[mythtv] Audio-upmixing questions regarding #5900

Yeechang Lee ylee at pobox.com
Mon Nov 17 19:19:37 UTC 2008

foo bar <foobum at gmail.com> says:
> 5.1 audio is never upmixed (no need). It will be passed thru
> unmolested unless there's timeshifting.

Right. My question was whether native-5.1 audio was being demixed then
remixed, even starting out at 1.0X. You answer that in the next part
. . .

> However, (as before) the AC-3 will still be decoded and then
> re-encoded if we timestretch and then revert to 1.0x. This makes
> very little difference.

Understood. I know the issue of not reverting to the native-5.1 track
when returning to 1.0X timestretch is an existing issue, not one
caused by your patch.

> > * What advantages/disadvantages, if any, are there to upmixing
> >   non-5.1
> >  audio to 5.1?
> Advantages: Sounds nice

So there is an actual sound improvement? How is this done? Does the
upmixer actually take the Dolby 2.0 mock-surround sound cues and
convert them into the appropriate 5.1 tracks through the respective
surround speakers? I'd have thought all upmixing non-5.1 audio would
do is to just turn it into a Dolby 5.1 carrier signal (forgive me for
not using the right terminology) but still with only stereo sound.

> Disadvantages: The upmix/encode to AC-3 costs a few cycles

In terms of performance, my Pentium 4 3.0GHz frontend has trouble with
upmixing 5.1 audio off HD recordings; the normal 55-70% CPU usage
turns into 90-99%.

Frontend:		P4 3.0GHz, 1.5TB software RAID 5 array
Backend:		Quad-core Xeon 1.6GHz, 6.6TB sw RAID 6
Video inputs:		Four high-definition over FireWire/OTA
Accessories:		47" 1080p LCD, 5.1 digital, and MX-600

More information about the mythtv-dev mailing list