MediaPortal Audio Renderer (1 Viewer)

[tourettes]

Not required anymore as the renderer will test few known speaker configurations. So, what are the bad results? Did it burn your house to ground or just play Backstreat Boy's Createst Hits album in non-stop mode? A bit more info would be useful (and a log too).

When the footage start the audio in jerky and the video run very slow.
The situation was so serious that I did not know how to start to explain... sorry

Probably related to that speaker mask. 1st stream format is going nicely, but then audio decoder changes stream format (such can happen without user interaction as decoders aren't always knowing what they are outputting until they receive data). Then somethign goes wrong as the drivers report constantly AUDCLNT_E_UNSUPPORTED_FORMAT.

Too bad that this was found out too late, I have no time to fix it for 1.3.0 alpha (in case if the MPAR is going to be included in it).

 

[red5goahead]

Too bad that this was found out too late, I have no time to fix it for 1.3.0 alpha (in case if the MPAR is going to be included in it).

Ok. As usual I'm glad to help you with the tests when it will possible.
A question: why the channels option in the setup ? Already I set my channel configuration (5+1) using control panel . so MPAR should not follow that automatically?

 

[doveman]

No one is able to bend the physical laws (time in this case). PD needs to have a time machine around if he really is able to use MPAR + timestretching with live tv on 60Hz display. Why? After 1 hour of real time the MPAR would have required already that the broadcaster would have sent 60/50 * 1 hour amount audio and video data. Quite impossible?

Had already disabled timestretch setting at this point following tourettes instructions, the fact my tv plays content without glitches, does not seem related to MPAR anyway

Aw, I was hoping you had a time machine

Isn't much of MPAR's functionality lost with timestretching disabled?

 

[pilehave]

Enabling MPAR on my Core i5 with Nvidia Geforce GT330M raises CPU from a normal range of 9-18% and pushes it up to 30-41%
Doing the same thing on my AMD Athlon X2 240e raises the CPU to 75-100%, making the system unresponsive.

The Core i5 uses LAV filters with CUVID, the AMD uses PowerDVD 12.

Must the computer be restarted when I set EnableTimestretching to 0 (disabled)?

What logs could be useful to investigate this further?

I

 

[tourettes]

Enabling MPAR on my Core i5 with Nvidia Geforce GT330M raises CPU from a normal range of 9-18% and pushes it up to 30-41%
Doing the same thing on my AMD Athlon X2 240e raises the CPU to 75-100%, making the system unresponsive.

Timestretching is quite CPU consuming. Here on 6 channel Blu-ray content quad code AMD Athlon II X4 640 3.0Ghz uses around 40% (althou now the BDReader and MPAR binaries are both debug builds. Release builds should be a bit less).

Must the computer be restarted when I set EnableTimestretching to 0 (disabled)?

No. Restarting MP is enough (actually in theory it should be enough just to change the registry value and then start to play another video file).

What logs could be useful to investigate this further?

Usually not. Only to check if something like 192Khz has been selected as a sampling rate (it surely will kill CPU and provide no actual benefits) or the selected resampling quality setting.

Only good way to check where the CPU cycles are spent is to use some profiling tool. Most likely the major part is spend in sample rate conversion and/or SoundTouch libraries that we cannot affect in anyway.

When aiming to low CPU usage yuo have already spend too much money on the current CPU. Higher CPU usage by MPAR allows higher sound quality in sample rate conversion and/or time strecthing.

 

[pilehave]

OK, thanks for the answers, I was hoping to get lower CPU, my HTPC (the AMD one) isn't very well cooled (quiet fan) so I think I'll stick to the default audio renderer. For what it's worth, there were no dropped frames at all on the Core i5 with MPAR, not even when activating the OSD. The possibility to upmix is awesome

 

[tourettes]

OK, thanks for the answers, I was hoping to get lower CPU, my HTPC (the AMD one) isn't very well cooled (quiet fan) so I think I'll stick to the default audio renderer. For what it's worth, there were no dropped frames at all on the Core i5 with MPAR, not even when activating the OSD. The possibility to upmix is awesome

To get smoother video playback just cool the HTPC better

 

[tourettes]

OK, thanks for the answers, I was hoping to get lower CPU, my HTPC (the AMD one) isn't very well cooled (quiet fan) so I think I'll stick to the default audio renderer. For what it's worth, there were no dropped frames at all on the Core i5 with MPAR, not even when activating the OSD. The possibility to upmix is awesome

Dropped frames itself is not enough to tell the whole thruth of the video playback quality when it comes to smoothness. Video renderer can present some frames on wrong v-syncs and that will lead into juddering. So when reverting mback to DirectShow default audio renderer please check more than just the dropped frames when considering keeping it

In my opinion the upmixing is really useless feature (it came as a bonus after implementing downmixing). If movie contains only stereo track then it is best to listen as it is since it will produce the audio like it was mixed in the studio when making the movie. Upmixing creates just fake sound scape

 

[pilehave]

OK, thanks for the answers, I was hoping to get lower CPU, my HTPC (the AMD one) isn't very well cooled (quiet fan) so I think I'll stick to the default audio renderer. For what it's worth, there were no dropped frames at all on the Core i5 with MPAR, not even when activating the OSD. The possibility to upmix is awesome

Dropped frames itself is not enough to tell the whole thruth of the video playback quality when it comes to smoothness. Video renderer can present some frames on wrong v-syncs and that will lead into juddering. So when reverting mback to DirectShow default audio renderer please check more than just the dropped frames when considering keeping it

In my opinion the upmixing is really useless feature (it came as a bonus after implementing downmixing). If movie contains only stereo track then it is best to listen as it is since it will produce the audio like it was mixed in the studio when making the movie. Upmixing creates just fake sound scape

Most of our tv-channels broadcast in stereo. I think it sounds odd when the dialog from news broadcast ands so comes from left/right, and prefer to have it from my center-speaker instead. Most movies with a digital soundtrack has dialogue in the center-speaker, so I guess I got used to that.

 

[JackTramiel]

OK, thanks for the answers, I was hoping to get lower CPU, my HTPC (the AMD one) isn't very well cooled (quiet fan) so I think I'll stick to the default audio renderer. For what it's worth, there were no dropped frames at all on the Core i5 with MPAR, not even when activating the OSD. The possibility to upmix is awesome

Dropped frames itself is not enough to tell the whole thruth of the video playback quality when it comes to smoothness. Video renderer can present some frames on wrong v-syncs and that will lead into juddering. So when reverting mback to DirectShow default audio renderer please check more than just the dropped frames when considering keeping it

In my opinion the upmixing is really useless feature (it came as a bonus after implementing downmixing). If movie contains only stereo track then it is best to listen as it is since it will produce the audio like it was mixed in the studio when making the movie. Upmixing creates just fake sound scape

Most of our tv-channels broadcast in stereo. I think it sounds odd when the dialog from news broadcast ands so comes from left/right, and prefer to have it from my center-speaker instead. Most movies with a digital soundtrack has dialogue in the center-speaker, so I guess I got used to that.

OT for sure, but if your sound setup works and is placed as it should then the voice should appear to come from a center speaker even with stereo playback.

http://buschmeier.org/bh/study/soundperception/

Back to topic, after a long term test now i can say mpar still works flawless for me.