Dolby Prologic II Question

[mic j]

I transcode to play my files on either an aTV3 or iPad 2 or possibly iTunes (but have never actually done that). I use the High Profile setting. For audio, the first line I get as the default is Dolby Prologic II. In the next line, I will select my source to be either Dolby passthrough or discreet 6.1 if it is a dts source. I have been changing the Prologic line to stereo, as I would use the iPad or MBP with headphones, therefore, stereo. Am I needlessly changing the setting from Prologic to stereo? Will the the device output be stereo (of good quality) even if the source is Prologic?

[Flo]

Pro Logic is fully compatible with Stereo output. There's no need to change it unless you want to dump the extra channels entirely.

[mic j]

Thank you!

[Deleted User 13735]

Actually, there are no "extra channels." Dolby Surround/Pro Logic uses matrix technology to simulate rear channel playback on so-equipped devices. Unlike discrete surround encoding (like Dolby Digital), the file sizes are identical whether you select Stereo or Surround/Pro Logic.

[Flo]

There are extra channels. They're just phase-shifted instead of discrete and thus aren't as "clean" as true multi-channel formats. But they are extra channels nonetheless.

The size is identical because the bitrate is fixed.

[Deleted User 13735]

No.
One of the existing stereo or mixdown channels is phase shifted 90 degrees to make it easier on the Surround/Pro Logic decoder. The "extra channels" are simulated at the decoder, not the encoder. I actually test this stuff.

[Flo]

The "extra channels" are simulated at the decoder, not the encoder.

The surround channels are phase-shifted and merged into the stereo signal. The decoder extracts them into separate channels.

When transcoding into a lossy format (in this case AAC as we're talking about Apple products), it's important that the encoder knows to preserve surround information, otherwise it might simply dump the phase-shifted information, degrading the surround channels.

[Deleted User 13735]

The surround channels are phase-shifted and merged into the stereo signal. The decoder extracts them into separate channels.

No.
The "surround channels" don't exist!
If they did, you would be able to tell me which is which in the test results above.

Once again, my sincerest apologies to the OP for enabling the discussion to drift so far off topic.
Flo's original statement (sans qualifier) was the correct response.

Pro Logic is fully compatible with Stereo output. There's no need to change it . . .

Best of luck.

[Flo]

The "surround channels" don't exist!

Just because they're not discrete doesn't make them any less real. We're not talking about some sort of virtual surround technology here that fakes a wider sound field, but deliberately designed additional channels. Yes, Pro Logic II decoders can do that, too. But that's not Dolby Surround/Pro Logic as originally intended.

[Deleted User 11865]

No.
One of the existing stereo or mixdown channels is phase shifted 90 degrees to make it easier on the Surround/Pro Logic decoder. The "extra channels" are simulated at the decoder, not the encoder. I actually test this stuff.

While a Dolby Pro Logic decoder will probably try extract additional channels from any Stereo stream ("magic surround"), multichannel sources should be downmixed in a specific way in order to properly and/or better work with it.

Also, AFAICT, phase shifting happens before a channel gets downmixed into Lt or Rt:

In this case, there are four "cardinal" input signals: Left, Center, Right, and Surround (L, C, R, S). The L and R inputs go straight to the Lt and Rt encoder outputs without modification. The C input is divided equally to Lt and Rt with a 3 dB level reduction (to maintain constant acoustic power in the mix). The S input is also reduced by 3 dB, but before being divided equally between Lt and Rt, the signal has 90-degree phase shift applied relative to L, C, and R. Finally, the S signals are carried in Lt/Rt with opposite polarities (note the "-" sign in the summing stage feeding the Lt output).

[Deleted User 13735]

At least with the Dolby Digital Pro encoder, the pre-mixdown phase shift is accomplished with a flag, and it is optional.

90-degree phase shift
Select this check box to apply a 90-degree phase shift to the surround channels. This phase shift is useful when generating multichannel Dolby Digital streams that may be downmixed in an external two-channel decoder.

Perhaps Handbrake is able to detect the presence or absence of this flag, and apply the shift if it is missing. I also know that the phase shift flag is standard with commercial DVD and BD production. I agree that PLII works better with Lt/Rt than Lo/Ro.
It would be interesting to see how this works with AC3 5.1 as well as multichannel discrete PCM, and I'll add some testing to my to-do list.

[Deleted User 11865]

At least with the Dolby Digital Pro encoder, the pre-mixdown phase shift is accomplished with a flag, and it is optional.

90-degree phase shift
Select this check box to apply a 90-degree phase shift to the surround channels. This phase shift is useful when generating multichannel Dolby Digital streams that may be downmixed in an external two-channel decoder.

But you're not encoding Dolby Surround here, this is 5.1.

Is there anywhere you can set the "Preferred Downmix" like in Dolby media Encoder? If set to Dolby Surround or Dolby Pro Logic 2, the decoder is probably expected to apply phase shift and attenuation when downmixing.

When encoding for Dolby Surround (2.0 - Lt/Rt), Dolby Media Encoder doesn't let you disable phase shifting of surround channels.

Also, this doesn't look like a flag to me.

Perhaps Handbrake is able to detect the presence or absence of this flag, and apply the shift if it is missing.

I think it always applies it. But in the case of AC3 sources, liba52 is downmixing rather than HandBrake's built-in downmix, and I'm not sure of what it does.

[Deleted User 13735]

I "think" the virtual downmix from AC3 5.1 is really 2.0 Lt/Rt, assuming the phase shift is there. I'm assuming (perhaps wrongly) that it is essentially the same as what we used to call Dolby Surround. Without the shift, whether the stereo mixdown is really Lo/Ro I don't know??

I dug into this a while back when a friend's DVDs and BluRays came back as unplayable on some machines, and it turned out the absence of the phase shift (in the Dolby Studio AC3 encoder) was the culprit; the discs were reportedly noncompliant. In digging into this, I read the white papers, and believe the phase shift is indeed a flag.

So although this thread has taken some turns, it brings up the question of what the PLII encoder in HB actually does with various sources: Lo/Ro, Lt/Rt, encoded 5.1 with and without the shift, and multichannel PCM. Until proven otherwise, I'll continue on the assumption that it can do no harm, and neither adds to nor discards information from the source, although the decoding at delivery will behave differently.

[SnickySnacks]

Not a definitive test in any way, but downmixing 5.1 to DPLII seemed to lose all the surround channel info in the tests I tried, which is why the ATV2 preset has the option to embed the 5.1 AC3 stream, I suppose.