Introduction
Previously, performing Speech Quality Monitoring (SQM) assessment on mobile phones was straightforward.
The setup involved a Jack port, a direct connection to the Digital Speech Level Analyzer (DSLA), signal capture, metric calculation, and Mean Opinion Score (MOS) scoring.
The process became complicated when the primary connection port on the devices transitioned to USB-C.
Initially, the assumption was that a simple USB-C to Jack adapter would suffice. Lacking expertise in electronics, the complexity of factors involved in selecting the appropriate adapter was initially underestimated.
While the initial focus was on Speech Quality, the growing challenges posed by new requirements, such as those in first responder and mission-critical communications, necessitated extending the analysis to end-to-end latency. Fortunately, the first adapter selected performed well in terms of Audio Quality and exhibited minimal latency drift. However, following the end-of-sale for the initial model, a second adapter, while offering good audio quality, yielded variable latency scores with significant delay differences, particularly during long calls. This outcome motivated a deeper investigation into characterizing latency costs and clock drifts.
This paper explains the components within a USB-C to Jack adapter, the necessary verification points, the methodology employed, the resulting conclusions for the adapter selection process, and the alternative solution proposed by Opale Systems for managing USB-C connections.
The USB-C to Jack adapter
We will focus on adapters that include a microphone. Note that not all adapters provide microphone capabilities. Attention must be paid to the adapter's audio characteristics to ensure it supports more than just a headset (i.e., microphone support).
Headset and microphone
TRRS CTIA or OMTP CTIA and OMTP are the two primary wiring standards for the TRRS (Tip-Ring-Ring-Sleeve) 3.5mm audio jacks used on smartphones and other devices to support stereo audio and a microphone (headsets). The only difference between the two standards is the swapping of the Ground and Microphone connections on the plug. R: Right, L: Left, G: Ground, M: Mic.
CTIA: LRGM. Most common today.
OMTP: LRMG.
DAC/ADC characteristics
Sample rate: The sample rate is the number of "snapshots" taken of an analog audio signal per second during analog-to-digital conversion. It is measured in Hertz (Hz) or kilohertz (kHz). The Nyquist-Shannon sampling theorem states that the sample rate must be at least twice the highest frequency you want to record. The standard for CD audio, for example, is 44.1 kHz.
Resolution: Resolution, often referred to as bit depth, determines the number of possible amplitude values (or volume steps) available for each sample. It's measured in bits (e.g., 16-bit, 24-bit). Higher bit depth means a finer representation of the audio signal's amplitude, resulting in a wider dynamic range and a lower noise floor.
Filtering: Filtering in digital audio conversion usually refers to the use of anti-aliasing filters. These are essential low-pass filters applied before the analog-to-digital conversion to remove frequencies that are higher than half the sample rate (the Nyquist frequency). Without this filtering, these high frequencies would cause an unwanted digital distortion called aliasing.
Clock: The clock is the master timing reference that dictates when the samples are taken and when they are played back. It ensures a constant, precise, and even spacing between each sample. A stable clock is crucial ; any instability, known as jitter, can cause timing errors and negatively affect the audio quality.
The USB-C adapter Clock
USB-C adapters can utilize two clock modes:
- Synchronous: Uses the host clock.
- Asynchronous (most common): Uses its own internal clock.
The test setup
MultiDSLA: Is the controller that drives tests and metric calculation.
VPP: Is the Opale Systems reference softphone, which includes a large codec library for VoIP, OTT, and mobile networks. VPP does not require any input/output audio system to play and record audio; everything is pure digital. VPP is therefore ideal for qualifying devices or softphones and for avoiding end-to-end testing, which can create blind spots in quality or latency scoring.
VoIP Softphone: A market softphone was selected with the following characteristics : handling high sampling rate and high bit rate codecs , visibility into the jitter buffer to identify latency costs , auto answer for automation , and no Session Border Controller (SBC) required for a direct SIP call between VPP and the softphone.
DSLA: Digital Speech Audio Analyzer is a high-precision signal generator/recorder used for speech quality assessment and signal integrity analysis.
Time sync: The MultiDSLA controller has a GPS connection for accurate clock synchronization. It serves the clock to the VPP/Softphone platform and DSLA through NTP over a dedicated 1 Gb/s network. The NTP clients running on the VPP/softphone platform and DSLA are custom-developed by Opale Systems to ensure high-frequency time updates and accuracy.
Performing Speech Quality test
For Speech Quality, we use the OPTICOM POLQA 3 library. This is the only method defined by ITU and adopted by all vendors or local standardization organizations to assess Speech Quality. The advantage of POLQA is its capability to handle high sample rates and to provide a Mean Opinion Score (MOS) for both speech quality and audio latency.
Calculating clock drifting
A clock is the fundamental timing reference that synchronizes all operations in digital audio transmission. It generates regular electrical pulses at a precise frequency, determining the sample rate, bit timing, and synchronization. Each component in the audio chain has its own clock, and no clock is perfect.
Clock drift refers to the difference between the actual frequency of a clock and its nominal (specified) frequency.
To characterize a system's clock drift, a 10,000 Hertz (10 kHz) signal is generated to the device under test (DUT), and the resulting frequency is captured at the output of the DUT. The peak frequency on the captured signal, which will inevitably deviate from 10 kHz, is then identified.
Expected results (Clock Drift in ppm):
- Consumer headsets: 50–200 ppm typical
- Professional audio gear: 10–50 ppm
Studio equipment:
Since a complete transmission chain incorporates the drift from every element, it is crucial that the recorder itself exhibits a very low clock drift. The DSLA is designed with a high-precision clock system to manage a very low drift. For the VPP/Softphone host, a hardware platform with a high-performance CPU is utilized.
Conclusion
The necessary transition of test equipment to the USB-C connector has introduced a critical variable into audio quality and latency measurements: clock drift. Our study confirms that standard USB-C to Jack adapters, particularly those operating in an asynchronous clock mode, represent a significant and variable source of latency drift. The drift levels observed in consumer-grade equipment (typically 50 to 200 ppm) are unacceptable for the speech quality and end-to-end latency assessment required by critical communications.
To ensure the reliability and precision of measurements, it is imperative to isolate the evaluation system from the synchronization instability inherent in these consumer components. The methodology implemented by Opale Systems, relying on the ultra-precise GPS-synchronized clock of the MultiDSLA and the purely digital VPP softphone, allows for the isolation and characterization of these drifts with unparalleled rigor.
Consequently, the use of unqualified USB-C adapters creates "blind spots" in latency scoring. The alternative solution proposed by Opale Systems is therefore not merely an option, but a technical necessity to eliminate clock drift in the measurement chain and deliver consistent, precise Quality of Service data that meets professional requirements.
Want to know more?
Contact our team today and benefit from our interactive interactive course about clock drifting
Clement Lauriat
Monday, 23rd March 2026 
