:: Re: [maemo-leste] Nokia N900 voice …
Forside
Slet denne besked
Besvar denne besked
Skribent: Merlijn Wajer
Dato:  
Til: Pavel Machek
CC: maemo-leste, Sicelo A . Mhlongo
Emne: Re: [maemo-leste] Nokia N900 voice calls
Hi,

On 26/03/2023 21:56, Pavel Machek wrote:
> Hi!
>
>> We spoke at FOSDEM a little bit about the Droid 4, power management and N900
>> voice calls. I'm actually currently charging my Droid 4 battery to 4.3V to
>> get a better battery life out of it, too.
>
> :-). I made a tool to switch between 4.1 and 4.3V and display
> estimated charge. It is at
>
> https://github.com/pavelmachek/unicsy_demo/blob/master/bat/batman


Sounds good, I think freemangordon is also making some progress on
battery related code recently.

> . I usually let it charge to 4.1V, then tell it to boost to
> 4.3V when I anticipate the need for battery life... Likely better
> solution would be "I'd like this battery to be full at 8am".
>
>> I recalled that you offered to point us in the right direction regarding
>> N900 voice calls - I think it would be great to see if we can take another
>> look at it. Do you have some pointers that can help us get started?
>
> I believe the steps are:
>
> 1) Verify pulseaudio works on n900, in full duplex mode, and verify
> the latency.
>
> 2) Verify the voice call quality is acceptable (it was not for me).
>
>> so maybe the first step would be to package it and load it. I remember when
>> I would load it, it would take a whole bunch of CPU on start.
>
> 3) If biggest problem is CPU load, that probably can be fixed. But I
> did not know how to fix 2), and did not verify 1).


Okay, I can try to build this for chimaera as a starter.

>> Anyway, I remember you had another path working, do you have some details on
>> it?
>
> I have this: Also not sure if it is useful...
>
> https://gitlab.com/libcmtspeechdata/libcmtspeechdata


We have this packaged on Chimaera:
https://maedevu.maemo.org/pkgweb/chimaera/main/binary-armhf/libcmtspeechdata.html

I tried to run cmt_alsa and I can hear my Fremantle N900 on the n900
speaker it seems, but I could not hear my leste n900 on my fremantle
n900 - there were a lot of errors though. This could perhaps just be
related to the N900 audio setup (I was also surprised it went to
speakers rather than earpiece). [1]

Unrelated, but we probably have to make UCM2 for the N900.

With cmt_pulse nothing happened during the call - it didn't pick up on
the call happening at all, but I think we have some chat about this on
this issue from a few years ago:
https://github.com/maemo-leste/bugtracker/issues/77

Regards,
Merlijn


[1]

NFS sucks, version 0.0.1
Scheduler set to Round Robin with priority 99...
CMTSPEECH: nokiamodem_backend: Opening device, libcmtspeechdata v2.0.0.
CMTSPEECH: backend_common: CMT Speech Data state machine activated with 
SSI_CONFIG_REQ.
ERROR: CMTSPEECH: backend_common: ERROR: sending cmd 
NEW_TIMING_CONFIG_REQ failed, res -1
Hardware PCM card 0 'RX-51' device 0 subdevice 0
Its setup is:
   stream       : PLAYBACK
   access       : RW_INTERLEAVED
   format       : S16_LE
   subformat    : STD
   channels     : 2
   rate         : 8000
   exact rate   : 8000 (8000/1)
   msbits       : 16
   buffer_size  : 2048
   period_size  : 1024
   period_time  : 128000
   tstamp_mode  : NONE
   tstamp_type  : MONOTONIC
   period_step  : 1
   avail_min    : 1024
   period_event : 0
   start_threshold  : 2147483647
   stop_threshold   : 2048
   silence_threshold: 0
   silence_size : 0
   boundary     : 1073741824
   appl_ptr     : 0
   hw_ptr       : 0
3.0000 gain, 0 overflows,      0..0..0
initial write: 3072
Hardware PCM card 0 'RX-51' device 0 subdevice 0
Its setup is:
   stream       : CAPTURE
   access       : RW_INTERLEAVED
   format       : S16_LE
   subformat    : STD
   channels     : 2
   rate         : 8000
   exact rate   : 8000 (8000/1)
   msbits       : 16
   buffer_size  : 2048
   period_size  : 1024
   period_time  : 128000
   tstamp_mode  : NONE
   tstamp_type  : MONOTONIC
   period_step  : 1
   avail_min    : 1024
   period_event : 0
   start_threshold  : 2147483647
   stop_threshold   : 2048
   silence_threshold: 0
   silence_size : 0
   boundary     : 1073741824
   appl_ptr     : 0
   hw_ptr       : 0
read = 16
3.0000 gain, 0 overflows,  -9900..-9900..-9900
could not fill incoming buffer
cc write failed: Bad file descriptor
read = 16
3.0000 gain, 0 overflows, -10044..-9900..-9762
could not fill incoming buffer
cc write failed: Bad file descriptor
Could notrelease ulbuf, says (-16)
CMTSPEECH: backend_common: CMT Speech Data state machine deactivated.