sweep vs captured

[sm52]

I calibrated the sound card. After that I looked at the 'Captured' tab. It seemed to me that the sample rate of the sweep is not the same as the sample rate captured. Is it so?

 

[John Mulcahy]

Difficult to tell from that image, phase shift also has an effect at low frequencies.

 

[sm52]

This is towards the end of the measurement. The same is the difference.

 

[sm52]

And this is the end of the measurement. The sweep is over and the capture continues. Or I don’t understand something.

 

[John Mulcahy]

The rate at which the waveforms beat is about right for 44.1 kHz vs 48 kHz. Worth checking the Windows default format sample rate for the input and the output if you are using an ASIO wrapper, since those use the underlying WDM drivers.

 

[sm52]

I thought so too. But everywhere is 48 kHz 24 bits. On the 'advanced' tab. All record devices are disabled except one. All playback devices are disabled except one. In the settings of the sound card panel 48 kHz, the ASIO buffer size is 1024.

 

[sm52]

Is there a way to save the measurement so that it stores information for the captured tab?

 

[John Mulcahy]

No, the captured data is not saved. Can you attach the mdat file for the measurement? May also be worth trying with everything set to 44.1 kHz.

 

[sm52]

John, do you have the ability to calibrate your sound card and take a screenshot like mine?

 

[John Mulcahy]

It looks pretty similar. I think it is just a visual artefact from the delay between the generated and captured signals. If you make a loopback measurement with the acoustic timing reference selected the delay is removed and the traces should align better.

 

[sm52]

I found out that the delay in the start of the capture graph is equal to the delay in the end of the capture graph. Interestingly, a sound card connected to the usb 2.0 ports gives a delay of about 77 microseconds, while when connected to a usb 3.0 port it gives a delay of about 97 microseconds.

it is just a visual artefact from the delay between the generated and captured signals

Ok. I will try to make a measurement with acoustic synchronization later and pay attention to this.

 

[sm52]

I played around with the sample rate settings and the ASIO buffer size of the sound card. Nothing depends on the sampling rate. Sweep's smallest Captured plot start delay is obtained if the ASIO buffer is 128 samples. If less, the delay increases, plus the sound card is unstable, it may turn off, or the SPL graph may have artifacts. If more, the delay increases. But in all cases the SPL and phase plots are the same.