Minimum Performance Requirements
Requirement
|
Value
|
Full-scale input voltage:
|
|
FSIV (A-D-PC) line input
|
1.0 Vrms
|
FSIV (A-D-PC) microphone input
|
100 mVrms
|
Full-scale output voltage:
|
|
FSOV (PC-D-A) line output
|
1.0 Vrms [note 1]
|
Digital playback (PC-D-A) for line output:
|
Frequency response (-3 dB)
48.0 kHz source material
|
20 Hz to 19.2 kHz [note 4]
|
Dynamic range (DR)
|
80 dB FS A [note 3, 4]
|
THD+N (-3 dB FS)
|
-65 dB FS [note 4]
|
Digital playback (PC-D-A) for speaker output with 8-ohm load:
|
Frequency response (-3 dB)
48.0 kHz source material
|
100 Hz to 19.2 kHz [note 4]
|
Dynamic range (DR)
|
80 dB FS A [note 3, 4]
|
THD+N (-3 dB FS)
|
-65 dB FS [note 4]
|
Digital recording (A-D-PC) for line input: [note 2]
|
Frequency response
48.0 kHz destination
|
20 Hz to 19.2 kHz [note 4]
|
Passband ripple
|
<+/-0.5 dB
|
Dynamic range (DR)
|
70 dB FS A [note 4]
|
THD+N (-3 dB FS)
|
-60 dB FS (input referenced) [note 4]
|
Digital recording (A-D-PC) for microphone input:
|
Frequency response (-3 dB)
22.05 kHz destination
|
100 Hz to 8.8 kHz
|
Passband ripple
|
<+/-0.5 dB
|
Dynamic range (DR)
|
70 dB FS A [note 4]
|
THD+N (-3 dB FS)
|
-60 dB FS (input referenced) [note 4]
|
Line output cross-talk:
|
Channel separation
between left and right line-out channels
(measured at 10 kHz)
|
60 dB [note 4]
|
Sampling frequency accuracy:
|
Playback
|
0.1%
|
Record
|
0.1%
|
Notes:
|
Note 1: For 3.3 V audio codec, the required Full Scale Output Voltage for line output is 0.7 Vrms.
Note 2: Line input to speaker output is a requirement only if a line output is not supported.
Note 3: Decibels relative to full scale (FS), measured using "A weighting" filters.
Note 4 - Mobile PC Notes:
The dynamic range requirements are relaxed from previous requirements by 10 dB FS.
The THD+N requirements are relaxed by 10 dB FS.
The required frequency response is 30 Hz to 15 kHz, measured using 3 dB corners.
The cross-talk requirements are relaxed by 10 dB FS.
|
Current "Designed for Windows XP" Requirements
B3.1.4.5 See B3.1.4.2 B3.1.4.6 Analog microphone input meets jack and circuit specifications
This requirement enables users with electret or dynamic microphones to connect the device to their PC and achieve consistent results. This requirement also maintains compatibility with the installed base of microphones.
If the PC has an analog microphone input, it must meet the following specifications:
-
Three-conductor 1/8-inch (3.5 millimeters) tip/ring/sleeve microphone jack where the microphone signal is on the tip, bias is on the ring, and the sleeve is grounded.
This design is optimized for electret microphones with three-conductor plugs, but will also support dynamic microphones with two-conductor plugs, where the ring and sleeve are shorted together.
-
Minimum AC input impedance between tip and ground: minimum, 4 kilohms.
-
Input voltages of 10–100 millivolts (mV) must deliver full-scale digital input, using software-programmable gain.
-
Maximum 5.5 V with no load, minimum 2.0 V with 0.8 milliampere (mA) load, direct current bias for electret microphones.
-
Minimum bias impedance between bias voltage source and ring: 2 kilohms.
-
AC coupled tip.
B3.1.4.7 Audio driver reports sample position for stream synchronization
The driver must be capable of reporting within 1 ms the current position of the buffer being rendered, in relation to the samples given to the codec. This requirement applies for both compressed and uncompressed data.
Maximum allowable latency is 1 ms.
B3.1.4.8 PCI audio device meets minimum requirements B3.1.4.8.1 Supports initiator, target, and block transfer.
For complete implementation details, see PCI 2.2. Full-duplex audio sample transport must be supported using separate PCI bus mastering hardware for playback and capture sample streams.
B3.1.4.8.2 PCI device supports efficient audio buffer management.
The audio device must be able to fully function when the system can only provide single pages of contiguous memory. In other words, the audio device can require many pages of memory, but must not require the largest block of contiguous memory to exceed one page.
This requirement ensures audio support in docking and dynamic loading scenarios where memory may be fragmented with respect to pages.
The audio device and associated device-specific driver must not introduce unnecessary latency. If the audio driver adds more than 2 milliseconds of computational latency between buffer transfer and queuing for rendering, the driver must provide a programmatic method for a latency sensitive application to temporarily disable the computation.
This requirement helps to ensure that telephony applications closely resemble the performance of a conventional phone and minimizes the possibility that audio and video streams will appear out of synch.
B3.1.4.8.3 Device does not use ISA-based resources. B3.1.4.8.4 Audio subsystem is digital ready.
Audio sources must be available as digital audio streams accessible to the system-wide kernel, that is, they must not rely exclusively on any analog mixing stage between the digital-to-analog converter and the speaker jack as the only means for output. Sources that continue to offer an analog mixing output configuration must also provide the user a configurable digital option.
One model for providing the user such an option is Windows support for CD Music.
The audio sources covered by this requirement, which must be available digitally to USB speakers if attached, are:
-
CD-ROM or DVD
-
TV tuner
-
FM radio
-
Voice modem
Analog Mic and Line In with available analog-to-digital converters (ADCs) are digital ready by definition. Devices for which the operating system supports emulation equivalents, such as hardware accelerated 3-D and MIDI synthesis (Microsoft DirectSound® 3D emulation and the Windows GS Wavetable SW Synth) are excepted.
This requirement assures that all audio content can be made available at both the analog jack and USB port. Elimination of the dependency on analog mixing for output is key to making PC audio easier to configure and use, and it removes a major obstacle for USB audio rendering devices.
USB audio devices must use an HID to control basic functions. If a USB audio device plans to implement volume adjustment controls, it must employ the HID usages provided on the consumer controls page in the USB HID Usage Tables, Version 1.1 specification and in “Windows Support for HID-based Audio Controls.”
B3.1.4.10 Support required for DirectX functions as specified in the DirectX SDK “DirectX Audio” and the Windows DDK “Audio Drivers” section B3.1.4.11 Device implements DRM support as defined in Windows DDK
To pass Windows Logo testing for Secure Audio Path for Digital Rights Management as DRM compliant with DRM Level = 1200 (which is the security level for Windows XP), audio drivers must not call DrmForwardContentToFileObject.
"Digital Rights Management Reference” in the Windows DDK.
B3.1.4.12 Audio subsystem requirements
Audio subsystem supports AEC reference inputs. Built-in or external audio devices that support full-duplex, stereo playback and record for speakers and microphone must support simultaneous capture of microphone and one or more AEC reference inputs.
At minimum, the audio device must support capture of microphone input in the left channel and a monaural output mix in the right channel, where the monaural output mix is the left and right channels of the main output merged into a single channel. This 1+1 channel interleaved format (similar to stereo) will be referred to as “mic+ref”, and can be easily achieved using existing stereo ADCs.
B3.1.4.12.1 Supports acoustic echo cancellation reference inputs.
See B3.1.4.12
B3.1.4.12.2 Does not rely on analog mixing.
See B3.1.4.8.4
B3.1.4.13 Headset microphone used for speech recognition meets performance requirements
The following requirements are for close-speaking headset microphones intended for use in speech-recognition applications. These requirements are compatible with most of the installed base of sound cards and audio-enabled system boards.
The requirements for a speech-recognition microphone are:
-
Close-speaking headset design positions microphone within 1.5 inches of the corner of the speaker’s mouth.
-
FSOV: 100 mV (0 dB FS).
-
Microphone connector meets requirements.
-
Operating bias voltage from 2.0–5.0 volts direct current (VDC) with a maximum current drain of 0.8 mA.
-
Capable of sustaining a maximum voltage of 10 VDC on tip or ring without damage.
-
Frequency response:
±3 dB from 100 Hz to 10kHz
0 dB at 1 kHz
-
Minimum sensitivity of –44 dB relative to 1 volt per pascal.
-
Maximum 2 percent THD+N 100 Hz to 10 kHz at 94 decibel sound pressure level (dBSPL).
-
Noise cancellation null sensitivity at 90 degrees and 270 degrees, ±10 degrees, with the following minimums:
20 dB at 100 Hz
20 dB at 400 Hz
20 dB at 1000 Hz
20 dB at 4000 Hz
10 dB at 10 kHz
-
Maximum wind noise sensitivity of –65 dB with 0 dB = 1 V (measured with wind speed of 1 meter per second at the 0 degree axis of microphone).
Maximum output impedance of 1 kilohm (using a 1-kHz full-scale test tone with 2.0 VDC bias).
|
