TGL (Tiger Lake) ALC257 ABI mismatch: firmware 3:22:1 vs kernel 6.17 ABI 3:23:1 - rain noise on DMIC

[10gajendra]

System Information:

Machine: Lenovo IdeaPad Flex 5 14ITL05 (LNVNB161216)
Subsystem ID: 17aa38a3
Codec: Realtek ALC257 (10ec0257)
DSP: Intel Tiger Lake (TGL)
OS: Ubuntu 24.04.4 LTS (Noble Numbat)
Kernel: 6.17.0-19-generic (#19~24.04.2-Ubuntu)
firmware-sof-signed: 2025.12.2-1
PipeWire: 1.0.5
WirePlumber: 0.4.17

Problem Description:
Internal DMIC microphone produces only white noise (rain-like sound) when recording. The issue is a SOF firmware ABI mismatch between the installed firmware (3:22:1) and what kernel 6.17 expects (3:23:1).
Kernel log shows:
sof-audio-pci-intel-tgl: Firmware: ABI 3:22:1 Kernel ABI 3:23:1
sof-audio-pci-intel-tgl: Topology: ABI 3:22:1 Kernel ABI 3:23:1
sof-audio-pci-intel-tgl: DMICs detected in NHLT tables: 2
sof-audio-pci-intel-tgl: using HDA machine driver skl_hda_dsp_generic
With IPC4 mode (snd_sof.ipc_type=1):
sof-audio-pci-intel-tgl: Topology: ABI 3:29:1 Kernel ABI 3:23:1
DMIC is visible but produces no audio output.
Expected behavior:
Internal DMIC should record audio cleanly without noise.
Actual behavior:

IPC3 mode: DMIC records only white noise/rain sound
IPC4 mode: DMIC is visible but silent

Workaround:
Currently using Bluetooth headset as temporary microphone workaround.

[lgirdwood]

@10gajendra its probably not ABI as its backwards compatible, do you know how many DMICs you have (as this can impact config and is not always known by driver) ?
Btw, with IPC4 mode do you capture silence i.e. 0s for the duration of capture.
@singalsu any inputs ?

[singalsu]

@10gajendra Please attach here the report generated by alsa-info.

[10gajendra]

@10gajendra its probably not ABI as its backwards compatible, do you know how many DMICs you have (as this can impact config and is not always known by driver) ? Btw, with IPC4 mode do you capture silence i.e. 0s for the duration of capture. @singalsu any inputs ?

I have 2 DMICs (as detected by NHLT tables).
With IPC4 mode, the capture device is visible but records silence (0s throughout the recording), not actual audio.

[10gajendra]

@10gajendra Please attach here the report generated by alsa-info.

Here is my ALSA info report:
http://alsa-project.org/db/?f=ab39eba6a6706d5ff898e67125df85ac82e1982e

Let me know if any additional logs are needed.

[singalsu]

@10gajendra Please attach here the report generated by alsa-info.

Here is my ALSA info report: http://alsa-project.org/db/?f=ab39eba6a6706d5ff898e67125df85ac82e1982e

Let me know if any additional logs are needed.

control.41 {
	iface MIXER
	name 'Dmic0 Capture Volume'
	value.0 70
	value.1 70
	comment {
		access 'read write'
		type INTEGER
		count 2
		range '0 - 70'
		dbmin -9999999
		dbmax 2000
		dbvalue.0 2000
		dbvalue.1 2000
	}
}
control.42 {
	iface MIXER
	name 'Dmic0 Capture Switch'
	value.0 true
	value.1 true
	comment {
		access 'read write'
		type BOOLEAN
		count 2
	}
}

It looks like you have max. volume and capture switches are enabled so it looks OK. I can't see from alsa-info which topology you are using, but I'm guessing it is this IPC3 sof-hda-generic-2ch.tplg. Can you confirm it from dmesg or journalctl -k.

One thing to check is the EQIIR10.0 in capture pipeline from DMIC0.IN. Do you have sof-ctl tool available?

Find the full control name of the IIR with command "amixer -c0 controls". Then do "sof-ctl -c name='x' where x is the full control name. I can check from the words values if the config blob that the IIR is using is sane. If the IIR and the gain are right then it is a lower level issue than firmware.

Are the by chance some mic mute or privacy switches in the PC?

[singalsu]

Ah,I saw now that you've tried both IPC3 and IPC4. It's really strange that you get rain with IPC3 and silence with IPC4. With IPC4 the topology would look like this:

With IPC4 would be useful to see control values for gain (alsa-info), eqiir, tdfb, drc (sof-ctl).

But that possible mute/privacy switch is best to check first. And do the check with low-level arecord command line instead of pulse/pipewire, e.g. "arecord -Dhw:0,6 -f S32_LE -c 2 -r 48000 -d 10 rec.wav".

[10gajendra]

Ah,I saw now that you've tried both IPC3 and IPC4. It's really strange that you get rain with IPC3 and silence with IPC4. With IPC4 the topology would look like this:

With IPC4 would be useful to see control values for gain (alsa-info), eqiir, tdfb, drc (sof-ctl).

But that possible mute/privacy switch is best to check first. And do the check with low-level arecord command line instead of pulse/pipewire, e.g. "arecord -Dhw:0,6 -f S32_LE -c 2 -r 48000 -d 10 rec.wav".

DMIC Capture Issue on Lenovo IdeaPad Flex 5 (Tiger Lake)

System Info:

OS: Ubuntu 24.04.4 LTS, Kernel 6.17.0-19-generic
Codec: Realtek ALC257, DSP: Intel Tiger Lake
Firmware: firmware-sof-signed 2025.12.2-1
PipeWire: 1.0.5 / WirePlumber: 0.4.17

Topology:

sudo dmesg | grep -i tplg

Output:

sof-audio-pci-intel-tgl 0000:00:1f.3: Topology file: intel/sof-tplg/sof-hda-generic-2ch.tplg
2 DMICs detected

Mixer / Capture Controls:

Dmic0 Capture Volume: 70 / 70 (max)
Dmic0 Capture Switch: on

Raw DMIC Capture Test (IPC4):

arecord -Dhw:0,6 -f S32_LE -c 2 -r 48000 -d 10 rec.raw
head -c 200 rec.raw | hd

Output (all zeros):

00000030 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|
*
000000c0 00 00 00 00 00 00 00 00 |........|

Observations:

DMIC detected and enabled
IPC4 capture visible but records silence
IPC3 capture produces white noise
Mixer settings correct (max volume, capture on)
sof-ctl could not be built on Ubuntu 24.04 due to xtensa toolchain

Conclusion: Likely a low-level DSP/IPC4 routing or firmware issue, not a volume/mixer problem.

[github-actions]

This issue has been marked as stale because it has been open (more than) 60 days with no activity. Remove the stale label or add a comment saying that you would like to have the label removed otherwise this issue will automatically be closed in 14 days. Note, that you can always re-open a closed issue at any time.