Takashi Iwai | c9746ee | 2016-11-10 22:23:58 +0100 | [diff] [blame] | 1 | ===================== |
Liam Girdwood | eb1a6af | 2006-10-06 18:34:51 +0200 | [diff] [blame] | 2 | Audio Pops and Clicks |
| 3 | ===================== |
| 4 | |
| 5 | Pops and clicks are unwanted audio artifacts caused by the powering up and down |
Matt LaPlante | 01dd2fb | 2007-10-20 01:34:40 +0200 | [diff] [blame] | 6 | of components within the audio subsystem. This is noticeable on PCs when an |
Liam Girdwood | 10b9852 | 2007-02-08 17:06:09 +0100 | [diff] [blame] | 7 | audio module is either loaded or unloaded (at module load time the sound card is |
Liam Girdwood | eb1a6af | 2006-10-06 18:34:51 +0200 | [diff] [blame] | 8 | powered up and causes a popping noise on the speakers). |
| 9 | |
Liam Girdwood | 10b9852 | 2007-02-08 17:06:09 +0100 | [diff] [blame] | 10 | Pops and clicks can be more frequent on portable systems with DAPM. This is |
| 11 | because the components within the subsystem are being dynamically powered |
| 12 | depending on the audio usage and this can subsequently cause a small pop or |
| 13 | click every time a component power state is changed. |
Liam Girdwood | eb1a6af | 2006-10-06 18:34:51 +0200 | [diff] [blame] | 14 | |
| 15 | |
| 16 | Minimising Playback Pops and Clicks |
| 17 | =================================== |
| 18 | |
Mark Brown | 7c4dbbd | 2008-01-23 08:41:46 +0100 | [diff] [blame] | 19 | Playback pops in portable audio subsystems cannot be completely eliminated |
| 20 | currently, however future audio codec hardware will have better pop and click |
| 21 | suppression. Pops can be reduced within playback by powering the audio |
| 22 | components in a specific order. This order is different for startup and |
| 23 | shutdown and follows some basic rules:- |
Takashi Iwai | c9746ee | 2016-11-10 22:23:58 +0100 | [diff] [blame] | 24 | :: |
Liam Girdwood | eb1a6af | 2006-10-06 18:34:51 +0200 | [diff] [blame] | 25 | |
Takashi Iwai | c9746ee | 2016-11-10 22:23:58 +0100 | [diff] [blame] | 26 | Startup Order :- DAC --> Mixers --> Output PGA --> Digital Unmute |
| 27 | |
| 28 | Shutdown Order :- Digital Mute --> Output PGA --> Mixers --> DAC |
Liam Girdwood | eb1a6af | 2006-10-06 18:34:51 +0200 | [diff] [blame] | 29 | |
| 30 | This assumes that the codec PCM output path from the DAC is via a mixer and then |
| 31 | a PGA (programmable gain amplifier) before being output to the speakers. |
| 32 | |
| 33 | |
| 34 | Minimising Capture Pops and Clicks |
| 35 | ================================== |
| 36 | |
| 37 | Capture artifacts are somewhat easier to get rid as we can delay activating the |
Matt LaPlante | 01dd2fb | 2007-10-20 01:34:40 +0200 | [diff] [blame] | 38 | ADC until all the pops have occurred. This follows similar power rules to |
Liam Girdwood | eb1a6af | 2006-10-06 18:34:51 +0200 | [diff] [blame] | 39 | playback in that components are powered in a sequence depending upon stream |
| 40 | startup or shutdown. |
Takashi Iwai | c9746ee | 2016-11-10 22:23:58 +0100 | [diff] [blame] | 41 | :: |
Liam Girdwood | eb1a6af | 2006-10-06 18:34:51 +0200 | [diff] [blame] | 42 | |
Takashi Iwai | c9746ee | 2016-11-10 22:23:58 +0100 | [diff] [blame] | 43 | Startup Order - Input PGA --> Mixers --> ADC |
| 44 | |
| 45 | Shutdown Order - ADC --> Mixers --> Input PGA |
Liam Girdwood | eb1a6af | 2006-10-06 18:34:51 +0200 | [diff] [blame] | 46 | |
| 47 | |
| 48 | Zipper Noise |
| 49 | ============ |
| 50 | An unwanted zipper noise can occur within the audio playback or capture stream |
| 51 | when a volume control is changed near its maximum gain value. The zipper noise |
| 52 | is heard when the gain increase or decrease changes the mean audio signal |
| 53 | amplitude too quickly. It can be minimised by enabling the zero cross setting |
| 54 | for each volume control. The ZC forces the gain change to occur when the signal |
| 55 | crosses the zero amplitude line. |