Robert Ly | 35f2fda | 2013-01-29 16:27:05 -0800 | [diff] [blame^] | 1 | page.title=Audio Warmup |
| 2 | @jd:body |
| 3 | |
| 4 | <!-- |
| 5 | Copyright 2010 The Android Open Source Project |
| 6 | |
| 7 | Licensed under the Apache License, Version 2.0 (the "License"); |
| 8 | you may not use this file except in compliance with the License. |
| 9 | You may obtain a copy of the License at |
| 10 | |
| 11 | http://www.apache.org/licenses/LICENSE-2.0 |
| 12 | |
| 13 | Unless required by applicable law or agreed to in writing, software |
| 14 | distributed under the License is distributed on an "AS IS" BASIS, |
| 15 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 16 | See the License for the specific language governing permissions and |
| 17 | limitations under the License. |
| 18 | --> |
| 19 | <div id="qv-wrapper"> |
| 20 | <div id="qv"> |
| 21 | <h2>In this document</h2> |
| 22 | <ol id="auto-toc"> |
| 23 | </ol> |
| 24 | </div> |
| 25 | </div> |
| 26 | |
| 27 | <p>Audio warmup is the time for the audio amplifier circuit in your device to |
| 28 | be fully powered and reach its normal operation state. The major contributors |
| 29 | to audio warmup time are power management and any "de-pop" logic to stabilize |
| 30 | the circuit. |
| 31 | </p> |
| 32 | |
| 33 | <p>This document describes how to measure audio warmup time and possible ways to decrease |
| 34 | warmup time.</p> |
| 35 | |
| 36 | <h2 id="measuringOutput">Measuring Output Warmup</h2> |
| 37 | |
| 38 | <p> |
| 39 | AudioFlinger's FastMixer thread automatically measures output warmup |
| 40 | and reports it as part of the output of the <code>dumpsys media.audio_flinger</code> command. |
| 41 | At warmup, FastMixer calls <code>write()</code> |
| 42 | repeatedly until the time between two <code>write()</code>s is the amount expected. |
| 43 | FastMixer determines audio warmup by seeing how long a HAL <code>write()</code> takes to stabilize. |
| 44 | </p> |
| 45 | |
| 46 | <p>To measure audio warmup, do the following |
| 47 | steps for the built-in speaker and wired headphones and at different times after booting. |
| 48 | Warmup times are usually different for each output device and right after booting the device:</p> |
| 49 | |
| 50 | <ol> |
| 51 | <li>Ensure that FastMixer is enabled.</li> |
| 52 | <li>Enable touch sounds by selecting <b>Settings > Sound > Touch sounds</b> on the device.</li> |
| 53 | <li>Ensure that audio has been off for at least three seconds. Five seconds or more is better, because |
| 54 | the hardware itself might have its own power logic beyond the three seconds that AudioFlinger has.</li> |
| 55 | <li>Press Home, and you should hear a click sound.</li> |
| 56 | <li>Run the following command to receive the measured warmup: |
| 57 | <pre>adb shell dumpsys media.audio_flinger | grep measuredWarmup</code></pre> |
| 58 | |
| 59 | <p> |
| 60 | You should see output like this: |
| 61 | </p> |
| 62 | |
| 63 | <pre> |
| 64 | sampleRate=44100 frameCount=256 measuredWarmup=X ms, warmupCycles=X |
| 65 | </pre> |
| 66 | |
| 67 | <p> |
| 68 | The <code>measuredWarmup=X</code> is X number of milliseconds |
| 69 | it took for the first set of HAL <code>write()</code>s to complete. |
| 70 | </p> |
| 71 | |
| 72 | <p> |
| 73 | The <code>warmupCycles=X</code> is how many HAL write requests it took |
| 74 | until the execution time of <code>write()</code> matches what is expected. |
| 75 | </p> |
| 76 | </li> |
| 77 | <li> |
| 78 | Take five measurements and report them all, as well as the mean. |
| 79 | If they are not all approximately the same, |
| 80 | then it’s likely that a measurement is incorrect. |
| 81 | For example, if you don’t wait long enough after the audio has been off, |
| 82 | you will see a lower warmup time than the mean value. |
| 83 | </li> |
| 84 | </ol> |
| 85 | |
| 86 | |
| 87 | <h2 id="measuringInput">Measuring Input Warmup</h2> |
| 88 | |
| 89 | <p> |
| 90 | There are currently no tools provided for measuring audio input warmup. |
| 91 | However, input warmup time can be estimated by observing |
| 92 | the time required for <a href="http://developer.android.com/reference/android/media/AudioRecord.html#startRecording()">startRecording()</a> |
| 93 | to return. |
| 94 | </p> |
| 95 | |
| 96 | |
| 97 | <h2 id="reducing">Reducing Warmup Time</h2> |
| 98 | |
| 99 | <p> |
| 100 | Warmup time can usually be reduced by a combination of: |
| 101 | <ul> |
| 102 | <li>Good circuit design</li> |
| 103 | <li>Accurate time delays in kernel device driver</li> |
| 104 | <li>Performing independent warmup operations concurrently rather than sequentially</li> |
| 105 | <li>Leaving circuits powered on or not reconfiguring clocks (increases idle power consumption). |
| 106 | <li>Caching computed parameters</li> |
| 107 | </ul> |
| 108 | However, beware of excessive optimization. You may find that you |
| 109 | need to tradeoff between low warmup time versus |
| 110 | lack of popping at power transitions. |
| 111 | </p> |