[vlc-devel] Re: Update: Mac OS X audio patch (also questions on audio sync)
Jean-Alexis Montignies
ja at sente.ch
Wed Jul 7 13:34:41 CEST 2004
Hi Derk,
VLC should work on a G5 at least for PCM output (no hog mode required
here).
I did notice that in all versions of VLC, sometimes the sound is
chopped. The work around is to disable then enable the audio track from
the menu.
I'm not sure but it may be because changing streams in coreaudio is
asynchronous (so you should wait for the notification to come to be
sure the stream have changed). I don't think the current VLC coreaudio
driver does this.
I think it's not related to the recent changes in coreaudio. The Mark
Well's version behave exactly the same way in this regard.
I'm not convinced that changing the sample rate is a user friendly
feature by default. For me it's more of an annoyance.
In fact my MOTU is synced to my CD player whose clock is at 44.1. And
running VLC when it thinks the interface is at 48khz but it's in fact
at 44.1 does not work very well. So I have to switch manually the clock
sources anyway.
It should, I think the audio resync algorithm from VLC is not very
good. In fact even with the internal clock source at 48, maybe once or
twice in a hour I get an accelerated sound for 1 second.
How VLC does when the output is AC3 as it can not resample?
Jean-Alexis
http://www.sente.ch/
On 7 juil. 04, at 12:00, Derk-Jan Hartman wrote:
> On 06 jul 2004, at 23:21, Jean-Alexis Montignies wrote:
>> It worked great on my setup!
>>
>> Motu 828 mkII , I tried all the speaker configs. (The configs for
>> 3F1R 2F1R were not there, but then it would give a 20 lines menu :),
>> not the greatest interface).
>> For me it's a go.
>>
>> Did anyone tested this on a G5 with optical out configured to output
>> AC3?
>>
>> Jean-Alexis
>
> Don't think so. There aren't much G5 users around. I guess because VLC
> doesn't work....
> One thing has me worried though. i noticed that since a couple of
> weeks ago, sometimes my normal stereo built-in device behaves weird.
> like it has the wrong samplerate or something.
>
> Could folks keep their ears open for such a thing?
>
> DJ
>
>> On 2 juil. 04, at 09:57, Mark Wells wrote:
>>
>>> I sent the following email yesterday afternoon, but for some reason
>>> it didnt come through on the list. I have removed the attachment
>>> incase this was the problem. Therefore if anyone wants the code then
>>> let me know where to send it!
>>>
>>> Mark
>>>
>>> Begin forwarded message:
>>>
>>>> From: Mark Wells <mark.wells at oxsemi.com>
>>>> Date: 1 July 2004 15:13:54 BST
>>>> To: vlc-devel at videolan.org
>>>> Subject: Update: Mac OS X audio patch
>>>>
>>>> Attached is my latest offering - an update to Matt Romaine's OS X
>>>> audio patch. Main functionality includes:
>>>>
>>>> 1) Fixed selection on the "Audio Device" submenu so now when you
>>>> select an option, it actually changes the output format to reflect
>>>> the selected option. (Previously, audio would be played in one
>>>> format only. all available formats were correctly detected, and
>>>> changing the selection on the audio option menu worked, but did not
>>>> actually change the output format!)
>>>>
>>>> 2) Added 5-channel option to Open() so that 3F2R is available.
>>>> (Matt added 3F2R detection in ChannelsToAoutVar() but there was no
>>>> initial 5-channel hardware detection so this menu option would
>>>> previously never happen)
>>>>
>>>> 3) Bug fix so that where multiple devices supporting the same
>>>> number of channels are connected to the system they can be selected
>>>> individually (Previously, selecting "stereo" would indicate in the
>>>> menus that every stereo-capable device was selected)
>>>>
>>>> 4) Error message in InitDevice: "Something may be wrong with
>>>> physical format indexing!" Downgraded to a warning message because
>>>> during testing with my G5's internal device I was getting this
>>>> error message popup, while there is perfect audio playback. I have
>>>> not looked into this issue extensively to see in what circumstances
>>>> this message should occur, but based on my testing it appears when
>>>> the audio stream is actually OK :o)
>>>>
>>>> If you have the available hardware then please test this module. I
>>>> have tested mainly with AC3 content, because I understand from
>>>> Jean-Alexis that the plan is to eventually implement OS X audio
>>>> using the HALOutputUnit (like Xine) but that approach will not
>>>> support direct stream output....so this module will still be
>>>> required even when the new one comes along.
>>>>
>>>> Something else to note is I have been working directly off Matt's
>>>> patch, which used the 20040524 snapshot. I notice that DJ updated
>>>> coreaudio.c in revision 8084 on 28 June, so bear in mind that this
>>>> copy of coreaudio.c DOES NOT include his updates (whatever they
>>>> were...) I guess if necessary I can integrate the two
>>>> copies....that depends how useful this patch is to everyone.
>>>>
>>>> I have tried to comment my changes well, but if there are any
>>>> questions then just shout! Other feedback/suggestions welcome,
>>>> although I am not certain how much more time I will have available
>>>> to continue on this module.
>>>>
>>>> Mark
>>>>
>>>> ----------------------------------
>>>> Mark Wells
>>>> Oxford Semiconductor Ltd
>>>> 25 Milton Park, Abingdon, OX14 4SH, UK
>>> ----------------------------------
>>> Mark Wells
>>> Oxford Semiconductor Ltd
>>> 25 Milton Park, Abingdon, OX14 4SH, UK
>>>
> ---
> Universiteit Twente
> Derk-Jan Hartman (d.hartman at student.utwente dot nl)
> http://home.student.utwente.nl/d.hartman
>
> --
> This is the vlc-devel mailing-list, see http://www.videolan.org/vlc/
> To unsubscribe, please read http://developers.videolan.org/lists.html
> If you are in trouble, please contact <postmaster at videolan.org>
>
--
This is the vlc-devel mailing-list, see http://www.videolan.org/vlc/
To unsubscribe, please read http://developers.videolan.org/lists.html
If you are in trouble, please contact <postmaster at videolan.org>
More information about the vlc-devel
mailing list