• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 30 subscribers
  • Views 1036 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Parameters to measure audio quality

Toby
Mastermind 35030 points

Hi team,

This is a general audio question, with target frequency range of human voice.

I have the following system, and trying to improve the output audio quality.

I start the test by playing a frequency sweep from this video: https://www.youtube.com/watch?v=dU80Fagdy28

The Matlab script: https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/6/read_5F00_evaluate_5F00_PCM.m

The PCM output from the device: pcm_data_sweep3.csv

The resulting output and tiles from matlab:

>> read_evaluate_PCM
------- PERFORMANCE METRICS -------
SNR (dB) --- -3.429719e+01
THD (percentage) --- 2.435858e+01

 

I am wondering if this is the correct way to measure the audio quality? 

Do these values make sense?

Thanks,
Toby

  • 0
    TI__Mastermind 35030 points

    Hi team,

    Some additional questions to help my understanding if the test setup/process/measurements need changes:

    What audio source (speaker) should be used to generate the input signal to the microphone? We are currently using the iPhone to generate the tone/sweep inputs to the microphone, but we are concern the frequency response of the speaker may not be the best to test our system and could be harming our results.

    What type of audio should be used to calculate the different metrics (SNR, THD, Frequency response): tone/sweep? For instance, calculation of THD using a sweep vs a tone is quite different.

    Is there a practical way to evaluate the frequency response of the filtering stage (decimation), without including inherited acoustic frequency responses of the speaker and microphone, we are thinking about inputting a “known” PDM signal of different frequencies? Do you have a way to help us generating this? What is the standardized / recommended way of evaluating the performance of the filtering stage?

    Thanks,
    Toby

  • 0 in reply to Toby
    TI__Mastermind 23030 points

    In my opinion giving anaccoustic input to the PDM Mic may induce the effects of the Amplifier of the sound source itself.

    I would suggest to have some kind of PDM Source to drive the system input. I know that Audio Generation such as manufactured by Audio Precision have such PDM sources present .

     A Sweep measuring SNR and THD is  good test.

    THDN vs amplitude is also a good test to check for distortion.

    https://www.ti.com/lit/ds/symlink/pcmd3180.pdf?ts=1706810164807&ref_url=https%253A%252F%252Fwww.google.com%252F

    Please refer to the datasheet of the TI ADC that can be used with a PDM Mic. The tests shown here could perhaps be used as a reference