Bluetooth mouse usage makes A2DP audio stutter

[Martin Emrich]

Binary package hint: bluez

When I use my Bluetooth A2DP headset (Philips SHB7102) together with my mouse (Lenovo Bluetooth Laser Mouse), I notice this:

Whenever I leave the mouse still for a few seconds (it seems to enter some kind of lower-power-mode) and then start to move it, my music most of the time skips for a second. This makes using these two bluetooth devices together quite annoying...

The adapter is a Lenovo "ThinkPad Bluetooth with Enhanced Data Rate II" (Bus 004 Device 005: ID 0a5c:2145 Broadcom Corp.)

I am running current karmic amd64 (linux 2.6.31-9).

[Matthew Woerly (nattgew)]

I'm using a Motorola S805 headset and
Bus 007 Device 003: ID 413c:8126 Dell Computer Corp. Wireless 355 Bluetooth
Same for me is Karmic with 2.6.31-11...

[Maui Muc (stefan-mauerberger)]

I can confirm this problem with a broadcom bluetooth controller also with enhanced data rate on a IBM T60 in combination with an motorola DC800 audiogateway and an noname bluetooth mouse.

From Jaunty to Karmic it has become worse.

[Matthew Woerly (nattgew)]

Changing to Confirmed, still getting this.

[Matthew Woerly (nattgew)]

Architecture: amd64
DistroRelease: Ubuntu 9.10
NonfreeKernelModules: wl
Package: bluez 4.51-0ubuntu2
PackageArchitecture: amd64
ProcEnviron:
 SHELL=/bin/bash
 PATH=(custom, no user)
 LANG=en_US.UTF-8
ProcVersionSignature: Ubuntu 2.6.31-13.45-generic
Uname: Linux 2.6.31-13-generic x86_64
UserGroups: adm admin cdrom dialout lpadmin netdev plugdev sambashare

[Dmitriy (dark81)]

Architecture: amd64
DistroRelease: Ubuntu 9.10
Package: bluez 4.51-0ubuntu2
PackageArchitecture: amd64
ProcEnviron:
 PATH=(custom, no user)
 LANGUAGE=ru_RU.UTF-8
 LANG=ru_RU.UTF-8
 SHELL=/usr/bin/zsh
ProcVersionSignature: Ubuntu 2.6.31-14.46-generic
Uname: Linux 2.6.31-14-generic x86_64
UserGroups: vboxusers

[Dmitriy (dark81)]

It is impossible to use Plantronics P590 at same time when Logitech M555b mouse in use. Sound shutdown on every mouse movement.
When I use only headset it work very good. Under Windows XP headset and mouse works perfectly together.

[Matthew Woerly (nattgew)]

If the audio destination is closer compared to the mouse it's much less of a problem... but once the audio destination is further, like 10 feet, the mouse really messes with it.

[Timur I. Davletshin (timur-davletshin)]

Almost the same problem for me. But my Motorola H350 headset is "stronger" than my mouse and very often kills other connections (mouse and phone). And then it doesn't sound just stutters.

[Paul Illes (illespal)]

Same problem with sony dr-21btg a gigabyte bluetooth mouse and 2.1 bluetooth notebook, and 2.0 notebook. Both skips while using them at the same time or just even ideling the mouse. (kernel 2.6.32, 4.60/4.61 bluez)

Any updates?

[Martin Emrich (emme)]

I just tried this under Windows Vista, same hardware. Here, there's no problem, the music plays fine while using the mouse. So it seems not to be a hardware or gereral Bluetooth protocol problem.

[Baptiste Mille-Mathias (bmillemathias)]

On Wed, Feb 17, 2010 at 10:54 AM, Martin Emrich <email address hidden> wrote:
> I just tried this under Windows Vista, same hardware. Here, there's no
> problem, the music plays fine while using the mouse. So it seems not to
> be a hardware or gereral Bluetooth protocol problem.
>

Perhaps you can ask on freenode IRC server on chan #bluez-users.

--
Baptiste Mille-Mathias
Les gens heureux ne sont pas pressés

[Paul Illes (illespal)]

Baptiste, i've already sent a mail to the linux-bluetooth mailing-list regarding that (after waiting an hour on the irc channel you mentioned). Now waiting for reply there on the mailing list. Link: http://marc.info/?l=linux-bluetooth&m=126633648012348&w=2

[Martin Emrich (emme)]

I just tried it on my old laptop (Thinkpad T41p running lucid i386 updated a few minutes ago), with the same headset and the same mouse. On lucid, it seems to work (not a single audio interruption for 5 minutes now).

I don't think there will be a fix released for 9.10, so let's wait for 10.04 :)

[Paul Illes (illespal)]

lucid uses bluez 4.60, 2.6.32. I'm using 4.60 too on my arch linux system. There I do have the skips, even having these versions...tho it can play for some minutes but after a while it gets bad. This is with BT 2.0 profile. On my 2.1 notebook it's much better, very few skips and little pitch shifting.

[Baptiste Mille-Mathias (bmillemathias)]

On Wed, Feb 17, 2010 at 11:37 PM, Martin Emrich <email address hidden> wrote:
> I just tried it on my old laptop (Thinkpad T41p running lucid i386
> updated a few minutes ago), with the same headset and the same mouse. On
> lucid, it seems to work (not a single audio interruption for 5 minutes
> now).
>
> I don't think there will be a fix released for 9.10, so let's wait for
> 10.04 :)
>

No there is too much to backports, so if it works for lucid, I guess I
can close the bug.

Thanks.
--
Baptiste Mille-Mathias
Les gens heureux ne sont pas pressés

[Maui Muc (stefan-mauerberger)]

I've tried the beta 2 a view days ago and have to report that the bug still exists in Lucid. This is so annoying because on Windows XP the audio gateway and the mouse together works properly.

[Vladimir Firenze (turino-lgua)]

have the same problem in 10.04 with 2.6.32 & bluez 4.60.

Having Trust 15352 mouse and Motorola S305 headset

[Paul Illes (illespal)]

2.6.34 kernel and latest bluez changes nothing either. It's bluez's problem with hifi audio and mouse usage at the same time. Looks like parallel open channels are still handled badly. (I'm following bluez's git commits to the source base, i couldnt spot anything changed related to this problem, so my guess it's gonna remain like this for another year or so... :( )

[Jens Janssen (jayjay)]

The problem is not bluez but the settings bluez uses. You can use hciconfig to put your device into the correct mode.

hciconfig hci0 lm ACCEPT,MASTER
hciconfig hci0 lp HOLD,SNIFF,PARK

should be enough where hci0 is your bluetooth device.

I put this in a udev rule so that these changes are applied every startup.

For that you can copy following text:

# put this file in /etc/udev/rules.d
# choose a higher rule number than the existing file at /lib/udev/rules.d/XX-bluetooth.rules
ACTION=="add", SUBSYSTEM=="bluetooth", KERNELS=="hci0", RUN+="/usr/sbin/hciconfig hci0 lm ACCEPT,MASTER", RUN+="/usr/sbin/hciconfig hci0 lp HOLD,SNIFF,PARK"

You can name the file 100-bluetooth.rules. Hope this helps.

[Paul Illes (illespal)]

This might be useful for some admittedly.

BUT that doenst really help the problem underlying in bluez which causes the problem with using mouse and A2DP parallelly.

And yes, I've tried a zillion configuration versions with hciconfig and none helped.
It's definitely bluez problem.

[Jens Janssen (jayjay)]

I use a bluetooth mouse and A2DP in parallel with this configuration without problems. This is the only configuration I know that makes bluetooth devices working in parallel.

[Matthew Woerly (nattgew)]

Thanks for the tip about the hciconfig settings.

hciconfig hci0 lm ACCEPT,MASTER
hciconfig hci0 lp HOLD,SNIFF,PARK

Running those makes things much better. I can actually use the Bluetooth mouse and play Bluetooth audio at the same time without totally losing the audio. I'm going to try the udev rules and hopefully that keeps things much better.

[Martin Emrich (emme)]

The "hciconfig hci0 lm ACCEPT,MASTER ; hciconfig hci0 lp HOLD,SNIFF,PARK" trick does _not_ work here. Audio still drops out when I move the mouse.

[Paul Illes (illespal)]

Yeah, indeed, the problem lies deeper in the code of bluez stack. I've tried all the hciconfig options and it does not help a bit.

(Interesting fact is that the troubles with bluez are underlined by the sudden pitch changes and skips happening on android based phones with bluetooth, which OS uses BLUEZ stack as well. I hope this fact may trigger something in the head of bluez devs as android users are more and more around the world, and i think a good chunk of them will swear if its revealed to them that bluetooth is crappy on the phone)

[Allan Branch (megamanexent7)]

I am get choppy and garbled sound and I am not even using a Bluetooth mouse ( or any other bluetooth device at the same time as A2DP stream). A2DP sound tends to pitch shift constantly, gets worse when the source ( my phone ) is far away from the computer. The hciconfig options help, but it did not solve it. Keep in mind I am streaming sound from my phone to the computer ( the opposite of what you guys are doing ). My first instinct was that pulseaudio was causing most of the trouble, but after reading this bug report, and Paul's mentioning that Android phone are experiencing the same problem, it must be the bluez stack, unless Android runs pulseaudio too.

Summary: Streaming Audio to computer still leads to pitch shifting, dropped sound, even without any other Bluetooth device connected.

[Charles A. Templeton III (ctempleton3)]

I am having the exact same problem. I tried the hciconfig commands and it didn't work. My mouse and headset work perfectly with the exact same hardware under windows 7.

Linux 2.6.32-24-generic #42-Ubuntu SMP Fri Aug 20 14:24:04 UTC 2010 i686 GNU/Linux

[Jens Janssen (jayjay)]

I hope you put "sudo" in front of the commands.
These commands need to be executed before connecting the devices.

If that does not help, try the udev rule.

[Charles A. Templeton III (ctempleton3)]

Ok, I got it working. I d/c the devices and then ran the commands and reconnected the devices.

[Jens Janssen (jayjay)]

Thank you for the feedback.

Sometimes it is also helps to make your own computer visible before connecting to a new/unknown bluetooth device.

[Andrei Emeltchenko (andrei-emeltchenko-finik)]

Can somebody check that changing bitpool in bluetooth daemon helps? Like (NOTE: dirty hack):

bluez repository:
git://git.kernel.org/pub/scm/bluetooth/bluez.git

diff --git a/audio/unix.c b/audio/unix.c
index 62eee31..e1a1c5a 100644
--- a/audio/unix.c
+++ b/audio/unix.c
@@ -557,7 +557,7 @@ static int a2dp_append_codec(struct bt_get_capabilities_rsp *
                sbc->subbands = sbc_cap->subbands;
                sbc->block_length = sbc_cap->block_length;
                sbc->min_bitpool = sbc_cap->min_bitpool;
- sbc->max_bitpool = sbc_cap->max_bitpool;
+ sbc->max_bitpool = sbc_cap->max_bitpool - 20;

                print_sbc(sbc_cap);
        } else if (codec_cap->media_codec_type == A2DP_CODEC_MPEG12) {

[Brian Rogers (brian-rogers)]

That seems to help a lot. In 30 minutes of testing, I had zero audio stutters, even if I frantically wiggled the mouse.

[robled (robled)]

I can reproduce this consistently on Ubuntu 10.10 x64. Why is the status of this bug set to "Fix Released"?

[Matthew Woerly (nattgew)]

Because someone 10 months ago said it was fine in 10.04... I agree, I haven't noticed much improvement in Bluetooth performance.

[Brian Rogers (brian-rogers)]

He later reported experiencing the bug again, so I'll reopen this.

[Brian Rogers (brian-rogers)]

OK, I can't actually change the bug status here.

[Afro (amilan-inlogic)]

No progress? I reported same problem on Sony Vaio CPCEA1SE. BT mouse + BT audio (a2dp) doesn't work together correctly. audio drop off stream. Ubuntu 2.6.35-25 x86_64

[Matthew Woerly (nattgew)]

I tried Ubuntu on my Cr-48, and it seems to be the other way around... my Logitech Bluetooth mouse seems to lag while the audio continues just fine.
Atheros AR5BBU12 Bluetooth. Ubuntu 10.10.

[Ian Halpern (ihh)]

The fix in comment #26 (https://bugs.launchpad.net/ubuntu/+source/bluez/+bug/424215/comments/26) used to work for me until I installed the latest updates and restarted last night...now the audio is once again choppy. I'm on 11.04. Any suggestions?

[Matt Hegarty (mrhegarty)]

Ubuntu 11.04
2.6.35-32-generic
bluez 4.69-0ubuntu

I'm trying to stream audio to hi-fi from Ubuntu. Experienced very choppy audio with a Belkin L25 bluetooth dongle, sending playback to a Belkin Music Receiver. This was in the absence of any mouse movements.

Tried the following:

* Disconnected device
* Added the fix in https://bugs.launchpad.net/ubuntu/+source/bluez/+bug/424215/comments/26 (with sudo)
* Reconnected device
This worked for about 60 seconds and then became choppy again.

Making the PC 'visible' vai bluetooth seemed to improve things

[Mathieu Trudel-Lapierre (cyphermox)]

Matt: please file a new bug report for your hardware. It's also possible that this is hardware-related, since different bluetooth chips may have different performance.

[Sebastian Malcolm (smalcolm)]

This bug still exists in Ubuntu 12.04 LTS and (re)reported in Red Hat Bugzilla's as bug 749366, so I believe we need a new bug report to get a fix for this into a 12.04.x point release?

If Andrei's patch to increase bitpool in bluetooth daemon fixes this problem then the problem _is_ in bluez and changing the settings bluez uses via options to hciconfig (that aren't documented in it's man page) is only a workaround.

I reproduced this choppy audio playback during mouse movements with the combination of:
* Kernel: 3.2.0-24-generic #39-Ubuntu SMP Mon May 21 16:52:17 UTC 2012 x86_64
* Bluez: 4.98-2ubuntu7
* Jabra BT620s
* Logitech Bluetooth Mouse M555b
* Onboard Bluetooth in a VAIO-Z2 (connected internally via USB): ID 8086:0189 Intel Corp.

Disconnecting my devices, then following Jen's instructions in comment #23 and re-pairing my devices worked for me as a workaround for this issue in bluez's link management. I will try using the uDev rule to make this workaround permanent across reboots. I've since streamed DI.fm without stutter for several hours whilst using my M555b mouse at same distance from the laptop as my Jabra bluetooth headset.

After applying the hciconfig changes, and reconnecting both headset & mouse (as only devices connected) my "hciconfig -a" output is:

hci0: Type: BR/EDR Bus: USB
        BD Address: 88:53:2E:29:C7:27 ACL MTU: 310:10 SCO MTU: 64:8
        UP RUNNING PSCAN ISCAN
        RX bytes:10740963 acl:308181 sco:0 events:784453 errors:0
        TX bytes:226808075 acl:779584 sco:0 commands:100 errors:0
        Features: 0xff 0xff 0x8f 0xfe 0x9b 0xff 0x59 0x87
        Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
        Link policy: HOLD SNIFF PARK
        Link mode: ACCEPT MASTER
        Name: 'z2-0'
        Class: 0x7e0100
        Service Classes: Networking, Rendering, Capturing, Object Transfer, Audio, Telephony
        Device Class: Computer, Uncategorized
        HCI Version: 3.0 (0x5) Revision: 0x1b07
        LMP Version: 3.0 (0x5) Subversion: 0x0
        Manufacturer: Intel Corp. (2)

[Nick B. (futurepilot)]

Confirmed on 12.04. The workaround in comment 23 worked for me after I disconnected all bluetooth devices and then ran the commands.

hci0: Type: BR/EDR Bus: USB
        BD Address: 00:02:72:DA:07:50 ACL MTU: 1021:8 SCO MTU: 64:1
        UP RUNNING PSCAN ISCAN
        RX bytes:1349414 acl:39653 sco:0 events:95733 errors:0
        TX bytes:116182568 acl:188689 sco:0 commands:57 errors:0
        Features: 0xbf 0xfe 0xcf 0xfe 0xdb 0xff 0x7b 0x87
        Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
        Link policy: HOLD SNIFF PARK
        Link mode: ACCEPT MASTER
        Name: 'xxxxxx'
        Class: 0x7e0100
        Service Classes: Networking, Rendering, Capturing, Object Transfer, Audio, Telephony
        Device Class: Computer, Uncategorized
        HCI Version: 4.0 (0x6) Revision: 0x1000
        LMP Version: 4.0 (0x6) Subversion: 0x220e
        Manufacturer: Broadcom Corporation (15)

Bus 002 Device 030: ID 0a5c:21e8 Broadcom Corp.

[Charles Edward Pax (pax)]

This appears to still be an issue in 13.04 daily (201302120. I'm running a Macbook Air1,1 with an Apple BT mouse (A1296) and Samsung BT headphones (SBH600).

[Pratik (pratik-prajapati)]

I am also seeing the same issue as mentioned by Matt Hegarty in comment #43 (https://bugs.launchpad.net/ubuntu/+source/bluez/+bug/424215/comments/43)

I have ubuntu 12.04, kernel 3.8.0-37-generic, bluez 4.98-2ubuntu7,

my BT mouse is : Gigabyte
and BT headset is: motorola SF600.

With fix mentioned in https://bugs.launchpad.net/ubuntu/+source/bluez/+bug/424215/comments/43,
It works for 10-15 seconds and then when i move mouse I get stutter audio.

Any possible fix?

[Uldis Kalniņš (ulcha)]

Confirmed on 14.10 alpha

Fix in #23 still working.

hci0: Type: BR/EDR Bus: USB
 BD Address: BC:77:37:D5:9F:89 ACL MTU: 310:10 SCO MTU: 64:8
 UP RUNNING PSCAN
 RX bytes:1364783 acl:46087 sco:0 events:89396 errors:0
 TX bytes:29127367 acl:132937 sco:0 commands:229 errors:0
 Features: 0xff 0xff 0x8f 0xfe 0x9b 0xff 0x59 0x87
 Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
 Link policy: HOLD SNIFF PARK
 Link mode: ACCEPT MASTER
 Name: 'Dell-System-Vostro-3450-0'
 Class: 0x7c0100
 Service Classes: Rendering, Capturing, Object Transfer, Audio, Telephony
 Device Class: Computer, Uncategorized
 HCI Version: 3.0 (0x5) Revision: 0x1b07
 LMP Version: 3.0 (0x5) Subversion: 0x0
 Manufacturer: Intel Corp. (2)

a2dp audio device - Motorola HT820
mouse - Apple A1197

[Uldis Kalniņš (ulcha)]

Bug is not fixed. I still have to apply #24 fix.

[mrvanes (mrvanes)]

Thanx to the fix in #23 I can now listen to uninterrupted A2DP audio from my laptop and use (or better, -not- use) the mouse without causing audio stutter. I'm running Kubuntu Wily (Frameworks 5.15.0, Applications 15.08.2, Plasma 5.4.2, Bluez 5.35).

[Ralph de Boom (t-ralph)]

Probably not much of use here but, heck sue me:

Can confirm #23 fixes stuff on Arch Linux too.

[Paolo Melchiorre (paulox)]

I can confirm the bug on Ubuntu 17.10

[Daniel van Vugt (vanvugt)]

This bug is closed, rightly or wrongly. And has been closed for several years.

If you have any ongoing issues then please log a new bug by running this command:

  ubuntu-bug bluez

[Vadim Andreyev (kolobast)]

Bug still exists in 18.10. How can you mark it fixed???

[Daniel van Vugt (vanvugt)]

I didn't. Someone else did 9 years ago in comment #19.

Also, the original bug was logged against Ubuntu 9.10 which reached end of life in 2011.

The most recent report of this bug I can see is 17.10, which also reached end of life in 2018.

If you have any ongoing problems then please open a new bug by running:

  ubuntu-bug bluez

[Oscar Carballal (oscar-carballal)]

I'm surprised I still get this issue. IT started happening since a couple weeks ago and the fix in #23 does not fix it.

First, specs:

Arch Linux
Kernel: 5.18.7-arch1
Mouse: MX Master 3
Headphones: Sony WF-1000XM4
Bluez / BluezUtils / Plugins: 5.64-2
HCIConfig: 5.63-1

When the mouse and the headphones are active, the headphones suffer extreme stuttering and buffering issues. As soon as the mouse is disconnected the headphones work perfectly fine.

I tries the fix in #23 with the following results:

Pre-hciconfig command output:
hci0: Type: Primary Bus: USB
 BD Address: 50:EB:71:FA:3C:2C ACL MTU: 1021:4 SCO MTU: 96:6
 UP RUNNING PSCAN
 RX bytes:13533405 acl:640925 sco:0 events:275306 errors:0
 TX bytes:183295302 acl:268774 sco:0 commands:5240 errors:0
 Features: 0xbf 0xfe 0x0f 0xfe 0xdb 0xff 0x7b 0x87
 Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
 Link policy: RSWITCH SNIFF
 Link mode: ACCEPT CENTRAL
 Name: 'impulse'
 Class: 0x7c0104
 Service Classes: Rendering, Capturing, Object Transfer, Audio, Telephony
 Device Class: Computer, Desktop workstation
 HCI Version: 5.2 (0xb) Revision: 0x2076
 LMP Version: 5.2 (0xb) Subversion: 0x2076
 Manufacturer: Intel Corp. (2)

I disconnect both devices and proceed to run the command "sudo hciconfig hci0 lm ACCEPT,MASTER"
Result (no changes):

sudo hciconfig -a

hci0: Type: Primary Bus: USB
 BD Address: 50:EB:71:FA:3C:2C ACL MTU: 1021:4 SCO MTU: 96:6
 UP RUNNING PSCAN
 RX bytes:13533405 acl:640925 sco:0 events:275306 errors:0
 TX bytes:183295302 acl:268774 sco:0 commands:5240 errors:0
 Features: 0xbf 0xfe 0x0f 0xfe 0xdb 0xff 0x7b 0x87
 Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
 Link policy: RSWITCH SNIFF
 Link mode: ACCEPT CENTRAL
 Name: 'impulse'
 Class: 0x7c0104
 Service Classes: Rendering, Capturing, Object Transfer, Audio, Telephony
 Device Class: Computer, Desktop workstation
 HCI Version: 5.2 (0xb) Revision: 0x2076
 LMP Version: 5.2 (0xb) Subversion: 0x2076
 Manufacturer: Intel Corp. (2)

Then I try to run the next command (sudo hciconfig hci0 lp HOLD,SNIFF,PARK) with the following result:

    Can't set link policy on hci0: Operation not supported (95)

Any idea what can I do about this? I mean, the bug is pretty old and it has been fixed multiple times but it seems to be a reocurring issue every now and then... I know we can't reopen this ticket, but this issue should be tracked somewhere IMHO

[Daniel van Vugt (vanvugt)]

This bug is closed. If you have any ongoing problems then please open a new bug by running:

  ubuntu-bug bluez

[Mårten Thornberg (marrowm)]

@Oscar Carballal

Try changing codec in pavucontrol (under the configuration tab).

---
I have opened a new bug here:
https://bugs.launchpad.net/ubuntu/+source/pulseaudio/+bug/1992306

...and upstream:
https://gitlab.freedesktop.org/pulseaudio/pulseaudio/-/issues/1400