struct snd_pcm_hardware_t values

struct snd_pcm_hardware_t values

[Markus Rechberger]

Hi,

a half year ago I bought an au600 device
(http://alsa.opensrc.org/au600) and it's still not supported... I made
some progress uppon the driver but there are still a few issues left.

does anyone know about the snd_pcm_hardware_t struct?
I'm not sure what values I should fill in there...
My current driver supports a choppy kind of playback and capturing choppy audio.
If I use following values in the struct my box locks up when the audio
stream is over (uninitialized buffer for sure)
static snd_pcm_hardware_t snd_au600_hw = {
    .info = SNDRV_PCM_INFO_BLOCK_TRANSFER | SNDRV_PCM_INFO_MMAP |
SNDRV_PCM_INFO_INTERLEAVED | SNDRV_PCM_INFO_MMAP_VALID,
    .formats = SNDRV_PCM_FMTBIT_S16_LE,
    .rates = SNDRV_PCM_RATE_KNOT | SNDRV_PCM_RATE_CONTINUOUS,
    .rate_min = 9600,
    .rate_max = 9600,
    .channels_min = 1,
    .channels_max = 1,
    .buffer_bytes_max = (32*1024),
    .period_bytes_min = 64,
    .period_bytes_max = 2048,
    .periods_min = 1,
    .periods_max = 255,
};

If I set the values lower it doesn't crash anymore but afterwards the
stream sounds garbled.

The streamformat itself is PCM, 9600hz, 16Bit, mono, S16_LE (capturing
as well as playback)

Markus

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

Re: struct snd_pcm_hardware_t values

[Markus Rechberger]

I did some more investigation uppon this
>     .buffer_bytes_max = (32*1024),
this one seems to be clear now, I set it to the max
transfer_buffer_size of 1 URB
>     .period_bytes_min = 64,
this one not .. in my case I set it to transfer_buffer_size/2
>     .period_bytes_max = 2048,
this one is again transfer_buffer
>     .periods_min = 1,
I set this one to 2 (people in IRC cleared me up about this one)
>     .periods_max = 255,
no idea again..

So what I get now is:
http://lost-souls.mercuryserver.de/files/data/test.wav.mp3

and a buffer underrun regarding aplay
aplay -D hw:2,1 5_5_2006__07_00_59.wav
Playing WAVE '5_5_2006__07_00_59.wav' : Signed 16 bit Little Endian,
Rate 9600 Hz, Mono
underrun!!! (at least 0,307 ms long)
underrun!!! (at least 0,285 ms long)
underrun!!! (at least 0,295 ms long)
underrun!!! (at least 0,279 ms long)
underrun!!! (at least 0,298 ms long)
underrun!!! (at least 0,287 ms long)

the strange thing here is the DTMF tones in the background sound fine,
so I don't really recognize an underrun when playing the file just
that strange noise which is constant all the time..

following is what my debug prints in dmesg:
I'm sure the PCM PLAYBACK TRIGGER STOP! are caused by the buffer
underrun which is reported by aplay..
The outgoing dataformat is 9600hz, S16_LE, mono ..

memcpy3 384 - 0--
ed ff ee ff ed ff f3 ff f2 ff f5 ff fa ff fa ff ff ff ff ff
00 00 01 00 03 00 06 00 05 00 0b 00 11 00 13 00 11 00 13 00
14 00 14 00 18 00 18 00 1c 00 1e 00 1e 00 22 00 20 00 21 00
27 00 25 00 25 00 2b 00 2c 00 2d 00 2d 00 2f 00 34 00 32 00
30 00 34 00 38 00 35 00 37 00 3a 00 39 00 38 00 3b 00 3c 00
37 00 3e 00 3e 00 3f 00 42 00 3f 00 41 00 3f 00 45 00 49 00
46 00 48 00 46 00 46 00 49 00 4c 00 4d 00 4b 00 4d 00 4b 00
4d 00 4e 00 4b 00 4d 00 4b 00 4f 00 53 00 4f 00 4b 00 4b 00
51 00 51 00 50 00 53 00 51 00 55 00 54 00 52 00 56 00 53 00
55 00 57 00 52 00 55 00 55 00 55 00 53 00 52 00 55 00 54 00
57 00 56 00 53 00 56 00 55 00 57 00 55 00 54 00 55 00 56 00
56 00 54 00 55 00 52 00 54 00 54 00 52 00 53 00 53 00 53 00
53 00 53 00 52 00 52 00 52 00 51 00 51 00 4f 00 4d 00 50 00
4f 00 4e 00 4f 00 51 00 50 00 4f 00 50 00 4d 00 4c 00 4e 00
4c 00 4c 00 4d 00 47 00 48 00 4c 00 49 00 4b 00 46 00 46 00
49 00 42 00 45 00 48 00 46 00 44 00 46 00 45 00 42 00 45 00
43 00 3e 00 3f 00 40 00 3d 00 3d 00 3d 00 3b 00 3e 00 40 00
39 00 3a 00 3b 00 35 00 36 00 36 00 36 00 33 00 35 00 37 00
34 00 30 00 2e 00 33 00 2e 00 2f 00 30 00 2d 00 2e 00 29 00
2b 00 2a 00
PCM PLAYBACK TRIGGER STOP!
--- PERIOD ELAPSED 0 1676736 96 --
snd_au600_prepare
PCM PLAYBACK TRIGGER START!
playback hwptr: 0 transfer_done: 1676928  192 psize: 96 stride: 2
hwptr too big: 192 - 192
memcpy3 384 - 0--
29 00 2a 00 29 00 28 00 27 00 25 00 22 00 26 00 24 00 21 00
21 00 1d 00 1d 00 1d 00 1e 00 1e 00 1d 00 1c 00 1a 00 1d 00
1c 00 18 00 19 00 13 00 15 00 16 00 10 00 13 00 13 00 11 00
10 00 0f 00 0c 00 0c 00 0b 00 09 00 0a 00 0b 00 0b 00 08 00
07 00 09 00 07 00 04 00 02 00 01 00 00 00 fe ff fb ff fb ff
fd ff fc ff fb ff f8 ff f8 ff fa ff f5 ff f3 ff f3 ff f4 ff
f4 ff f0 ff f1 ff f1 ff ee ff ee ff eb ff eb ff ec ff e7 ff
e8 ff e5 ff e3 ff e4 ff e4 ff e3 ff e1 ff e4 ff df ff dd ff
de ff da ff dc ff dc ff d9 ff d8 ff d8 ff d5 ff d3 ff d4 ff
cd ff d4 ff d3 ff cb ff d0 ff cc ff cc ff cc ff ca ff cc ff
c5 ff c6 ff c9 ff c5 ff c2 ff c1 ff c0 ff c3 ff c1 ff bf ff
bf ff bf ff be ff bc ff bb ff b9 ff b9 ff b6 ff b5 ff b7 ff
b3 ff b2 ff b2 ff af ff b0 ff b1 ff ac ff af ff ab ff a6 ff
ab ff a6 ff a7 ff a9 ff a6 ff a4 ff a1 ff a3 ff ac ff 04 01
f3 08 02 19 6b 22 d5 19 91 06 ab f3 83 ec 13 f3 cc fe 05 06
2b 02 d6 f7 e6 f0 8e f5 25 05 aa 15 11 1b 62 0f 69 f8 8b e3
df dd 4d ea 23 01 4f 14 b3 18 29 0e d1 fd db f3 cc f5 72 ff
b0 06 cb 03 c8 f7 14 ec 37 eb ba f8 18 0e 9a 1d 88 1c 6c 0a
b3 f1 5a e1 5d e2 7f f2 97 06 e3 11 f9 0e 13 03 0c f9 02 f9
09 02 19 0b
PCM PLAYBACK TRIGGER STOP!
--- PERIOD ELAPSED 0 1676928 96 --
snd_au600_prepare
PCM PLAYBACK TRIGGER START!
playback hwptr: 0 transfer_done: 1677120  192 psize: 96 stride: 2
hwptr too big: 192 - 192
memcpy3 384 - 0--
etc.

Any help that could lead me further is appreciated...

thanks,
Markus

On 8/5/06, Markus Rechberger <mrechberger@gmail.com> wrote:
> Hi,
>
> a half year ago I bought an au600 device
> (http://alsa.opensrc.org/au600) and it's still not supported... I made
> some progress uppon the driver but there are still a few issues left.
>
> does anyone know about the snd_pcm_hardware_t struct?
> I'm not sure what values I should fill in there...
> My current driver supports a choppy kind of playback and capturing choppy
> audio.
> If I use following values in the struct my box locks up when the audio
> stream is over (uninitialized buffer for sure)
> static snd_pcm_hardware_t snd_au600_hw = {
>     .info = SNDRV_PCM_INFO_BLOCK_TRANSFER | SNDRV_PCM_INFO_MMAP |
> SNDRV_PCM_INFO_INTERLEAVED | SNDRV_PCM_INFO_MMAP_VALID,
>     .formats = SNDRV_PCM_FMTBIT_S16_LE,
>     .rates = SNDRV_PCM_RATE_KNOT | SNDRV_PCM_RATE_CONTINUOUS,
>     .rate_min = 9600,
>     .rate_max = 9600,
>     .channels_min = 1,
>     .channels_max = 1,
>     .buffer_bytes_max = (32*1024),
>     .period_bytes_min = 64,
>     .period_bytes_max = 2048,
>     .periods_min = 1,
>     .periods_max = 255,
> };
>
> If I set the values lower it doesn't crash anymore but afterwards the
> stream sounds garbled.
>
> The streamformat itself is PCM, 9600hz, 16Bit, mono, S16_LE (capturing
> as well as playback)
>
> Markus
>

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

Re: struct snd_pcm_hardware_t values

[Ben Dooks]

On Sun, Aug 06, 2006 at 08:38:53PM +0200, Markus Rechberger wrote:
> I did some more investigation uppon this
> >     .buffer_bytes_max = (32*1024),
> this one seems to be clear now, I set it to the max
> transfer_buffer_size of 1 URB
> >     .period_bytes_min = 64,
> this one not .. in my case I set it to transfer_buffer_size/2
> >     .period_bytes_max = 2048,
> this one is again transfer_buffer
> >     .periods_min = 1,
> I set this one to 2 (people in IRC cleared me up about this one)
> >     .periods_max = 255,
> no idea again..
> 
> So what I get now is:
> http://lost-souls.mercuryserver.de/files/data/test.wav.mp3
> 
> and a buffer underrun regarding aplay
> aplay -D hw:2,1 5_5_2006__07_00_59.wav
> Playing WAVE '5_5_2006__07_00_59.wav' : Signed 16 bit Little Endian,
> Rate 9600 Hz, Mono
> underrun!!! (at least 0,307 ms long)
> underrun!!! (at least 0,285 ms long)
> underrun!!! (at least 0,295 ms long)
> underrun!!! (at least 0,279 ms long)
> underrun!!! (at least 0,298 ms long)
> underrun!!! (at least 0,287 ms long)
> 
> the strange thing here is the DTMF tones in the background sound fine,
> so I don't really recognize an underrun when playing the file just
> that strange noise which is constant all the time..
> 
> following is what my debug prints in dmesg:
> I'm sure the PCM PLAYBACK TRIGGER STOP! are caused by the buffer
> underrun which is reported by aplay..
> The outgoing dataformat is 9600hz, S16_LE, mono ..

I'm having trouble with ALSA intermittently getting into a state
where the trigger is being repeatedly called for start then stop
and can only sometimes get out of it. I will post a seperate report
on this later.

-- 
Ben

Q:      What's a light-year?
A:      One-third less calories than a regular year.


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV