Question: Why use 44.1 kHz audio sampling rate instead of 48 kHz in DNMS Teensy firmware?

[KevinAndani]

Hello,

I noticed that the DNMS Teensy firmware in this repo uses a 44.1 kHz audio sampling rate. From my understanding and industry standards, 48 kHz is generally recommended for professional audio applications, especially those involving video, film, or monitoring systems.

48 kHz provides slightly better anti-aliasing filter characteristics, better aligns with video frame rates, and offers a bit more headroom for processing without introducing artifacts. Meanwhile, 44.1 kHz is typically tied to audio CD standards and music production.

Could you please explain the reasoning behind choosing 44.1 kHz for DNMS audio monitoring? Are there specific hardware, software, or legacy compatibility considerations that influenced this decision?

Thank you for your insight and for the great project!

Best regards.

[hbitter]

Hello Kevin,

Yes, the DNMS Teensy does indeed use a sampling rate of 44.1 kHz. The reason for this is very simple. When we started the project, the underlying Audio Lib for the Teensy was fixed at 44.1 kHz. Later on, we switched to the 32-bit floating-point version of the audio library, which was also fixed at this rate. The ability to adjust the sampling rate in both libraries only became available later.
For our application, it doesn't really matter, but switching to a sampling rate of 48 kHz would require considerable effort and wouldn't improve things. We would not only need to change the sampling rate parameter for the Audio-Lib, but also need to recalculate the frequency correction, among other things.
The DNMS is not a class 1 SPL reference measurement system. It is a DIY system for measuring ambient noise, so there is no need to exceed 20,000 Hz. With a sampling rate of 44.1 kHz, however, the upper frequency limit of 20,000 kHz can be reached.
It was more important to switch to the 32-bit floating-point version of the Audio-Lib than to change the sampling rate. This makes it possible to utilise the full dynamic range of the microphone.

Best regards,

Helmut

[KevinAndani]

Hello @hbitter ,

Thanks a lot for the explanation about the 44.1 kHz choice — I understand the reasoning now.

I do have one follow‑up question out of curiosity: how did you determine the correction factors for the Z and A weightings in the firmware? And why did you choose to implement only these weightings?

I’m guessing part of the reason might be that this is a DIY project and Z/A cover the most common use cases, but I’m also wondering about the other standard curves. If I wanted to add additional weightings like B or C myself, how would I go about finding or designing the corresponding filter coefficients/correction factors for those?

Any pointers (papers, tools, or code references) would be greatly appreciated. Thanks again for sharing this project and for your time!

[hbitter]

Hello Kevin,

I was just wondering if you could tell me how the OIZOM commercial Class 1 noise monitoring project is going?

Kind regards,

Helmut

[KevinAndani]

Hi @hbitter, sorry for the late reply, and thank you for taking the time to raise this. We truly appreciate your effort.

We’re currently investigating why the Teensy firmware uses 44.1 kHz (checking Teensy Audio Library defaults, ADC/DAC/resampling behavior, timer/interrupt alignment, and any legacy/configuration constraints). We’ll assess the impact and effort required to move to 48 kHz and will follow up on this issue soon with our findings and recommended next steps (and a PR if a change is appropriate).