First of all i'd like to thank Meltus and jesus_c and gadgetcheck for their work.
http://forum.xda-developers.com/showthread.php?t=1722136
http://forum.xda-developers.com/showpost.php?p=27741216&postcount=54
i merely studied the file, i'm no expert.
i just know intermediate stuff.
i didn't know what hpf was, so i googled it and found it is high pass filter so i disabled it.
that alone didn't do the job well, the rest was educated guess and i'm glad it was right.
difference may not be that noticeable on phone speaker but can be clearly noticed when using headset.
the experts can now tweak and sort out more stuff after.
here's a comparison of vids:
old is from default
http://www.mediafire.com/?1a0wcd0rdodadd3
new is from tweaked
http://www.mediafire.com/?aob456t4qoyz25c
attached zip is cwm flashable.
UPDATE:
vid recording of sgs3 placed on laptop playing music.
default BFLB:
http://www.mediafire.com/?ke87s5nj2tpfhfb
tweaked BFLB:
http://www.mediafire.com/?63wzv2jd51bzy6o
NEW!!! Default BLG6/8:
http://www.mediafire.com/?se4ioy7x4206n1l
as you can see, tweaked BLFB and default BLG6/8 are pretty close.
the default BFLB's audio in vid was very bad, comparing it to default BLG6/8 there is a marked improvement.
samsung has indeed improved audio recording in vids now.
default vid starts around 2seconds, the tweaked at 6th second.
as can be seen on default, this phone is weak at picking up those between 32hz and 64hz.
if you can play the two vids and sync them, you'll see the tweaked picking up more bass than default.
app used in sgs3 is rta audio analyzer found in google play.
There is a notable difference between both videos, but as the sound source is just noise it's hard to say if this is a good or bad improvement with naked ear.
I will be making some test later with your mod, I hope for the best!
Cheers!
i have before and after vids of cars passing by just like gsmarena's. as this was where we all first noticed sgs3's bad audio in vid recording.
but sad to say my upload speed here is very slow and the files are over 100mb cos i recorded each over a minute to be sure.
all i can say is the vid with tweaked is better.
i've listened to both using headset to differentiate well, the tweaked one has more depth which many has complained the audio in sgs3 was lacking.
and there is less walkie-talkie effect scratchy noise than before.
You can use VLC to extract the audio channel from the vid files and upload only the audio files.
Then you can record 5 minutes on each, and be left with 5mb for each
i used mp4muxer, i extracted the aac audio in raw.
default
http://www.mediafire.com/?bbcjfj8j6zkmynr
tweaked
http://www.mediafire.com/?cns7j748i3y2uhj
the best way i can describe it is that the mic captures better now. it was windy when i recorded both a while ago, and only the tweaked version as you can hear captured the wind.
radeonxt said:
First of all i'd like to thank Meltus and jesus_c and gadgetcheck for their work.
http://forum.xda-developers.com/showthread.php?t=1722136
http://forum.xda-developers.com/showpost.php?p=27741216&postcount=54
i merely studied the file, i'm no expert.
i just know intermediate stuff.
i didn't know what hpf was, so i googled it and found it is high pass filter so i disabled it.
that alone didn't do the job well, the rest was educated guess and i'm glad it was right.
difference may not be that noticeable on phone speaker but can be clearly noticed when using headset.
the experts can now tweak and sort out more stuff after.
here's a comparison of vids:
old is default
http://www.mediafire.com/?1a0wcd0rdodadd3
new is tweaked
http://www.mediafire.com/?aob456t4qoyz25c
attached zip is cwm flashable.
Click to expand...
Click to collapse
Amazing! This was one of my biggest complaints about video recording with this phone. The dynamic range was cut with a high pass filter and you solved it. Good job and thanks for this!
there is definitely an improvement , but still cant record any bass
Haven't had the chance to experiment but I understand you completely disabled the High Pass Filter by using HPF Switch and set it to 0?
{ "AIF1ADC1 HPF Switch", 0 },
But another interesting line is the one above it:
{ "AIF1ADC1 HPF Mode", 2 },
Appearantly there are 3 or 4 modes for the HPF:
0 = Hi-Fi
1 = Voice 1
2 = Voice 2
(3 = Voice 3)?
Perhaps setting it to Hi-Fi (value of 0) instead of 1 or 2 (voice) is worth a try as well. I can imagine when the focus is set to voice a lot of surrounding sound is filtered out and Hi-Fi might be more 'general' or 'pure' but without issues of too much noise for the mic?
Will try it out some later myself btw
dagrim1 said:
Haven't had the chance to experiment but I understand you completely disabled the High Pass Filter by using HPF Switch and set it to 0?
{ "AIF1ADC1 HPF Switch", 0 },
But another interesting line is the one above it:
{ "AIF1ADC1 HPF Mode", 2 },
Appearantly there are 3 or 4 modes for the HPF:
0 = Hi-Fi
1 = Voice 1
2 = Voice 2
(3 = Voice 3)?
Perhaps setting it to Hi-Fi (value of 0) instead of 1 or 2 (voice) is worth a try as well. I can imagine when the focus is set to voice a lot of surrounding sound is filtered out and Hi-Fi might be more 'general' or 'pure' but without issues of too much noise for the mic?
Will try it out some later myself btw
Click to expand...
Click to collapse
I tried setting it to 0 last night. Recording quality was about the same. Something I never noticed before. .. even in the audio recorder, or using an app life tape machine records with no low end bass at all.. .. hell even when recording the bloody radio. ... its a global problem with this phone on all recording levels. .. my sgs 2 does not suffer from this.. . Might be kernel related as all the options I try in this file still do not give me any thing lower then
0.5khz. .. pretty stupid limitation.. ..
it wasn't just hpf that i tried to tweak guys.
i did say in first post
i didn't know what hpf was, so i googled it and found it is high pass filter so i disabled it.
that alone didn't do the job well, the rest was educated guess and i'm glad it was right.
difference may not be that noticeable on phone speaker but can be clearly noticed when using headset.
just glad that many have confirmed the clarity,volume and such is better.
i'm hoping the experts will include tweaking this more along with their other audio recording fixes, since i can do no more further as i've mentioned i'm no expert and also too busy with business these days.
Ugh, yeah... thought I was able to read better
Do you know if you have to reboot when changing this conf file? Or is it done on the fly. Might want to play around with it some more.
I've been looking into this recently, you beat me to it
This is what I've found (I'm not saying I'm right or anything, these are just my conclusions from research).
"AIF1ADC1 HPF" is indeed a High-Pass filter but would disabling the High-Pass filter be a good idea for audio recording?
Well, we have no idea what the current cut-off frequency is for the HPF (but we can take an educated guess, see below) and indeed, as 'dagrim1' pointed out, whether the cut-off frequency changes depending on the mode set. I've been unable to find a technical specification of the Galaxy S III's microphone but I'm guessing it wont be very good at handling the lower frequencies. Another reason is the majority of time we want the audio from a video recording it's when we are recording people. The recording of voices certainly wouldn't benefit from the extra bass frequencies that are being recorded from disabling the high-pass filter as the human voice is around 85Hz to 255Hz (including both male and female range - I've chosen this range because it is, roughly, what I was taught. There are people who claim different but it doesn't really matter that much).
If we look at the HPF modes (that dagrim1 pointed out), we can see there are 3 modes in use; 0, 1 and 2. Now 1 seems only to be used in the telephony sections ("Communication" modifiers in the file) whilst 2 is used for pretty much everything else (apart from where it isn't used at all, '0'. Although, '0' could itself be another HPF mode and whether it's on or off could be controlled entirely by the switch). The standard Frequency Band used for telephones is, approximately, 300hz to 3400hz (NOTE: this could just be for landline telephones and not mobiles, I'm not entirely sure), so we can logically assume the cut-off frequency of mode 1 is around the 300hz mark. Judging by the fact that telephone frequencies need to be quite heavily limited, I'm guessing that Band 1 will have a lower cut-off point (maybe around 50-100hz or so?), allowing more bass frequencies through.
So, you could remove the low-pass filter entirely and capture the whole range of sounds (no idea how the microphone will cope with this as I can't find any technical information about it), you could set it to the standard '2' or you can set the HPF to mode '1', limiting the available frequency range even further but potentially improving the clarity of the recording (it will probably sound tinny as hell too!).
That being said, with the 2 recordings 'radeonxt' made, the one with the disabled HPF did sound much better. It could sound worse in certain situations but it could also be an improvement across the board.
I see you also added in "AIF1ADC1L Mixer ADC/DMIC Switch". This seems to be a special modifier reserved for In-Call recording. I haven't had much luck adding extra parameters into this file (it seems to be quite strict) but I have no idea whether this will affect the sound or not. In theory, as the microphone on the SGSIII is indeed stereo (I think it is anyway, or am I just making that up?) it could enable both channels if they are not enabled by default. But, if both channels are already used and are simply controlled by 1 HPF and not 2, then this line could just do the same as "AIF1ADC1 HPF Switch".
There's also 2 separate sections for the camcorder microphone recording; "built-in mic", the normal one, I'm guessing, and the "back mic". Now I have absolutely no idea what the "back mic" is referring to, unless it means the secondary "noise cancelling mic" but it's not really on the "back" of the device.
EDIT: It seems the "back mic" records the right signal and the "built-in mic" records the left.
Of course, for any of this to actually be relevant completely depends on whether any of it is actually correct. I guess we'll just have to do some more testing
TL;DR version: Does disabling the HPF improve video recording audio? Dunno, lol. Depending upon the situation, perhaps.
dagrim1 said:
Ugh, yeah... thought I was able to read better
Do you know if you have to reboot when changing this conf file? Or is it done on the fly. Might want to play around with it some more.
Click to expand...
Click to collapse
You can either reboot or do "adb killall mediaserver" (or just "killall mediaserver" from a terminal) to re-load all the audio files
Could someone possibly do a recording of pink noise out of studio monitor speakers, so he could then do an analysis of the recorded frequency range?
–—–—–—–—–—–—–—–—
tapatalked from Galaxy SIII
welcome meltus.
yes i didn't even know what the hpf acronym was.
i googled just hpf and got many results, so i added what is hpf in audio and the search result yielded this wiki:
A high-pass filter (HPF) is an electronic filter that passes high-frequency signals but attenuates (reduces the amplitude of) signals with frequencies lower than the cutoff frequency. The actual amount of attenuation for each frequency varies from filter to filter. A high-pass filter is usually modeled as a linear time-invariant system. It is sometimes called a low-cut filter or bass-cut filter.[1] High-pass filters have many uses, such as blocking DC from circuitry sensitive to non-zero average voltages or RF devices. They can also be used in conjunction with a low-pass filter to make a bandpass filter.
was about to search what the hell attenuate means but luckily inside the parenthesis it showed it means reduces.
so knowing the basic stuff that 0 would mean disable and 1 would mean enable, i changed it to 0.
then after that i tried recording with hpf value changed to 0, i didn't notice much difference.
the only time it made a difference like shown in those before,after vids was when i added AIF1ADC1L and AIF1ADC1R on camcorder's built in mic and back mic. i decided to retain the change i made to hpf to 0 since getting a few bass perhaps is better than nothing.
since i was just interested with camcorder, i just looked at what was in there. then i saw settings for built in mic and back mic, one was just L the other was R. so i figured one was used for Left and other was used for Right. the aif1adc was simply just aif1adc, i found an aif1adcL and aif1adcR so i tried putting both on built in and back mic because i was thinking builtin mic is the one on the bottom and back mic is the one on top.
thank you meltus, i noticed another thing maybe we can try to tweak in those mic settings and it's
{ "IN1L Volume", 30 },
{ "MIXINL IN1L Volume", 0 },
and
{ "IN1R Volume", 30 },
{ "MIXINR IN1R Volume", 0 },
how about we try playing around those and upping those volumes?
dagrim1 said:
Ugh, yeah... thought I was able to read better
Do you know if you have to reboot when changing this conf file? Or is it done on the fly. Might want to play around with it some more.
Click to expand...
Click to collapse
i use cwm manager and select flash update then select that zip file.
so it loads since it restarts after it finishes flashing.
its fast since it's only a 197kb zip file.
FadeFx said:
Could someone possibly do a recording of pink noise out of studio monitor speakers, so he could then do an analysis of the recorded frequency range?
–—–—–—–—–—–—–—–—
tapatalked from Galaxy SIII
Click to expand...
Click to collapse
yes, there are many experts and audiophiles here in xda that have the knowledge and the gadgets to do this stuff.
i for myself can only use rta audio analyzer which can be found in google play.
the number of bands i chose in options is 20 - 1/2 octave.
then i played on my laptop black eyed peas let's get it started and place the sgs3 on the laptop's keyboard.
the thing i noticed is that 250hz to 16k range all get picked up and can reach about -40db.
the weakest pick up is between 32hz and 64hz.
my guesswork says the ones on the left side are for bass right?
center to right are mid range to high pitches?
radeonxt said:
welcome meltus.
yes i didn't even know what the hpf acronym was.
i googled just hpf and got many results, so i added what is hpf in audio and the search result yielded this wiki:
A high-pass filter (HPF) is an electronic filter that passes high-frequency signals but attenuates (reduces the amplitude of) signals with frequencies lower than the cutoff frequency. The actual amount of attenuation for each frequency varies from filter to filter. A high-pass filter is usually modeled as a linear time-invariant system. It is sometimes called a low-cut filter or bass-cut filter.[1] High-pass filters have many uses, such as blocking DC from circuitry sensitive to non-zero average voltages or RF devices. They can also be used in conjunction with a low-pass filter to make a bandpass filter.
was about to search what the hell attenuate means but luckily inside the parenthesis it showed it means reduces.
so knowing the basic stuff that 0 would mean disable and 1 would mean enable, i changed it to 0.
then after that i tried recording with hpf value changed to 0, i didn't notice much difference.
the only time it made a difference like shown in those before,after vids was when i added AIF1ADC1L and AIF1ADC1R on camcorder's built in mic and back mic. i decided to retain the change i made to hpf to 0 since getting a few bass perhaps is better than nothing.
since i was just interested with camcorder, i just looked at what was in there. then i saw settings for built in mic and back mic, one was just L the other was R. so i figured one was used for Left and other was used for Right. the aif1adc was simply just aif1adc, i found an aif1adcL and aif1adcR so i tried putting both on built in and back mic because i was thinking builtin mic is the one on the bottom and back mic is the one on top.
thank you meltus, i noticed another thing maybe we can try to tweak in those mic settings and it's
{ "IN1L Volume", 30 },
{ "MIXINL IN1L Volume", 0 },
and
{ "IN1R Volume", 30 },
{ "MIXINR IN1R Volume", 0 },
how about we try playing around those and upping those volumes?
Click to expand...
Click to collapse
Ah, so it was the "AIF1ADC1L/R" lines that actually made the difference? That's interesting, I wonder what other lines we could add to gain more control?
I think that the Volume parameters will simply be microphone gain, so to achieve clearer results we could try lowering that value so the microphone distorts less. Or, the "AIF1ADC1" value could be something to do with the HPF and could perhaps control the cut-off value?
There's a lot we don't know, it's a shame Samsung didn't provide us with any documentation about it!
I also have a pair of ADAM A7X studio monitors so I'll do the "Pink Noise" testing, as suggested by FadeFx, with those later and see if changing the HPF mode value actually does make any difference.
yes AIF1ADC1L/R that made the difference.
can be clearly heard when we use headset.
Meltus said:
Ah, so it was the "AIF1ADC1L/R" lines that actually made the difference? That's interesting, I wonder what other lines we could add to gain more control?
I think that the Volume parameters will simply be microphone gain, so to achieve clearer results we could try lowering that value so the microphone distorts less. Or, the "AIF1ADC1" value could be something to do with the HPF and could perhaps control the cut-off value?
There's a lot we don't know, it's a shame Samsung didn't provide us with any documentation about it!
I also have a pair of ADAM A7X studio monitors so I'll do the "Pink Noise" testing, as suggested by FadeFx, with those later and see if changing the HPF mode value actually does make any difference.
Click to expand...
Click to collapse
As for low end recording,
Changing AIF1ADC1 to AIF2ADC2 gave me low end roll-off around 200Hz lowest point at -30db was 163Hz.
What I did:
I recorded a few bass kicks, drops and rumbles from a few VST plugins, put the phone next to my Monitors (and the Sub) - recorded audio/video and then I used the linear frequency graph in Cubase to see what my recorded waveform was doing.
This is the Mod I tested :
Modifier "Camcorder" {
SupportedDevice {
"Builtin Mic"
}
Enable {
{ "AIF2ADC2 HPF Mode", 0 },
{ "AIF2ADC2 HPF Switch", 0 },
{ "IN1L Volume", 30 },
{ "MIXINL IN1L Volume", 0 },
{ "AIF2ADC2 Volume", 96 },
{ "AIF2ADC2L Mixer ADC/DMIC Switch", 1 },
{ "AIF2ADC2R Mixer ADC/DMIC Switch", 1 },
AND REPEAT FOR THE "BACK MIC"
I modded the values in Voice for recorder to match what I had done in Camcorder Modifier.
Try it, record bass from a sub and see - I got results (High-end was untouched)
EDIT::::::.........
This did not do what I thought it was doing. Turns out thaqt Sub frequencys where recorded below 300Hz, but up from 90 to 30Hz - there is a 'Scoop-out' so this didn't really do anything significant.
BTW - in AIF2ADC2 the ADC stands for 'A'nalogue to 'D'igital 'C'onverter ... no filter preferences there - but maybe switching it to a different ADC will give 'Full Range' processing???
yes i did try 0'ing or disabling those hpf in the enable parts last night and compared results.
wasn't sure since there is no noticeable difference and knowing laptops don't give out good bass as compared to speakers with subs.
and yes that's another one who has noticed that, from 32hz it gets better than nothing bass then onwards to 64hz the eq bar of rta audio analyzer for android shows it mostly stays at the bottom most of the time.
Related
Hi there
last week I got a new Samsung Galaxy S GT I-9000, and although I am generally happy with it, I am very disappointed with the ADC recording quality.
As others, I noticed it after asking a friend to film me playing at a gig, and the audio was completely unuseable.
After digging around and some experiments I found so far 2 serious issues:
1 - As others pointed out, the microphone input level seems to be unadjustable. I have digged around in the hidden menus, where I could find a way to adjust the input level for speech calls but not for audio recording.
(Audio -> Handset ->I2S1 Rx Gain, although I am not sure if the signal clips before or after that point when doing audio recording ).
2 - This is an issue I have not seen documented anywhere: no matter what application I use, I am not able to record at any higher sampling rate than 22.050kHz. This is with both handset and headset microphone. But what is probably worse, is that the signal is not (or not very well) low pass filtered (antialised) before being downsampled!! I tried recording a frequency sweep from 0 to 16kHz, and the first 8kHz (11025 Hz inn 22050 Hz sampling rate mode) were Ok, the next 8kHz (11025Hz) where aliased like hell. Try record a 10 khz and a 12 kHz tone at 11khz adc rate and you will get a 10khz back in both cases, almost at the same amplitude ! Not sure why I can't acces 48kHz, as I read somewhere that the audio chipset should support that.
Anyone know if there is a fix for either?
Just uploaded the recorded sample for reference.
This was a 100Hz to 16kHz frequency sweep of 30 seconds.
The test setting was far from ideal (just played it out on my PC) but it gives an indication. Note that aliasing is so bad that even harmonic distortion gets aliased, and it becomes really loud when all harmonics get imaged back to the same frequency by the aliasing process. If the situation is not fixed, there is no hope for high quality audio for the galaxy... Anyone knows if with 2.2 upgrade this will work?
Do HTC desire or wildfire have similar issues? I am seriously considering swapping before my 14 days trial runs out: no point in having HD video (main selling point of the phone) if the audio is close to unuseable (and I don't want to go to Apple just yet)
The voice quality on my Galaxy S is really bad. I've had several people say to me that on calls it's difficult to understand what I'm saying and that the sound is either muffled or distorted.
I tried some recordings with the voice recorder app and these two don't sound clear at all.
Could this be caused by the same thing you're mentioning, or do I have a problem with my phone?
How do other people find the call quality (that is, someone listening to you calling *from* your Galaxy.)?
Meza1 said:
The voice quality on my Galaxy S is really bad. I've had several people say to me that on calls it's difficult to understand what I'm saying and that the sound is either muffled or distorted.
I tried some recordings with the voice recorder app and these two don't sound clear at all.
Could this be caused by the same thing you're mentioning, or do I have a problem with my phone?
How do other people find the call quality (that is, someone listening to you calling *from* your Galaxy.)?
Click to expand...
Click to collapse
I think this is related to the same bugs mentioned as the subject of this thread.
BTW, it looks like HTC desire has the same issue wrt video recording, so it might be an Android issue rather than a Samsung one...
You can find some additional details in the modaco forum, on thread samsung-galaxy-s-i9000-mic-sensitivity-recording.
(can't add the link here)
ilcello said:
(...)
2 - This is an issue I have not seen documented anywhere: no matter what application I use, I am not able to record at any higher sampling rate than 22.050kHz. (...)
Click to expand...
Click to collapse
After flashing the european firmware JM1 from Kies I am able to record at 44.1 kHz now!
Haven't checked aliasing yet.
I'm on JM2 and this is a topic that I would very much be interested in.
Sent from my GT-I9000 using XDA App
There is another older post but still active here:
http://ip208-100-42-21.static.xda-developers.com/showthread.php?t=721069
Same problem gggrrr
I recorded a friend of mine singing last night and the audio on the video is terrible and i mean really terrible, the picture is fine but the audio, nope.
I have heard that there is a way to load in a new kernal to be able to change the sensitivity of the mic but i'm not sure if that is the way to go or not, i just want to be able to record video that sounds good, please help.
Flash a kernel with voodoo sound. Then the audio quality will be amazing
Sent from my GT-I9000 using XDA App
Read this thread
http://forum.xda-developers.com/showthread.php?t=806195&page=103
I've used voodoo sound ever since I go this phone. You don't even have to flash a kernel anymore, supercurio's made it even easier, as long as you have root just go onto the market and buy Voodoo Control Plus and unlock the true audio capabilities of your Galaxy S
I started out with JPY and using Tape Machine I've been able to record just fine at 44100Hz. Same with JVB.
Thanks for that, i installed voodoo control plus, really easy to do, at a gig tonight so i will record and see what the sound is like on the video, thanks again
russellhearn69 said:
Thanks for that, i installed voodoo control plus, really easy to do, at a gig tonight so i will record and see what the sound is like on the video, thanks again
Click to expand...
Click to collapse
Thank supercurio not me
I figured I'd start this thread for everyone to discuss problems relating to call audio.
Today I tried from a fresh CM10 copying all the audio cfg files from Holics 1.3 (fixing files names to match CM10 where required), reboot, make calls.
I am certain that I got the back speaker louder for notifications music (to the point now where it distorts), however speaker phone and headset vol are still about the same.
The volume up/down has no effect on the volume. I can see in the live log that when I press the vol rocker, the phone is "setting" the volume step, but it has no audible effect.
I also tried copying the AIC and TPA and Adie cfg files from an HTC One V (which I'm guessing uses the same audio chips since the cfg files are named the same). No change detected.
While testing calls with a friend, he noticed that when he spoke in a low voice, or made non-word noises (banging, clapping, etc) he did not hear that echo. Whispering didn't echo. Just "regular" volume speech.
I'm starting to wonder if the changes made to the A1026 to enable in-call for AT&T phones have caused the echo issue (and possibly the broken vol control and/or seemingly single-level volume).
That's it for tonight. Feel free to post any steps you followed to test the issue here.
Im up for this
Yea I keep getting that. In calls it. Echoes or sounds muffled I have noise cancelation checked
Sent from my Vivid 4G using Tapatalk 2
Being that the remote user hears themselves echo, it almost sounds like the gain on the second mic is too high. So when their voice comes through on the handset speaker, the second mic pics it up, and the echo is produced. What carrier and device does everyone have? Im on ATT with a Vivid and from what I understand there's no echo, at least none that has been reported.
homeslice976 said:
Being that the remote user hears themselves echo, it almost sounds like the gain on the second mic is too high. So when their voice comes through on the handset speaker, the second mic pics it up, and the echo is produced. What carrier and device does everyone have? Im on ATT with a Vivid and from what I understand there's no echo, at least none that has been reported.
Click to expand...
Click to collapse
I'm on ATT and Vivid. Sometimes it happens sometimes it doesn't. Mostly people report I sound garbled but not with every conversation.
I'm getting reports that people are hearing themselves echo..
In Toronto on the chatr network..but the phones were originally from Rogers. On my end I have zero echo, everything is crystal clear, if a little quiet.
Sent from my Vivid 4G using xda premium
Second mic gain is what I suspect too, since quitter noises don't echo, but in my testing, I don't seem to be able to affect the mic/headset speaker gains using the configs. Though with that said, I wasn't specifically trying to reduce the top mic gain last time. I suppose that's the next step, as well as reviewing the code that loads the CSV values and sets values in software.
I'm wondering if we are not actually loading the values that are in the CSV, and are instead relying on defaults that were built in. Or the values are only loaded once. Or we have bypassed the loading or checking for changes, or something like that.
If anyone has the time to sleuth around github/XDA or wherever to find examples of kernel audio adjustments for other phones (on the same or similar boards) that would be a big help. I have had a hard time finding things on github. The advanced search seems a little less than useless to me.
I dont hear myself echo the other person does with noise cancelation unchecked and with it checked they say i sound muffled
Sent from my HTC Vivid 4G
I wonder if this is our problem - maybe the phone is ignoring the tpa config files and defaulting to stock board values, which are probably wack. I'd like to see dcacklam's git to see exactly what he's doing, but haven't found it yet. haven't looked too hard tbh. I have an idea of what he's trying to do though, I think, and may play around later - otherwise i'll try some stuff tomorrow.
He appears to be building a kernel for CM10 for the Amaze and just posted this a few hours ago (maybe helping maintain cm10, haven't searched too hard. I'm feeling lazy this evening, can you tell?):
dcacklam said:
[/COLOR]Think I have the audio-volume licked...
The phone was IGNORING the TPA2051_CFG.csv file and going with the hardcoded defaults in the driver source-file...
So I put a mix of HTC defaults, and EclipzeRemix's SpeedRom amplifier-config values (using SpeedRom numbers where no data existed in the stock TPA file) into the source-file...
And I am waiting for the kernel to build with these changes!
Click to expand...
Click to collapse
So as it turns out my remote parties hear themselves echo too
Lol - no one ever cared to tell me until tonight...
Same symptoms though - they hear themselves echo. But with me it doesn't seem to matter whether noise suppression is on or off, they still hear an echo of themselves, but hear me ok, and i hear them ok.
This may make it a little easier for me to troubleshoot though.
Sorry I've been telling everyone I'm not affected
I didnt know either until someone told me. At least we knowhere now
Sent from my HTC Vivid 4G
Hmmmm.... Could this method in board-audio.c have something to do with the volume being stuck at one volume?
Code:
int holiday_get_rx_vol(uint8_t hw, int network, int level)
{
int vol = 0;
/* to be implemented */
pr_aud_info("%s(%d, %d, %d) => %d\n", __func__, hw, network, level, vol);
return vol;
}
ess.boyer said:
Hmmmm.... Could this method in board-audio.c have something to do with the volume being stuck at one volume?
Code:
int holiday_get_rx_vol(uint8_t hw, int network, int level)
{
int vol = 0;
/* to be implemented */
pr_aud_info("%s(%d, %d, %d) => %d\n", __func__, hw, network, level, vol);
return vol;
}
Click to expand...
Click to collapse
I saw this the other day - that was the only int vol reference I found actually. I don't understand the language well enough to know exactly what's going on or how it would need to be edited. Have you tried editing the 0 value to see if anything changes? Or do you know the language wee enough to tell it to look at the tpa configs for values...or even if that's where it would need to be done?
homeslice976 said:
I saw this the other day - that was the only int vol reference I found actually. I don't understand the language well enough to know exactly what's going on or how it would need to be edited. Have you tried editing the 0 value to see if anything changes? Or do you know the language wee enough to tell it to look at the tpa configs for values...or even if that's where it would need to be done?
Click to expand...
Click to collapse
I'm doing a build right now with that value set to 3 to see if it changes anything.
I'm wondering if that is the method that catches the vol up/down commands and passes on the volume change.
If that's the case, what I would expect is that it has a usable range of 0-5. 0 is loudest, 5 is quietest. If that's true, here is where we would fix the volume control not working. However, that's not going to help us get the in-call volume up any higher.
ess.boyer said:
I'm doing a build right now with that value set to 3 to see if it changes anything.
I'm wondering if that is the method that catches the vol up/down commands and passes on the volume change.
If that's the case, what I would expect is that it has a usable range of 0-5. 0 is loudest, 5 is quietest. If that's true, here is where we would fix the volume control not working. However, that's not going to help us get the in-call volume up any higher.
Click to expand...
Click to collapse
I'm also testing a theory right now based on reverting this (https://github.com/Sportsstar89/and...mmit/9261a3aab974db0a376d3a398e4302e6ba53ca76) and this (https://github.com/Sportsstar89/and...mmit/56d56f55bf08dd25295bd93ae40b2df78ec30f52) (but leaving the att in call audio fix alone)
Comparing to HTC's sources (which could very possibly be very incomplete), these changes weren't there (obviously, or else they wouldn't be changes). Of course, we didn't have ATT in call audio with HTC's sources either. In the first commit, htc acoustic table values are directed to new sources (.csv files). In ATT's source, both audience and non audience were directed to aic3254 txt files. I'm thinking, since my previous "audiofix" which played with some TPA2051 values seemed to improve other audio, maybe these dsp .csv configs are only for non-call audio, and the in call audio requires a the aic .txt files? With the second commit - again, in HTC's source, they didn't have any indicators for ATT SKU 1 or ATT SKU 2, just ATT SKU3 had support_audience=1. What if not all of the ATT Vivid's have the audience chip? Only a select batch.(Edit: New theory for this is the support_audience=1 that was listed is actually defined for all 3 ATT SKU's, and the other entry "Default" is for the other models. So it's possible that the in call mic always worked if you screamed in it for ATT users, but maybe no one ever tried?) IFFFFF these theories were to pan out - it would possibly explain why some have echo and some don't, why in call audio seems worse for some people (and can't be adjusted), and possibly even why some ATT users have no trouble with in call audio on the first try, and others do. All theories though. h8rift seems like a smart dude, so chances are reverting these changes won't help a bit. But honestly, the ratio of things I've done that haven't worked/things I've fixed is like 200/3, so I won't be disappointed if nothing happens
Edit: Also - in HTC's sources "int vol = 0" as well
Edit: Well audio did initialize on the first call after flashing the new kernel and rebooting. Also rebooted again after that to test, and in call audio kicked in on the first call again. In the past after any kind of reboot, first call didn't work. Audio adjustment still doesn't work. I'll call my wife shortly to see if the echo still exists. But this appears to have at least killed the first call audio not working bug. If anyone else wants to test here you go:
zImage and bcmdhd.ko are here (https://docs.google.com/folder/d/0B8HUFb41pK2eQjFZMWx6UE43VnM/edit)
Don't know about other recoveries, but with WCX, in Developer Options, you can make and flash boot from zImage. Follow on screen instructions. maybe someone else can make a boot.img and I'll make a flashable package.
Edit 2: Still echoes on remote end. Slightly, but it's there. I don't know how bad it was before though - didn't try it for myself til now. Also my wife just said it still echoes, but it's not as bad. Please test and post results (in which case I have a few other thoughts as well, but I'm going to lunch now )
The couple of calls I had over lunch said there wasn't any echo. Anyone give this a go yet?
Feedback on this^ flashed it and reports are crystal clear. Only complaint was that I was really quite. The person I called had just gotten off a call with someone else and the volume was fine so it looks like it is on my side.
Thanks for the feedback. Hopefully we get a few more that report the same, then I'll start trying to tweak volumes
Sent from my Vivid 4G using Tapatalk 2
mind posting your boot.img esloan? I tried repacking homeslice's zImage into mine but it won't boot. I must be missing something.
Is there any way to disable the secondary mic? This would (a) verify that the mic is causing the echo, and (b) provide a temporary workaround until a proper solution can be found.
The microphone for the Nexus 4 is very noisy. A user on another thread described a video he recorded
to sound as if it was "recorded underwater".
I looked up the Android SDK and the MediaRecorder object has access to the following audio streams:
CAMCORDER, DEFAULT, MIC, VOICE_CALL, VOICE_COMMUNICATION, VOICE_DOWNLINK, VOICE_RECOGNITION, VOICE_UPLINK.
http://developer.android.com/reference/android/media/MediaRecorder.AudioSource.html
I noticed that while audio and video recorders sound bad on Nexus4, voice calls are of decent quality.
Not all phones have the hardware to handle all these so most sound recorders use MIC or DEFAULT.
I found the source code to the MIUI Sound Recorder and it was easy to add a setting that lets you
manually select which hardware stream to record from.
I ran some tests on the Nexus 4 and came to the following conclusions:
- MIC, DEFAULT and VOICE RECOGNITION are probably the same stream and sound bad.
- the CAMCORDER audio stream is the cleanest.
I propose that until Google and LG fix this issue ROM developers alter the Android API on Nexus 4 (only)
so that the CAMCORDER stream always overwrites MIC and DEFAULT.
I'm attaching the modified MIUI Sound Recorder with manual stream select if anyone wants to do
tests on their own. Here's an amplified view of MIC vs CAMCORDER.
-Mindroid- said:
The microphone for the Nexus 4 is very noisy. A user on another thread described a video he recorded
to sound as if it was "recorded underwater".
I looked up the Android SDK and the MediaRecorder object has access to the following audio streams:
CAMCORDER, DEFAULT, MIC, VOICE_CALL, VOICE_COMMUNICATION, VOICE_DOWNLINK, VOICE_RECOGNITION, VOICE_UPLINK.
http://developer.android.com/reference/android/media/MediaRecorder.AudioSource.html
I noticed that while audio and video recorders sound bad on Nexus4, voice calls are of decent quality.
Not all phones have the hardware to handle all these so most sound recorders use MIC or DEFAULT.
I found the source code to the MIUI Sound Recorder and it was easy to add a setting that lets you
manually select which hardware stream to record from.
I ran some tests on the Nexus 4 and came to the following conclusions:
- MIC, DEFAULT and VOICE RECOGNITION are probably the same stream and sound bad.
- the CAMCORDER audio stream is the cleanest.
I propose that until Google and LG fix this issue ROM developers alter the Android API on Nexus 4 (only)
so that the CAMCORDER stream always overwrites MIC and DEFAULT.
I'm attaching the modified MIUI Sound Recorder with manual stream select if anyone wants to do
tests on their own. Here's an amplified view of MIC vs CAMCORDER.
Click to expand...
Click to collapse
I modified my kernel to fix this and I used your apk to test for both mic and camcorder and I don't notice the difference with my updated kernel.
http://faux.androidro.ms/mako/beta/mako-jb-kernel-004b06.zip
check for me to see if I really fix it or not since I am not an audio expert
Thank you, faux!
I've been meaning to try your kernel for a while so now I just have to.
Please treat my ideas about how Android exposes the audio stream as a speculation for now.
I didn't run any in-depth tests and I didn't trace the kernel code.
To my ears the MIC/DEFAULT streams are very noisy. The CAMCORDER one is noisy also
and but somewhat less so. They all sound bandpassed (probably a microphone limitation).
I'd like to know what other users think of the sound.
Very interesting.
I agree that N4 mic is so sensitive that it picks up a lot of background noises. I have been tweaking with mic volume to offset it.
If OP is true, then there's a hope for VoIP users who use Csipsimple. Without changing the kernel, csip can use CAMCORDER mode by changing the settings under media>audio troubleshooting.
The very high mic gain is the main issue with the sound quality. It needs to be adjusted from
the kernel driver which is not a trivial procedure. Gain from apps works on top of the hardware
gain and does not affect distortion that happens earlier in the signal chain.
One user put scotch tape on the mic to mechanically attenuate the sound and reported good results.
If the the sound system on the nexus 4 is built on top of ALSA, I.e. if the directory /proc/asound exists, then microphone levels can be adjusted in user-space. There is an app (ALSAMixer) available on the market that could be used here.
What would be the effect of covering the mic with something to reduce it's sensitivity?
chdloc said:
If the the sound system on the nexus 4 is built on top of ALSA, I.e. if the directory /proc/asound exists, then microphone levels can be adjusted in user-space. There is an app (ALSAMixer) available on the market that could be used here.
Click to expand...
Click to collapse
I'm not sure if that will do it. The textbook audio recording chain goes like this:
microphone -> preamp -> analog-to-digital converter
If I'm not mistaken ALSA(the driver) will take over afterwards. We need a way to control
the hardware gain on the preamp. It's possible that preamp gain is permanently fixed by design.
On the other hand almost all modern preamps have automatic variable gain so it might just be
a matter of figuring out how to turn it on.
EDIT:
ALSAMixer doesn't work at all on my phone (stock 4.2.1 - rooted)
Do you know of any custom kernels for the Nexus4 that use ALSA?
donec said:
What would be the effect of covering the mic with something to reduce it's sensitivity?
Click to expand...
Click to collapse
Aside from the ugly look it should work OK. I would suggest using some kind of soft thread fabric.
Wool is one of the best materials for sound absorption but you should try different materials.
The downside is that mechanical methods don't attenuate all frequencies linearly - they work much better
on higher frequencies so you will have to try and see what material works best.
If this works out it would be ideal to open the phone and plug the microphone hole on the inside with the proper material.
I don't have a Nexus 4, but I do know that the Nexus 7 uses ALSA. Have you checked for /proc/asound? If it exists, then you may need the proper ALSA libraries first.
Sent from my Nexus One using xda app-developers app
im starting to think LG just used cheap mics to keep cost down. From the buzzing to the sub par recording quality
-Mindroid- said:
Aside from the ugly look it should work OK. I would suggest using some kind of soft thread fabric.
Wool is one of the best materials for sound absorption but you should try different materials.
The downside is that mechanical methods don't attenuate all frequencies linearly - they work much better
on higher frequencies so you will have to try and see what material works best.
If this works out it would be ideal to open the phone and plug the microphone hole on the inside with the proper material.
Click to expand...
Click to collapse
I have my Nexus 4 in a case and thought maybe I could improve the quality by placing a piece of cloth between the case and the phone.
Lockeskidney said:
im starting to think LG just used cheap mics to keep cost down. From the buzzing to the sub par recording quality
Click to expand...
Click to collapse
The specs for the microphone do not look bad either. http://www.knowles.com/search/prods_pdf/SPU0410LR5H.pdf is the spec sheet and it should be working to 116 dB, which is louder than I'd want to be in - though it is only the loud end of rock concerts. Noise also looks okay on the spec sheet, but you cannot expect Schoeps quality of course..
The underwater sounds I'm hearing with the sound recording tool remind me of processing, heavy compression andf such. Not what I heard on my camcorder tests, which had a pleasant sound.
Overload is more likelely to be caused by audio codec settings. A voice call has the mouth close to the mic, that is loud too, and works well. But applications like movie recordings crank up the gain to record sounds further away.
jutezak said:
The specs for the microphone do not look bad either. http://dl-bacon.bbqdroid.org/GeebCM/EN_LG-E973_SVC_ENG_121127.pdf is the spec sheet and it should be working to 116 dB, which is louder than I'd want to be in. Noise also looks okay on the spec sheet, but you cannot expect Schoeps quality of course..
The underwater sounds I'm hearing with the sound recording tool remind me of processing, heavy compression andf such. Not what I heard on my camcorder tests, which had a pleasant sound.
Overload is more likelely to be caused by audio codec settings. A voice call has the mouth close to the mic, that is loud too, and works well. But applications like movie recordings crank up the gain to record sounds further away.
Click to expand...
Click to collapse
Nice find! Thank you for sharing!
Microphones in general are almost never bad. It's the preamps and digital processing that add more to the mix.
Regarding the "underwater" sound I now suspect a lack of a Low Pass filter. Check out my other thread:
http://forum.xda-developers.com/showpost.php?p=36856100&postcount=18
jutezak said:
The specs for the microphone do not look bad either. http://www.knowles.com/search/prods_pdf/SPU0410LR5H.pdf is the spec sheet and it should be working to 116 dB, which is louder than I'd want to be in - though it is only the loud end of rock concerts. Noise also looks okay on the spec sheet, but you cannot expect Schoeps quality of course..
Click to expand...
Click to collapse
Human speech can generate 116dB SPL measured at 1-2" distance fairly easily. I just measured the level my speech at 1" distance (with a lab-grade device) and I did not have to shout to produce 116dB SPL. The spec sheet says that the microphone produces about 10% THD typically at 118SPL for a 1kHz tone.
jutezak said:
Overload is more likelely to be caused by audio codec settings. A voice call has the mouth close to the mic, that is loud too, and works well. But applications like movie recordings crank up the gain to record sounds further away.
Click to expand...
Click to collapse
Distortion of the microphone signal is generated (or amplified) in at least four locations:
1. The microphone itself (analog or digital if the microphone is a digital microphone as many MEMS are; the above referenced Knowles mike is analog)
2. The amplifier (analog)
3. The analog-to-digital converter (ADC) which is spec'ed with a maximum input voltage to generate a signal that hits "digital rail", i.e. 0dB FS (full-scale)
4. digital gain along the digital processing path
Potential "clipping" (i.e. signals producing 0dB FS at the ADC) can only be avoided if the analog amplifier (if there is one) can be controlled. By looking at the ALSA mixer settings for the Nexus 4 my only hope lies in the controls that are labeled "ADCX Volume" where X=[1,6], i.e. numid. I believe -Mindroid- shared his "idle" settings vs. settings during a call and only X={1,3}, i.e. numid={48,50} have changed. Give changing those two parameters a try...
I posted this in a different thread, but thought it would be useful here as well as a potential lead to a workaround:
Looks like the audio chip in question is a Qualcomm WCD9310.
It's used in a few other devices as well (eg: GS3 USA). If their audio runs fine, perhaps we could try running their audio drivers on our device?
I did this on my N5 and I recall it being mixer_paths.xml and then you find something along the lines of path name=handset or incall etc. followed by a line RX(X) Digital Volume value= (something along the lines of 80). I did a search and that was about it so I went into this expecting to find something much the same.
I can find reference to RX digital volumes in the listing of initial values at the start of the file and these have an expected value of 84 but I cant find a path name= with a likely suspect like handset, incall, voice, etx. followed by this line where I would make the actual change.
Anyone else dug into this yet? I'm wondering if I can simply make the changes in the initial values? I cant remember if the N5 had those lines but I'm guessing it must have. There are seven of them as you would expect with the stock call volume stepping. I've been contemplating giving them a small nudge to 86 and see what happens, good idea or not?
As an aside I found a line for in call recording mode which is currently set to 0...
krabman said:
I did this on my N5 and I recall it being mixer_paths.xml and then you find something along the lines of path name=handset or incall etc. followed by a line RX(X) Digital Volume value= (something along the lines of 80). I did a search and that was about it so I went into this expecting to find something much the same.
I can find reference to RX digital volumes in the listing of initial values at the start of the file and these have an expected value of 84 but I cant find a path name= with a likely suspect like handset, incall, voice, etx. followed by this line where I would make the actual change.
Anyone else dug into this yet? I'm wondering if I can simply make the changes in the initial values? I cant remember if the N5 had those lines but I'm guessing it must have. There are seven of them as you would expect with the stock call volume stepping. I've been contemplating giving them a small nudge to 86 and see what happens, good idea or not?
As an aside I found a line for in call recording mode which is currently set to 0...
Click to expand...
Click to collapse
take a look here
if u need somethig tell me,but without root we can t do more
https://drive.google.com/file/d/0BwTULBGmnkwqWjJKWlRZUFJ0bm8/view?usp=sharing
I have root (maxx) and I've been going through the build.prop and some other files including the mixer_paths.xml and mixer_path_i2s.xml and I cannot find the digital volume= X in any likely looking path=. I think it has to be there and I'm just missing it or else it may be we would adjust the initial volume? In i2s.xml the digital volume shown only has 4 steps and still doesn't come after any path that looks likely. It would seem to me that rx1 through 7 are certainly a smoking gun. DEC should be speaker, ADC mic, and HPHL/R would be headphone correct?
Since no one jumped in with a warning I went ahead and backed up mixer_path.xml and then edited the values mentioned changing the stock initial value of 84 to a value of 86 on RX1 through RX7 and it appears to have worked over a couple calls I made although it may be I got lucky with a couple of unusually good connections. If someone wants to give it a go and verify the results we may have a working method of increasing the in call volume. Anyone testing that hasn't done this before should be aware that small changes in the values have a large effect so its best to take baby steps, keep in mind the stock speaker may not have much headroom so play it safe unless you're feeling brave. Be sure to backup the file before editing.
krabman said:
Since no one jumped in with a warning I went ahead and backed up mixer_path.xml and then edited the values mentioned changing the stock initial value of 84 to a value of 86 on RX1 through RX7 and it appears to have worked over a couple calls I made although it may be I got lucky with a couple of unusually good connections. If someone wants to give it a go and verify the results we may have a working method of increasing the in call volume. Anyone testing that hasn't done this before should be aware that small changes in the values have a large effect so its best to take baby steps, keep in mind the stock speaker may not have much headroom so play it safe unless you're feeling brave. Be sure to backup the file before editing.
Click to expand...
Click to collapse
i will test it
tell me what to do
Edited both files and raised volumes to 95 for RX values and to 90 the Dec values that were arrounf 70 and 80 something. Clearly a noticeable improovement. Even call volume is raised and sounds excellent.
---------- Post added at 07:34 PM ---------- Previous post was at 07:28 PM ----------
These are my edited Moto Maxx mixer files. Use if you like and check if they work for you. Remember, only for Maxx, and i am not responsible if your phone catches the Ebola virus. Just unzip files and move to system/etc folder and set correct permissions. Let me know.
https://copy.com/x3TmPejOwitAwDab
The DEC should be speaker, headphones can also be adjusted in the same way which should be HPHL and HPHR. I thought we had a runner but I wasn't positive because call quality is so hit and miss around here. I'd be wary of making that large an adjustment but I may be overly paranoid about it because I blew the speaker on my sgs3 this way. The thing there being I only made a modest change much like I did here but in hindsight it would later become apparent by the number of blow speakers people had that they had limited the volume on those release handsets to where they did because that was as much as the speaker could take.
You need root danger, I think you are on the Turbo aren't you? Maybe I'm mistaken, at any rate all you have to do is make a backup, mount the file, edit the initial values, save it and reboot.
krabman said:
The DEC should be speaker, headphones can also be adjusted in the same way which should be HPHL and HPHR. I thought we had a runner but I wasn't positive because call quality is so hit and miss around here. I'd be wary of making that large an adjustment but I may be overly paranoid about it because I blew the speaker on my sgs3 this way. The thing there being I only made a modest change much like I did here but in hindsight it would later become apparent by the number of blow speakers people had that they had limited the volume on those release handsets to where they did because that was as much as the speaker could take.
You need root danger, I think you are on the Turbo aren't you? Maybe I'm mistaken, at any rate all you have to do is make a backup, mount the file, edit the initial values, save it and reboot.
Click to expand...
Click to collapse
we don t have root on Droid turbo
The Galaxy S9 is an audio powerhouse. It has the first set of stereo speakers on a Samsung flagship, and it even comes standard with a set of AKG-tuned earbuds that would normally cost $99. But if you want to further enhance your audio experience, there's a feature that will customize audio output to your own specific hearing.
The feature, aptly called Adapt Sound, is hidden deep within the settings menu on your S9 or S9+, but will make a world of difference in audio quality once enabled. As its name implies, Adapt Sound will fine tune your S9's audio and customize it to perfectly match your hearing, which dramatically enhances the perceived quality of any set of headphones.
Step 1 Find the 'Adapt Sound' Menu
To start, head to your phone's main Settings menu, then select "Sounds and Vibration." From there, scroll down and tap on "Sound quality and effects" found along the bottom half of the menu.
Now, tap on "Adapt Sound" near the bottom of the screen. You'll be asked to grant phone call related permissions, so tap "Allow" on the prompt. Within "Adapt Sound," leave the available age-based sound profiles alone, and tap on "Add Personalized Sound Profile."
Step 2 Take the Hearing Test
Now, go to a totally quiet location and plug in your headphones. Tap on the "Start" button on the bottom when you're ready. Adapt Sound will now run a hearing test, which consists of a series of beeps of varying frequencies from the left and right earbuds. It'll ask if you can hear each tone along the way, so take your time and answer honestly in order to create an accurate profile.
Once the test reaches 100%, you'll now notice a personalized equalizer catered to each ear. Now that you've completed the test, you're free to move on to the final step.
Step 3 Finalize Your Audio Profile
After finishing the tests, Adapt Sound will ask for you to choose your preferred ear for phone calls. Once you've made your decision, your custom audio profile will be activated and made available for preview. So tap on "Preview" and check out the difference Adapt Sound makes by toggling between "Personalized" and "Original," along with "Left," "Right," and "Both" ears.
In addition to this, you have the option of naming your sound profile — simply tap on "Profile name" and enter your desired label when prompted. After that, tap on "Save" in the upper-right corner to save your profile and finish up with setup.
With Adapt Sound enabled, we highly recommend playing your favorite song to determine exactly how much of an impact it's made on sound quality. It'll make a huge difference on how you hear audio, especially if your hearing isn't great to begin with.
Xperience Z said:
The Galaxy S9 is an audio powerhouse. It has the first set of stereo speakers on a Samsung flagship, and it even comes standard with a set of AKG-tuned earbuds that would normally cost $99. But if you want to further enhance your audio experience, there's a feature that will customize audio output to your own specific hearing.
The feature, aptly called Adapt Sound, is hidden deep within the settings menu on your S9 or S9+, but will make a world of difference in audio quality once enabled. As its name implies, Adapt Sound will fine tune your S9's audio and customize it to perfectly match your hearing, which dramatically enhances the perceived quality of any set of headphones.
Step 1 Find the 'Adapt Sound' Menu
To start, head to your phone's main Settings menu, then select "Sounds and Vibration." From there, scroll down and tap on "Sound quality and effects" found along the bottom half of the menu.
Now, tap on "Adapt Sound" near the bottom of the screen. You'll be asked to grant phone call related permissions, so tap "Allow" on the prompt. Within "Adapt Sound," leave the available age-based sound profiles alone, and tap on "Add Personalized Sound Profile."
Step 2 Take the Hearing Test
Now, go to a totally quiet location and plug in your headphones. Tap on the "Start" button on the bottom when you're ready. Adapt Sound will now run a hearing test, which consists of a series of beeps of varying frequencies from the left and right earbuds. It'll ask if you can hear each tone along the way, so take your time and answer honestly in order to create an accurate profile.
Once the test reaches 100%, you'll now notice a personalized equalizer catered to each ear. Now that you've completed the test, you're free to move on to the final step.
Step 3 Finalize Your Audio Profile
After finishing the tests, Adapt Sound will ask for you to choose your preferred ear for phone calls. Once you've made your decision, your custom audio profile will be activated and made available for preview. So tap on "Preview" and check out the difference Adapt Sound makes by toggling between "Personalized" and "Original," along with "Left," "Right," and "Both" ears.
In addition to this, you have the option of naming your sound profile — simply tap on "Profile name" and enter your desired label when prompted. After that, tap on "Save" in the upper-right corner to save your profile and finish up with setup.
With Adapt Sound enabled, we highly recommend playing your favorite song to determine exactly how much of an impact it's made on sound quality. It'll make a huge difference on how you hear audio, especially if your hearing isn't great to begin with.
Click to expand...
Click to collapse
How is the audio over head/ear phones? I previously had the Galaxy S7 and while it was a great device, I found the audio output via the 3.5mm socket to be pretty underwhelming. And I had tuned it using the similar software function above, but it was still under par for me. I haven't tried the S8 but seen a review that said the audio/earphone quality still wasn't that great. So hoping the S9 has been improved, as they have been focusing on improving audio albeit via the external speakers. Im in the UK so believe it would be the Exynos model if that makes any difference.
Sent from my Redmi Note 4 using Tapatalk
I just did this and it made quite a difference. Thanks for sharing this.
gsmyth said:
How is the audio over head/ear phones? I previously had the Galaxy S7 and while it was a great device, I found the audio output via the 3.5mm socket to be pretty underwhelming. And I had tuned it using the similar software function above, but it was still under par for me. I haven't tried the S8 but seen a review that said the audio/earphone quality still wasn't that great. So hoping the S9 has been improved, as they have been focusing on improving audio albeit via the external speakers. Im in the UK so believe it would be the Exynos model if that makes any difference.
Click to expand...
Click to collapse
Juice3250 said:
I just did this and it made quite a difference. Thanks for sharing this.
Click to expand...
Click to collapse
For me made it a quite difference also.
The sound is very nice. And with this small trick much better.
I turned on the Atmos setting, and immediately noticed the difference, just on speakers as the headset is in the car.
Dolby Atmos is terrible
wmharley said:
I turned on the Atmos setting, and immediately noticed the difference, just on speakers as the headset is in the car.
Click to expand...
Click to collapse
if you use high quality headphones, you can noticeably tell that atmos is terrible. it compresses the sound and gives it an unnatural vibe. I find that none of the extra features actually enhances, so i leave everything off. fellow members at head-fi have found the same results I have.
I need to try this with my beats x
adapt sound is also dependent on your headphones. you may think you have lost some hearing if you use headphones that have narrower frequencies, because that pair is just unable to output certain frequencies, so do this test with headphones that can output more, if you want a more accurate result of your current hearing.
radiohead14 said:
if you use high quality headphones, you can noticeably tell that atmos is terrible. it compresses the sound and gives it an unnatural vibe. I find that none of the extra features actually enhances, so i leave everything off. fellow members at head-fi have found the same results I have.
Click to expand...
Click to collapse
For me the phones amp is kind of weak so atoms gives it a volume boost in a conveniently easy push of 1 button.
Without it over ear headphones that are 32 ohms are slightly below my preferred listening volume. With ear buds the volume is fine as long as I have a great seal. I find the included AKG's decent in sound quality but not very compatible with my ears shape so I'm constantly pushing them in.
I did the Headphine Sound setup but didn't find it to have made a vast difference in my listening experience. Maybe if toggling between profiles I could notice a little difference but nothing that makes me feel like a whole new world has opened up to me.
I've used the SHPS 9500s over ears, Westone w40 IEM, and the included AKG IEM.
I listened to a HiRes Flac 192/24 and felt like the track lacked life without the Atmos turned on, this was especially the case when using over ear headphones.
Will try with my PSB M4U 1 closed back over ear phones soon.
mc_365 said:
For me the phones amp is kind of weak so atoms gives it a volume boost in a conveniently easy push of 1 button.
Without it over ear headphones that are 32 ohms are slightly below my preferred listening volume. With ear buds the volume is fine as long as I have a great seal. I find the included AKG's decent in sound quality but not very compatible with my ears shape so I'm constantly pushing them in.
I did the Headphine Sound setup but didn't find it to have made a vast difference in my listening experience. Maybe if toggling between profiles I could notice a little difference but nothing that makes me feel like a whole new world has opened up to me.
I've used the SHPS 9500s over ears, Westone w40 IEM, and the included AKG IEM.
I listened to a HiRes Flac 192/24 and felt like the track lacked life without the Atmos turned on, this was especially the case when using over ear headphones.
Will try with my PSB M4U 1 closed back over ear phones soon.
Click to expand...
Click to collapse
agree that the phone's amp is weak. the earphone i used for testing, a Sennheiser IE80, doesn't have any issues, but the stock AKG and various Sony I tried were on the weaker side. Atmos, I found, also varies on the track played, but the increase in volume you hear is probably due to the compression, so it'll give you that increase in volume, but you may also not notice that certain frequencies have been dampened. Testing various tracks with different earphones, I found that in most cases, Atmos just muddies songs.. almost like a blanket was put over it. The only times I found Atmos to be preferable was when playing older tracks that weren't mixed well and were on the trebly side. Atmos hid some sibilance due to my previous finding of Atmos just kinda putting a dampen on most tracks.
also agree on the stock AKG's not being the ideal in-ear fit, as i too have found myself constantly having to adjust (i commute in NYC transit, so this was a good test of isolation). the littlest movement, and you lose a lot of low end and volume.
i suggest you do the adapt sound test with the headphones you will use the most with the phone, as it is highly dependent on the frequencies. and if your hearing is still really good, then you may not find much benefit to it. it's mainly pushing frequencies higher if your hearing has lost some ability to discern those. so if your hearing is great, then you may not find much difference.
gsmyth said:
How is the audio over head/ear phones? I previously had the Galaxy S7 and while it was a great device, I found the audio output via the 3.5mm socket to be pretty underwhelming. And I had tuned it using the similar software function above, but it was still under par for me. I haven't tried the S8 but seen a review that said the audio/earphone quality still wasn't that great. So hoping the S9 has been improved, as they have been focusing on improving audio albeit via the external speakers. Im in the UK so believe it would be the Exynos model if that makes any difference.
Click to expand...
Click to collapse
I seriously feel like my S7 was way louder... Im even considering to root this device just to edit the mixer gains xml for more volume... this is really dissapointing.
Rule n5
5. Create a thread topic or post a message only once, this includes external links & streaming media.
As a large forum, we don't need unnecessary clutter. You're free to edit your message as you like, so if you do not receive an answer, revisit your message and see if you can describe your problem better. Not everyone is online at the same time so it might take a while before you receive an answer.
You can bump your unanswered question once every 24 hours
Duplicate threads and posts will be removed
Always post in an existing thread if a topic already exists, before creating a new thread.
Use our search function to find the best forum for your device.
Links to an external source are only allowed if relevant to the topic in hand. A description must be included, no copy & pasting from the original source.
Self-promotion is forbidden, this includes blogs, social media and video channels etc. Random links will be removed.
Since this thread is a copy/paste from original source, is now closed!