Solved Microphone Recording Quality Issue on Fedora Linux with OBS Studio (but not only OBS)

Solved issue

KeYunLong

New Member
Joined
Jul 22, 2022
Messages
29
Reaction score
12
Credits
630
Hello everyone,

I am using Fedora Linux and OBS Studio for recording, and I am facing issues with the headset microphone recording quality. In the past, I used Windows with GeForce Experience, and the recording quality was significantly better. I am hoping to get some advice on how to improve the recording quality on Fedora.

Please find my specs attached.

My question is what should I look for to improve the recorded voice? I can record a sample if you want.

PS. I also checked Audacity and the recorded voice is also terrible.

Thanks!
KeYunLong
 

Attachments



This suggests your mic is low quality, if it's part of a headset.
Do you have specs about your mic?

Does your distro use pipewire or pulseaudio for audio?
pipewire is better.

Hi CaffeineAddict,

Thank you for your response!I understand your concern, and I want to clarify that I'm not claiming that my microphone is of high quality. However, the significant difference in recording quality between Windows and Fedora Linux with the same equipment is what's puzzling me. On Windows, the recordings were clear and usable, but on Fedora, the recorded voice sounds very poor, almost like it was recorded by a calculator. I attached some specs already in .txt file, but I guess that people might not want to open them in this way. So:

System Information​



Software Information​


  • OBS Studio Version: 31.0.0 (64 bit)
  • ALSA Version:
aplay -v /dev/zero

Playing raw data '/dev/zero' : Unsigned 8 bit, Rate 8000 Hz, Mono

ALSA <-> PipeWire PCM I/O Plugin

Its setup is:

stream : PLAYBACK

access : RW_INTERLEAVED

format : U8

subformat : STD

channels : 1

rate : 8000

exact rate : 8000 (8000/1)

msbits : 8

buffer_size : 4000

period_size : 1000

period_time : 125000

tstamp_mode : ENABLE

tstamp_type : GETTIMEOFDAY

period_step : 1

avail_min : 1000

period_event : 0

start_threshold : 4000

stop_threshold : 4000

silence_threshold: 0

silence_size : 0

boundary : 9007199254740992000


pipewire --version

pipewire

Compiled with libpipewire 1.2.7

Linked with libpipewire 1.2.7

  • PulseAudio Version: not installed

sudo dmesg | grep -i audio
[ 4.429437] snd_hda_intel 0000:01:00.1: Handle vga_switcheroo audio client
[ 4.474234] snd_hda_codec_realtek hdaudioC0D0: autoconfig for ALC897: line_outs=1 (0x14/0x0/0x0/0x0/0x0) type:line
[ 4.474240] snd_hda_codec_realtek hdaudioC0D0: speaker_outs=0 (0x0/0x0/0x0/0x0/0x0)
[ 4.474242] snd_hda_codec_realtek hdaudioC0D0: hp_outs=1 (0x1b/0x0/0x0/0x0/0x0)
[ 4.474245] snd_hda_codec_realtek hdaudioC0D0: mono: mono_out=0x0
[ 4.474246] snd_hda_codec_realtek hdaudioC0D0: dig-out=0x11/0x0
[ 4.474247] snd_hda_codec_realtek hdaudioC0D0: inputs:
[ 4.474249] snd_hda_codec_realtek hdaudioC0D0: Rear Mic=0x18
[ 4.474250] snd_hda_codec_realtek hdaudioC0D0: Front Mic=0x19
[ 4.474251] snd_hda_codec_realtek hdaudioC0D0: Line=0x1a
[ 12.149883] usbcore: registered new interface driver snd-usb-audio


Comparison with Windows​


  • Windows Setup:
    • Used GeForce Experience for recording.
    • Microphone settings were set to default.
    • The recording quality was clear with no background noise or distortion.

Specific Issues​


  • Quality Issues:
    • Background noise (this can be removed in OBS, so it's not a problem)
    • Distortion <-- this is the biggest problem
I don't have to have a radio quality voice. Just to have at least similar to the one from Windows. Thanks in advance!
 
Last edited:
I attached some specs already in .txt file, but I guess that people might not want to open them in this way.
Apologies, I opened it with book editor by mistake and it showed me only a portion, didn't see other specs.

I had similar problem with my mic, turned out I had to have it close to my mouth to get better recording, you can try if this makes a difference.

I don't know what to suggest, I'd start by adjusting which ever UI settings you have.
 
Apologies, I opened it with book editor by mistake and it showed me only a portion, didn't see other specs.

I had similar problem with my mic, turned out I had to have it close to my mouth to get better recording, you can try if this makes a difference.

I don't know what to suggest, I'd start by adjusting which ever UI settings you have.
I appreciate the suggestion about positioning the microphone closer to my mouth. However, the issue I'm experiencing is not related to the voice level or distance. The distortion and poor quality are present regardless of how close the microphone is to my mouth.The main issue is that the recordings on Fedora sound significantly worse compared to the same setup on Windows. On Windows, using GeForce Experience, the recordings were clear and free of distortion. This leads me to believe that the problem lies in the configuration or drivers on Fedora.
 
if you were using GeForce Experience before, you've got an nvidia card. if it's a newer model you could try their open source driver package - there's been some discussion on here about it recently.
 
Isn't geforce experience windows.exe only? How are you running this? Are you running the beta version?
Native Linux application will always run better than software running through an emulator.

Which geforce video card are you using? I don't see this in your specs.
If you simply need to do screen recording spectacle recorder generally works very well.
(There are different ones for different desktops)
 
Last edited:
Hi theLegionWithin and dos2unix,

Thank you both for your responses and for taking the time to help me out.
I apologize for any confusion. I mentioned GeForce Experience as an example of a recording tool I used on Windows, where the recording quality was significantly better. The purpose of this thread is to improve the recording quality on Fedora using OBS Studio, but I am open to using any other tool, such as Audacity, which I have also tried without success.
GeForce Experience: It is indeed a Windows-only application. I am not running it on Linux, and I am not looking to troubleshoot it. The comparison with Windows was to highlight the quality difference with the same equipment.

sudo lspci | grep -i vga
01:00.0 VGA compatible controller: NVIDIA Corporation GA104 [GeForce RTX 3060 Ti Lite Hash Rate] (rev a1)
glxinfo | grep -i vendor
server glx vendor string: SGI
client glx vendor string: NVIDIA Corporation
OpenGL vendor string: NVIDIA Corporation

sudo glxinfo | grep -i version
server glx version string: 1.4
client glx version string: 1.4
GLX version: 1.4
OpenGL core profile version string: 4.6.0 NVIDIA 565.77
OpenGL core profile shading language version string: 4.60 NVIDIA
OpenGL version string: 4.6.0 NVIDIA 565.77
OpenGL shading language version string: 4.60 NVIDIA
OpenGL ES profile version string: OpenGL ES 3.2 NVIDIA 565.77
OpenGL ES profile shading language version string: OpenGL ES GLSL ES 3.20
GL_EXT_shader_group_vote, GL_EXT_shader_implicit_conversions,


I am using the proprietary NVIDIA driver because the open-source Nouveau driver bricked my PC. The proprietary driver has been working well for me so far.
 
Last edited:
I'm not sure I understand. Are you recording video, or audio? Or both?

Your video driver shouldn't affect audio really.
When you watch videos. Does the audio playback reasonably well?
 
What does your sound input look like?

1737041423582.png
 
I'm not sure I understand. Are you recording video, or audio? Or both?

Your video driver shouldn't affect audio really.
When you watch videos. Does the audio playback reasonably well?
Thank you for your response!

To be honest, I’m not entirely sure why we’re discussing the GPU either, but theLegionWithin asked about it earlier, so I provided the information, thinking they might see some connection to the microphone issue. I don’t personally see how the GPU would affect the microphone, but I wanted to be thorough in case there was something I was missing.
I’ve tested both scenarios:
  1. Screen capture with microphone (e.g., recording the screen and audio together).
  2. Microphone-only recording (e.g., just recording audio without video).
In both cases, the result was the same: poor microphone quality with distortion. So, the issue doesn’t seem to be tied to video recording or GPU-related tasks.

When I watch videos in the browser (e.g., YouTube), I don’t notice any issues with audio playback. Similarly, playing MP3 files locally on Fedora works perfectly fine. This suggests that the playback side of the audio system is functioning as expected.
I suspect the issue might lie in how my system is handling the microphone input. For example, it might be trying to use the microphone in an unusual way (e.g., through a different input method or configuration). I’ve attached screenshots of my system settings below, which might help identify the problem.
1737042446235.png


1737042474869.png

1737042490533.png


1737042502392.png

1737042551315.png
 
I see you have this plugged into the front audio port. (That should work OK)

But usually for headsets the input (microphone) and the output (speakers) go through the same audio port.

Just for testing purposes, can you try the ports on the back of the computer? Some computers
have separate microphone and speaker ports. Maybe try recording with the microphone port?
 
I see you have this plugged into the front audio port. (That should work OK)

But usually for headsets the input (microphone) and the output (speakers) go through the same audio port.

Just for testing purposes, can you try the ports on the back of the computer? Some computers
have separate microphone and speaker ports. Maybe try recording with the microphone port?
Thank you for your suggestion!

You are correct that the headset is currently plugged into the front audio port. However, based on the documentation for the Tronsmart Glary Alpha Gaming Headset, it seems that a Y-adapter cable is required to properly connect the microphone and audio output.

Documentation​


  • Model: Glary Alpha
  • Input Power: 30mW
  • Sensitivity: 101dB ± 3dB
  • Driver Unit: 50mm dynamic
  • Impedance: 32 Ohms ± 3
  • Microphone Impedance: 2.2K Ohms
  • Microphone Sensitivity: -42 ± 3dB
  • Package Contents: 1 x Glary Alpha gaming headset, 1 x Y-adapter cable for PC, 1 x Warranty card, 1 x User manual

FAQ​


  • Why doesn't my microphone work when connecting to a PC?
    • When connecting to a PC, you need to plug in a Y-adapter cable to use audio and microphone normally.

Setup at the time I created this topic​

  • Headset Jack: Front audio port
  • Headset USB: Front USB port

Setup I have now​

  • Use the Y-adapter cable. (it's taken from the internet but this type)
1737059265159.png

  • Audio Jack: Connected to the green audio port on the back of the computer.
  • Microphone Jack: Connected to the pink microphone port on the back of the computer.
  • USB: Connected to a USB port on the computer. (but this is only for LED)
Previously I haven't used the Y adapter - totally forgot about this.

Originally, what I noticed after switching to back ports and started using Y adapter is that when I recorded the via Audacity I had less distortion (but much quiet and with bigger noise I think). On OBS Studio I didn't notice a difference.

After spending much time and playing with OBS Studio filters now, I can't say if it's better than before, bit better, or totally no progress. I can only say that it still doesn't sound like on Windows + Geforce Experience.

I need to go sleep and have a "fresh ear" to tell if there is a progress.
 
I can only say that it still doesn't sound like on Windows + Geforce Experience.

Is this only true when you're using OBS, or is this the case in other applications?

(You can quickly record a track with Audacity - which also works on Windows. You could then compare the two to see if there's a significant difference.)
 
Is this only true when you're using OBS, or is this the case in other applications?

(You can quickly record a track with Audacity - which also works on Windows. You could then compare the two to see if there's a significant difference.)
I don't have Windows anymore, because I fully moved to Linux. I will try to record samples using my headset in both applications on my friend's PC and compare it with samples from my Fedora.

Stay tuned! (this can take some time)
 
Previously I haven't used the Y adapter - totally forgot about this.

That could make a difference, but what I meant was, can you use the microphone of the headset alone, without plugging the speakers in.

Try to record something with only the microphone plugged in. No speakers. Use the microphone port.
Then once it's recorded, plug the speakers back in, and listen normally.

If this works, then we know it isn't drivers. Maybe it's the adapter.
 
That could make a difference, but what I meant was, can you use the microphone of the headset alone, without plugging the speakers in.

Try to record something with only the microphone plugged in. No speakers. Use the microphone port.
Then once it's recorded, plug the speakers back in, and listen normally.

If this works, then we know it isn't drivers. Maybe it's the adapter.
Hi dos2unix,

This specific model has gold-plated 3.5 mm + USB connector. Honestly I forgot/didn't notice the Y-adapter cable when I bought the headset and I used the headset in my PC in the way that I just plugged in 3.5 mm and USB to the front port (the top of my PC tower) without the Y-adapter at all. It worked good enough when I was on Windows.

1737115322043.png


Partially I tried to do what you said - I connected only to the pink and recorded the sound, but of course I couldn't listen to the recorded file, probably I deleted the sample and started recording with another configuration - thinking about it now it wasn't smart at all... I'll check your suggestion later today, if I don't see any difference, I'll take KGIII's advice and record samples using the same application on my friend's Windows PC. I think he has a good quality standalone microphone which he could also lend me.
 
That could make a difference, but what I meant was, can you use the microphone of the headset alone, without plugging the speakers in.

Try to record something with only the microphone plugged in. No speakers. Use the microphone port.
Then once it's recorded, plug the speakers back in, and listen normally.

If this works, then we know it isn't drivers. Maybe it's the adapter.
I tested connecting only the microphone and recorded that. Then, I connected the audio part using the Y-adapter.

My observation is that in OBS, without any filters, the voice would sound good if it were louder and had much less noise.

When I applied the noise suppression filter in OBS, I didn't have any noise, but the problems with audio clipping/distortion returned.

I started to play with some filters and volume levels in the system and OBS, but I haven't found the right settings yet.

I'll update you after I meet with my friend—it may take a couple of days.
 
I visited my friend and we tested my headsets on his Windows machine using both GeForce Experience and OBS Studio. The quality was good on both, which ensured me that the headset and OBS itself are functioning correctly.

Before trying to tinker with drivers, the kernel, or any advanced options in my system, I decided to go back to the basics.
  1. Volume Control:
    • Opened Volume Control.
    • Navigated to Input Devices.
    • Built-in Audio Analog Stereo -> Port: Rear Microphone (plugged in).
    • By default, the input level was set to 100% (0 dB), which theoretically should be the most balanced setting. However, when I started talking into the microphone, the sound bar level reached the maximum. It seems that for Fedora, the 0 dB setting was too high.

1737578223033.png


I significantly decreased the input device level. This adjustment eliminated the distortion.



1737578956234.png


The issue was very simple. The input level was set too high, causing distortion. Now, the audio is a bit quieter, but I can find a proper balance. Topic can be closed - thanks to all who helped me, especially dos2unix and KGIII.
 

Members online


Top