Is AAC multi-channel quicker to transcode to AC3 or eAC3?

Plex Media Server Desktops & Laptops
,
1

I have a movie with two versions of the same audio track: 1 is AAC 7.1 and the other is a downmixed AAC 5.1 copy.

Neither version is compatible with my devices so the first version is transcoded to eAC3 7.1 and the other AC3 5.1 in Plex.

Which conversion is easier for Plex to deal with?

2

You can probably test this yourself :wink:
For the most accurate comparison you might want to run it through FFmpeg – considering FFmpeg has no user interface you could use Handbrake instead. Create 2 transcodes for each of the scenarios (with the audio source/target being the only difference) and let it run. Afterwards you can compare how long it took.

I expect to take the 7.1 transcoding slightly more time/power (simply because there’s “more to do”). That being said… it’s probably negligible.

3

The best method would be to go back to the source. I am pretty sure that the multi-channel AAC streams are already the results of a conversion from either Dolby or DTS audio.
Converting them yet again will only make their quality worse.

4

Thanks that’s a good idea. I might do that next time I rip a blu-ray with Handbrake.

I’m using a Roku which has a “transcode health” number value during playback. The higher the number, the easier it is on the server it would seem.

5

Thanks and yep you’re right these two tracks were made from DTS-HD MA 7.1.

I think next time I will just create two tracks from the lossless source. A downmixed AAC 5.1 track for surround sound to be transcoded to standard Dolby 5.1 (I only have a soundbar, should I keep all 7.1 channels?) and a downmixed AAC stereo track for direct playing on Roku. Hopefully that should cover all bases compatibility-wise.

6

I recommend you to use the AAC codec only for stereo mixdowns.
(preferably while normalizing the dialog loudness, but that is unfortunately not possible with Handbrake. So you’ll have to revert to other tools.)

A 7.1 stream doesn’t make any sense for soundbar users (even though their marketing wants to make you believe otherwise).

It would be easier, if the soundbar would understand DTS (there is always a “DTS core” component in every DTS HD stream, so a conversion is not necessary – only an extraction).

But if they only support AC3, then you will have to do a conversion, yes.

1 Like
7

Thanks Otto for the advice.

My AVR supports both AC3 and DTS but Roku doesn’t support lossless. The only reason I was considering converting DTS MA to AAC is because the DTS core is so huge and the Roku struggles to receive it.

Maybe I should just keep the 5.1 core though so I can direct play and lose the two extra channels rather than having Plex transcode AAC back to AC3?

8

I usually convert multi-channel audio to a single 5.1 AC3 track. I used to include a stereo AAC track, but most new TVs and other devices handle the 5.1 AC3 track quite well and downmix it to stereo for anyone just using the speakers on the TV/phone/etc.

My experience is that most devices only work on stereo AAC, multichannel tracks (and sometimes even mono) are more miss than hit.

9

This would be my preferred method. But I care very much about sound quality.
You will have to decide for yourself if the results justify the additional bandwidth.

10

Thanks again,

I’m using Staxrip to encode and downmix the lossless audio with QAAC encoder to stereo. Will ticking the “normalize” option do what you meant or will it change things other than dialogue as well?

I’ve tested a couple of tracks with this option ticked downmixing to 5.1 and Stereo.

On the stereo versions with “normalize”, some dialogue comes out slightly louder and some tracks it doesn’t change all that much. The 5.1 is much much louder overall. Maybe this is just a limitation of stereo channels?

11

I can’t tell you that, because I don’t know what Staxrip does there.

I am using XMedia Recode for this (which is actually just a frontend for ffmpeg).
It has a normalization option with a target value you can set in dB. I just use the recommended 89dB. This makes my ripped tv shows’ audio quite consistent (and usually louder) in loudness (without overdoing it).
This process requires a separate run where the loudness of the source is first analysed.
In the second step the conversion is done with the correction factor applied.

The actual difference it causes, depends of course on the source material. You will most often need it, when you downmix surround 5.1 or 7.1 to stereo.

12

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.