MediaPortal Audio Renderer (1 Viewer)

[doveman]

Ah, so you dont need to have the both outputs active at the same time. That would work in similar way MPAR - just change the registry settings before starting the playback with similar way as you are now changing.

I do need to have both outputs active at the same time. If I'm using "Stereo to two outputs" it's sending the Stereo signal to the L&R output and the Centre&Subwoofer output at the same time. If I'm using "5.1 mixed down to three channels" it's also using two outputs, the L&R output for everything but the Centre channel content, and the Centre&Subwoofer output for the Centre channel content.

Whilst I could still use ffdshow's mixer profiles to mixdown as I am now before sending it to MPAR, the way ARC works is it loads the ffdshow registry settings for one or the other profile and switches renderer to trigger a rebuild of the graph. Currently it's set to switch between Default Directsound and Directsound: Realtek Speakers, which isn't really a switch at all as they're both the same thing, but the different name is enough to trick ARC/Windows into rebuilding. Obviously I can't do this with MPAR as there's only one instance of it, so I'd have no way of switching the ffdshow profiles.

MPAR will output stereo as stereo if no force mixing option is selected.

Oops, should've been able to work that out for myself from the first post! So I could use ffdshow to mix "Stereo to two outputs" and "5.1 mixed down to three channels" and feed that to MPAR with ForceChannelMixing disabled and it would just output whatever it receives (both profiles actually use four channels/two outputs so I suppose I'd set SpeakerConfig 51). I'd still need a way to switch ffdshow profiles and trigger a graph rebuild (using a remote button) though.

 

[edterbak]

SO. Here it is. I finally got some time to test and do what is asked.
I put together a lot of information. Meant for Tourettes and Red5 to see what might be differenct with his Xonar setup.
I use the ATI HD5670 with CCC 12.3.
SAF 6.10 unlocked (LAV using native dxva)

I did not do a Graphedit check on the actual renderer. It is already in the log and displayed during shift 1 stats. If there is a reason to add it, I will do so ofcourse.

During playback there were a number of droped frames (3or4 I think) There was also an audio hickup around frame ~33200 (+/-200). I hope this helps the search.

If you need anything else, just say so. .





As an extra, I added a screenshot with my display (DRC) set at 50hz instead of 23.976hz.
The result is below. I see or hear no issue. Screenshot does indicate the video is sped up to 24.9fps.

 

[red5goahead]

I don't understand why you add a speaker setup in this version. I can't explain because the setup should be already handled by operating system through audio card driver.

For example to allow up mixing or to allow people to use separate spekaer config in different applications.

I understood.
Imho an "auto" option would be nice to passthrough the original chosen setup in the operating system.

 

[tourettes]

SO. Here it is. I finally got some time to test and do what is asked.
I put together a lot of information.

Meant for Tourettes and Red5 to see what might be differenct with his Xonar setup.

red5goahead could you test same settings as edterbak has?

edterbak, seems that there is something happening on approx every 250 seconds. At least one audio sample gets dropped on that interval. I hope it is related to WASAPIUseEventMode = 0. I'll try to reproduce the issue.

 

[red5goahead]

SO. Here it is. I finally got some time to test and do what is asked.
I put together a lot of information.

Meant for Tourettes and Red5 to see what might be differenct with his Xonar setup.
red5goahead could you test same settings as edterbak has?

Yes. for sure. later . But as I wrote in PM to edterbak he choose 6 ch-> 2ch into Xonar setup. I have 6 ch-> 5+1 instead.

 

[tourettes]

Yes. for sure. later . But as I wrote in PM to edterbak he choose 6 ch-> 2ch into Xonar setup. I have 6 ch-> 5+1 instead.

Trying to rule out what causes the issue would be helpful for the Xonar owners

 

[red5goahead]

SO. Here it is. I finally got some time to test and do what is asked.
I put together a lot of information.

Meant for Tourettes and Red5 to see what might be differenct with his Xonar setup.

red5goahead could you test same settings as edterbak has?

edterbak, seems that there is something happening on approx every 250 seconds. At least one audio sample gets dropped on that interval. I hope it is related to WASAPIUseEventMode = 0. I'll try to reproduce the issue.

two logs.

it works if I put 2 ch on Xonar setup and stereo on MPAR (as edterbak configuration) (with some kind of audio/video lag, anymay not too much)

it doesn't work if I put 5+1 ch on Xonar setup and 5+1 on MPAR

 

[tourettes]

it works if I put 2 ch on Xonar setup and stereo on MPAR (as edterbak configuration) (with some kind of audio/video lag, anymay not too much)

it doesn't work if I put 5+1 ch on Xonar setup and 5+1 on MPAR

21-05-2012 18:59:29.332 [ 136c] REN - applying : 48000Hz 24i (24)bits 6ch -- ch mask: 1551 align: 18 avgbytes: 864000
...
21-05-2012 18:59:29.765 [ 11a8] CWASAPIRenderFilter::NegotiateFormat WASAPI client refused the format: (0x88890008)

Renderer tries to use the correct speaker mask but still it fails. I guess I'll have to add some more logging to find out why the drivers will ignore the audio format.

 

[edterbak]

edterbak, seems that there is something happening on approx every 250 seconds. At least one audio sample gets dropped on that interval. I hope it is related to WASAPIUseEventMode = 0. I'll try to reproduce the issue.

If you need me to test something, you'll let me know? Im happy to helpout.

 

[tourettes]

edterbak, seems that there is something happening on approx every 250 seconds. At least one audio sample gets dropped on that interval. I hope it is related to WASAPIUseEventMode = 0. I'll try to reproduce the issue.

If you need me to test something, you'll let me know? Im happy to helpout.

I managed to get few dropped samples during 2 hour long movie with my dev PC by having WASAPIUseEventMode = 0.

Do you have any other audio device in your PC? It doesn't have to be connected to anything just that if you could test with some device with WASAPIUseEventMode = 1 and provide a similar log. That way we could see a bit more about the issue. When event mode is used the audio HW / drivers are responsible for requesting new data at correct times and then even mode is not used the MPAR itself needs to generate the timer events. That way we could try to limit the error a bit (could be some system wide error if it happens also with the HW provided timing).