure with the other channel of the soundcard loopbacked Or use the left and right outputt towards the preamp of my Hifi ?

Many thanks, I didn't find a clear answer fo this starting point in the last Arta manual. I assume the problem remains if I want measure a single driver when I will make the probe voltage protection thingy : ho to setup the volume plot for spl curve at acurate db spl scale on the Arta windows ?

Many thanks if any links or inputs for that
 
What is your measurement setup?
 

Attachments

  • screenshot3.PNG
    screenshot3.PNG
    29.6 KB · Views: 352
My measurement setup is:
desktop computer with integrated sound card, the computer is not in the same room as my hifi system, so I have a 3m connection cable from the computer to the microphone preamp and a 10m cable for the microphone.

setup.jpg


I prepare for another of the usual measures from the listening point and ..
I get this measure.



I forgot to connect the microphone cable to the microphone preamp, so I measured without a microphone.
Intrigued by the upward trend of the frequency response, I also disconnect the microphone preamp and repeat the measurement with overlay, so I am measuring with the 3m cable connected to the computer sound card, yellow without mic and green without mic preamp.

An externally hosted image should be here but it was not working when we last tested it.


Here now the response curve rises less.
So I reconnect everything and do the first measurement with overlap,
yellow without microphone and green all connected measured from the listening point.



So I was thinking that the microphone preamp doesn't have a linear frequency response.
Should I think that the overall result of the measurement is not what I see, in reality the frequency response has a decreasing trend?
any thoughts?