[PATCH] ASoC: nau8825: erase pop noise by soft mute

[PATCH] ASoC: nau8825: erase pop noise by soft mute

[John Hsu]

The patch is a solution to fix the following issue.
Chrome OS Issue 54078: Chell_headphone pop back from S3

When the system resume from S3 and playing music, there is a pop noise 
happened. It looks like system changes the sound immediately from maximum 
volume to the volume configured by user in playback beginning. The sound 
signal changes sharply, and there is a DC offset in the signal to make 
pop noise. Therefore, the driver introduces the soft mute to avoid the 
side effect of the unstable signal.

Signed-off-by: John Hsu <KCHSU0@nuvoton.com>
---
 sound/soc/codecs/nau8825.c | 52 +++++++++++++++++++++++++++++++++++++++++++++-
 sound/soc/codecs/nau8825.h |  3 +++
 2 files changed, 54 insertions(+), 1 deletion(-)

diff --git a/sound/soc/codecs/nau8825.c b/sound/soc/codecs/nau8825.c
index e643be9..3bcba14 100644
--- a/sound/soc/codecs/nau8825.c
+++ b/sound/soc/codecs/nau8825.c
@@ -754,8 +754,9 @@ static void nau8825_xtalk_measure(struct nau8825 *nau8825)
 			nau8825->imp_rms[NAU8825_XTALK_HPR_R2L],
 			nau8825->imp_rms[NAU8825_XTALK_HPL_R2L]);
 		dev_dbg(nau8825->dev, "cross talk sidetone: %x\n", sidetone);
+		nau8825->sidetone = (sidetone << 8) | sidetone;
 		regmap_write(nau8825->regmap, NAU8825_REG_DAC_DGAIN_CTRL,
-					(sidetone << 8) | sidetone);
+			nau8825->sidetone);
 		nau8825_xtalk_clean(nau8825);
 		nau8825->xtalk_state = NAU8825_XTALK_DONE;
 		break;
@@ -1334,6 +1335,50 @@ int nau8825_enable_jack_detect(struct snd_soc_codec *codec,
 }
 EXPORT_SYMBOL_GPL(nau8825_enable_jack_detect);
 
+/**
+ * nau8825_soft_mute - provide soft mute function
+ *
+ * @component: codec component
+ * @mute: 1 for mute; 0 for unmute
+ *
+ * Enable soft mute to gradually lower DAC volume to zero;
+ * Soft unmute will gradually increase DAC volume to volume setting.
+ * We found that signal at beginning is not stable in some application.
+ * That signal changes sharply will make pop noise; but the steps of soft
+ * mute is not enough in codec to cover the change period. Thus, the driver
+ * deffers 20ms to do soft unmute for the pop noise issue.
+ */
+int nau8825_soft_mute(struct snd_soc_codec *codec, int mute)
+{
+	struct nau8825 *nau8825 = snd_soc_codec_get_drvdata(codec);
+
+	if (mute) {
+		cancel_delayed_work(&nau8825->softmute_work);
+		/* Disable sidetone to avoid it affect the mute function */
+		regmap_write(nau8825->regmap, NAU8825_REG_DAC_DGAIN_CTRL, 0);
+		regmap_update_bits(nau8825->regmap, NAU8825_REG_MUTE_CTRL,
+			NAU8825_DAC_SOFT_MUTE, NAU8825_DAC_SOFT_MUTE);
+	} else {
+		/* Defer 20ms to do soft unmute to skip the unstable signal */
+		schedule_delayed_work(&nau8825->softmute_work,
+			msecs_to_jiffies(20));
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(nau8825_soft_mute);
+
+static void nau8825_unmute_deferred(struct work_struct *work)
+{
+	struct nau8825 *nau8825 = container_of(work,
+		struct nau8825, softmute_work.work);
+
+	/* Restore sidetone */
+	regmap_write(nau8825->regmap, NAU8825_REG_DAC_DGAIN_CTRL,
+		nau8825->sidetone);
+	regmap_update_bits(nau8825->regmap, NAU8825_REG_MUTE_CTRL,
+		NAU8825_DAC_SOFT_MUTE, 0);
+}
 
 static bool nau8825_is_jack_inserted(struct regmap *regmap)
 {
@@ -2150,6 +2195,7 @@ static int nau8825_set_sysclk(struct snd_soc_codec *codec, int clk_id,
 
 static int nau8825_resume_setup(struct nau8825 *nau8825)
 {
+	struct snd_soc_codec *codec = snd_soc_dapm_to_codec(nau8825->dapm);
 	struct regmap *regmap = nau8825->regmap;
 
 	/* Close clock when jack type detection at manual mode */
@@ -2169,6 +2215,8 @@ static int nau8825_resume_setup(struct nau8825 *nau8825)
 		NAU8825_JACK_DET_DB_BYPASS, NAU8825_JACK_DET_DB_BYPASS);
 	regmap_update_bits(regmap, NAU8825_REG_INTERRUPT_DIS_CTRL,
 		NAU8825_IRQ_INSERT_DIS | NAU8825_IRQ_EJECT_DIS, 0);
+	/* Enable codec soft mute */
+	nau8825_soft_mute(codec, 1);
 
 	return 0;
 }
@@ -2393,8 +2441,10 @@ static int nau8825_i2c_probe(struct i2c_client *i2c,
 	 */
 	nau8825->xtalk_state = NAU8825_XTALK_DONE;
 	nau8825->xtalk_protect = false;
+	nau8825->sidetone = 0;
 	sema_init(&nau8825->xtalk_sem, 1);
 	INIT_WORK(&nau8825->xtalk_work, nau8825_xtalk_work);
+	INIT_DELAYED_WORK(&nau8825->softmute_work, nau8825_unmute_deferred);
 
 	nau8825_print_device_properties(nau8825);
 
diff --git a/sound/soc/codecs/nau8825.h b/sound/soc/codecs/nau8825.h
index 1c63e2a..e15ea5c 100644
--- a/sound/soc/codecs/nau8825.h
+++ b/sound/soc/codecs/nau8825.h
@@ -433,6 +433,7 @@ struct nau8825 {
 	struct snd_soc_dapm_context *dapm;
 	struct snd_soc_jack *jack;
 	struct clk *mclk;
+	struct delayed_work softmute_work;
 	struct work_struct xtalk_work;
 	struct semaphore xtalk_sem;
 	int irq;
@@ -459,10 +460,12 @@ struct nau8825 {
 	int xtalk_event_mask;
 	bool xtalk_protect;
 	int imp_rms[NAU8825_XTALK_IMM];
+	int sidetone;
 };
 
 int nau8825_enable_jack_detect(struct snd_soc_codec *codec,
 				struct snd_soc_jack *jack);
+int nau8825_soft_mute(struct snd_soc_codec *codec, int mute);
 
 
 #endif  /* __NAU8825_H__ */
-- 
2.6.4

Re: [PATCH] ASoC: nau8825: erase pop noise by soft mute

[Mark Brown]

[-- Attachment #1.1: Type: text/plain, Size: 856 bytes --]

On Wed, Sep 14, 2016 at 03:47:32PM +0800, John Hsu wrote:

> + * Enable soft mute to gradually lower DAC volume to zero;
> + * Soft unmute will gradually increase DAC volume to volume setting.

> +		regmap_write(nau8825->regmap, NAU8825_REG_DAC_DGAIN_CTRL, 0);
> +		regmap_update_bits(nau8825->regmap, NAU8825_REG_MUTE_CTRL,
> +			NAU8825_DAC_SOFT_MUTE, NAU8825_DAC_SOFT_MUTE);

Why are we not just exposing soft mute as a userspace control like other
drivers do?  It seems like there's some weird interaction between
sidetones and the soft mute which this is trying to work around but
that's not really explained, it seems to be the main point here.
Basically I can't figure out what's intended here.

> +EXPORT_SYMBOL_GPL(nau8825_soft_mute);

Why is this exported, how will it be used?  The function seems to be
called from inside this driver as well...

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 455 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: [PATCH] ASoC: nau8825: erase pop noise by soft mute

[John Hsu]

On 9/25/2016 3:11 AM, Mark Brown wrote:
> On Wed, Sep 14, 2016 at 03:47:32PM +0800, John Hsu wrote:
>
>   
>> + * Enable soft mute to gradually lower DAC volume to zero;
>> + * Soft unmute will gradually increase DAC volume to volume setting.
>>     
>
>   
>> +		regmap_write(nau8825->regmap, NAU8825_REG_DAC_DGAIN_CTRL, 0);
>> +		regmap_update_bits(nau8825->regmap, NAU8825_REG_MUTE_CTRL,
>> +			NAU8825_DAC_SOFT_MUTE, NAU8825_DAC_SOFT_MUTE);
>>     
>
> Why are we not just exposing soft mute as a userspace control like other
> drivers do?  It seems like there's some weird interaction between
> sidetones and the soft mute which this is trying to work around but
> that's not really explained, it seems to be the main point here.
> Basically I can't figure out what's intended here.
>
>   

The issue only happens in the begin of playback. The amplitude of
output signal has abnormal change in a short time and then recover
to the level by user configuration. The amplitude change makes pop
noise. For the reason, we use soft mute to fix the pop noise because
the signal will output from small to normal case gradually. But the
steps of soft mute is not enough to cover the period of amplitude
change. Thus, we need to defer the soft mute and put the triggered
point near the signal output. The userspace control can't catch the
point as we wish.

>> +EXPORT_SYMBOL_GPL(nau8825_soft_mute);
>>     
>
> Why is this exported, how will it be used?  The function seems to be
> called from inside this driver as well...
>   

The function is exported for the machine driver. When the platform
stops the playback and it makes soft mute. When the playback starts,
the playform makes soft unmute in trigger start function. The trigger
function is the closest point of output signal, and the codec driver
can make the soft unmute deferring as short as possible.

Re: [PATCH] ASoC: nau8825: erase pop noise by soft mute

[John Hsu]

On 10/3/2016 7:21 PM, John Hsu wrote:
> On 9/25/2016 3:11 AM, Mark Brown wrote:
>> On Wed, Sep 14, 2016 at 03:47:32PM +0800, John Hsu wrote:
>>
>>  
>>> + * Enable soft mute to gradually lower DAC volume to zero;
>>> + * Soft unmute will gradually increase DAC volume to volume setting.
>>>     
>>
>>  
>>> +        regmap_write(nau8825->regmap, NAU8825_REG_DAC_DGAIN_CTRL, 0);
>>> +        regmap_update_bits(nau8825->regmap, NAU8825_REG_MUTE_CTRL,
>>> +            NAU8825_DAC_SOFT_MUTE, NAU8825_DAC_SOFT_MUTE);
>>>     
>>
>> Why are we not just exposing soft mute as a userspace control like other
>> drivers do?  It seems like there's some weird interaction between
>> sidetones and the soft mute which this is trying to work around but
>> that's not really explained, it seems to be the main point here.
>> Basically I can't figure out what's intended here.
>>
>>   
>
> The issue only happens in the begin of playback. The amplitude of
> output signal has abnormal change in a short time and then recover
> to the level by user configuration. The amplitude change makes pop
> noise. For the reason, we use soft mute to fix the pop noise because
> the signal will output from small to normal case gradually. But the
> steps of soft mute is not enough to cover the period of amplitude
> change. Thus, we need to defer the soft mute and put the triggered
> point near the signal output. The userspace control can't catch the
> point as we wish.
>
>>> +EXPORT_SYMBOL_GPL(nau8825_soft_mute);
>>>     
>>
>> Why is this exported, how will it be used?  The function seems to be
>> called from inside this driver as well...
>>   
>
> The function is exported for the machine driver. When the platform
> stops the playback and it makes soft mute. When the playback starts,
> the playform makes soft unmute in trigger start function. The trigger
> function is the closest point of output signal, and the codec driver
> can make the soft unmute deferring as short as possible.
>
>

We want to abandon the patch under reviewing because there is another
study and solution for the pop noise. What should I do to abandon the
patch? Thank you for the advice.